TRANSCRIPT OF THE RECORD
THE MARITIME AND COMMERCIAL COURT
On 22 November 1996 the court in the cases
H 185/93 1. Yarc Systems Corporation
(Peter Thykier, Counsel)
(Main Intervener, hereinafter referred to as Plaintiff 1 by the Court)
2. Danrip Int. A/S
(Peter Thykier, Counsel)
1) Flemming Stanley
(Anders Worsoe, Counsel)
2) Stantech I/S rep. by Flemming Stanley
and Peter Tolboell
3) Peter Tolboell
Parties joined as Third Party Plaintiffs:
1) Karsten Jeppesen
(Peter Thykier, Counsel)
2) Erik B. Hansen
3) Stig Mikkelsen
4) Hans Peter Glunz
H 32/94 1) Yarc Systems Corporation
(Peter Thykier, Counsel)
2) Danrip Int. A/S
3) Trevor G. Marshall
1) Flemming Stanley
(Anders Worsoe, Counsel)
2) Stantech I/S rep. by Flemming Stanley
3) Peter Tolboell
Parties joined as Third Party Plaintiffs:
1) Karsten Jeppesen
(Peter Thykier, Counsel)
2) Erik B. Hansen
3) Stig Mikkelsen
4) Hans Peter Glunz
gave the following
J U D G M E N T:
These cases concern partly the question as to whether the Plaintiffs’ and the Defendants’ rip products entail mutual infringements of intellectual property rights (H 185/93), partly the question of whether a number of statements are libellous (H 32/94). Each party has his own Post-Script rip product.
The case (H 185/94) was originally filed by Plaintiff 2, Danrip Int. A/S, against Defendant 1, Flemming Stanley, and concerned the right to use the trademark “Danrip”.
This question was decided upon in the interlocutory judgment of the Maritime and Commercial Court of 28 June 1994 against which no appeal lay while the claims set up in this case in the reply, the writ by which the main intervener joined the case, the third party notice and later were deferred to a later hearing.
In re H 185/93:
In this case Plaintiff 2, Danrip Int. A/S, and Plaintiff 1, (Main Intervener), Yarc Systems Corporation, have at the beginning of the trial set up the following claims against Defendant 1, Flemming Stanley, Defendant 2, StanTech I/S rep. by Flemming Stanley and Peter Tolboell, and Defendant 3, Peter Tolboell:
1) That the Defendants be ordered to cease the copying and distribution of products comprising a copy of Plaintiff 1’s FIFO technology or copies of the parts of the software files “Danrip Super”, “All Boot. Aut.” and “All Main Out” to which Plaintiff 1 has the exclusive rights.
2) That the Defendants be ordered to pay jointly and severally to Plaintiff 1 DKK 100,000 plus litigation interest from the filing of the writ of the main intervener on 30 August 1994.
These claims were withdrawn at the closing of the evidence.
Defendants 1, 2 and 3 have asked for dismissal of the Plaintiffs’ claims and set up the following counterclaims against the Plaintiffs and Third Party 1) Karsten Jeppesen, Third Party 2) Erik B. Hansen, Third Party 3) Stig Mikkelsen and Third Party 4) Hans Peter Glunz.
1) That the Plaintiffs and Third Parties be ordered to cease any direct or indirect production, marketing and sale of RIP products in or from Denmark, especially as has been the case so far under the designations Yarc/Danrip version 1.2, Yarc RIP rev. 1 a and Power RIP/Zumazee. That the Plaintiffs and Third Parties in addition be ordered to recognise that if in future they are to participate directly or indirectly in the production, marketing and sale of RIP products, it can only be done if evidence is produced beforehand - and the market is informed beforehand of the fact that these RIP products are no longer wholly or partially based on the copying of and are in no other way connected to the Danrips belonging to the Defendants.
2) That the Plaintiffs and Third Parties in addition be ordered to pay jointly and severally to the Defendants damages of an amount at the discretion of the court plus litigation interest from the institution of proceedings on 24 August 1993, and likewise that the Plaintiffs and Third Parties be sentenced pursuant to both the Danish Marketing Practices Act and the Danish Copyright Act.
The Plaintiffs and Third Parties 1,2,3 and 4 have against this asked for dismissal of the Defendants’ claims.
Defendant 2, StanTech I/S rep. by Flemming Stanley and Peter Tolboell, has against Plaintiff 2, Danrip Int. A/S, set up the following counterclaim:
That Plaintiff 2 be ordered to pay to the Defendant, Stantech I/S rep. by Flemming Stanley and Peter Tolboell, DKK 302,560.00 plus litigation interest from 1 December 1992.
Plaintiff 2 has asked for dismissal of the Defendant’s claim.
In re H 32/94:
In this case which was instituted on 8 February 1994 and in which the Plaintiffs and the Third Parties with the Defendants’ acceptance during the trial declared that the third party notice filed in case no. H 185/93 also applies to this case Third Party 1, Karsten Jeppesen, has against Defendant 1, Flemming Stanley, set up the following claims:
1) That Defendant 1 be sentenced pursuant to Section 267, Subsections 1 and 3 and/or Section 268 of the Danish Criminal Code.
2) That of Defendant 1’s statements the following are held to be unfounded and that Defendant 1 be ordered to pay an amount fixed by the Court for the payment of the costs of the announcement hereof:
a) “He is of the opinion that Karsten Jeppesen or the firm for which he now works, Rip International, has stolen it.”
b) “For your information our software was stolen from the estate by a person who held a key to the premises. Karsten as one of two had not returned his key: Do you not think that an unusually clear picture is gradually forming of what Karsten is involved in?”
c) “To complete the picture, I inform you that he has now attempted to defraud the bankrupt estate, but for that he will, of course, be called to account as deserved.”
3) That Defendant 1 be ordered to pay DKK 25,000.00 plus litigation interest from 29 August 1994.
Plaintiff 1, Yarc Systems Corporation, Plaintiff 2, Danrip Int. A/S, and Plaintiff 3, Trevor G. Marshall, have set up the following claims against Defendant 1, Flemming Stanley:
1) That Defendant 1 be sentenced pursuant to Section 267, Subsections 1 and 3 and/or Section 268 of the Danish Criminal Code.
2) That of Defendant 1’s statements the following are held to unfounded, cf. Section 273 of the Danish Criminal Code in order that Defendant 1 be ordered to pay to the Plaintiffs an amount fixed by the Court for the payment of the costs of the announcement hereof
a) “Approximately ten days later, a press release was distributed worldwide by a new company called Danrip International, a subsidiary of Yarc Systems Corp., claiming that the company was able to support and further develop the Danrip product! I am sure that any of your readers with only a slight insight into technology will understand that such a claim is either a false marketing statement or proof that the company is illegally in possession of the source code of the DanRIP system!”
b) “I can further inform you that Yarc Systems Corp., with the cooperation of Mr. Peter Glunz, has illegally copied the DanRIP hardware and that the President and CEO of Yarc Systems Corp., Mr. Trevor Marshall, has admitted to the receiver at a meeting that this action had indeed taken place.”
c) “Without anticipating events it appears from the course of actions in the enclosed letter that the managing director of DanRIP International A/S, Mr Peter Glunz, and the CEO of Yarc Systems Corporation in the USA, Mr Trevor Marshall, must have copied the property of the bankrupt estate before the rights were put up for sale.”
d) “We keep up and insist even though the result is a most unusual abuse of a court in which such drastic measures as e.g. perjury before the Superior Court of the State of California are used (documentation of such an allegation is of course available).”
e) We believe that DanRIP International A/S illegally has obtained possession of DanRIP hardware and the source code. A few days prior to the sale of the rights nine DanRIP boards as well as the source code for the program were removed during a break-in at the Stantext estate and we assume that these items are now with DanRIP International and Yarc Systems Corporation.”
f) “According to a press release from Plaintiff - but with the written clause that it is “not to published before Seybold, oct. 21, 1993” distributed at the IPEX93 to various companies present at the exhibition, Plaintiff has now “again developed” another DanRIP with two processors. Although Plaintiff by this “new innovation” tries to hide away the copied DanRIP board into three “proprietary ASICs - Defendants have from the above already stated good reasons to believe in the light of the above evidency that it is still - Defendants board/assets which Plaintiff has illegally copied. For your information Defendants shall emphasize that the DanRIP project so far has requested 25 man years of development. Plaintiffs now claims that they just managed to develop a system with the same features in a few months!”.
g) “Rights “stolen”
The two new sole owners (!) of the DanRIP project now knew that Mr Trevor Marshall and Mr Peter Glunz had illegal possession of all their hard-earned ownership rights and when they immediately thereupon read that even the world-famed graphic art magazine “Seybold Report on Publishing” - on the basis of a press release from Yarc and DanRIP Intl. - announced to the entire graphic art world that it was now the Yarc/DanRIP Intl. partnership that owned and carried on the DanRIP project, Flemming Stanley wrote a letter to the main editor of Seybold, Mr Stephen Edwards, saying the he and his readers had been defrauded. Furthermore, with relevant documentation prepared by the liquidator he wrote that the DanRIP belonged to the two new Danish owners alone in the form of a company in the course of formation called StanTech.”
3) That Defendant 1 be ordered to pay to the Plaintiffs DKK 100,000.00 plus litigation interest from 8 February 1994.
Plaintiff 1, Yarc Systems Corporation, and Plaintiff 2, Danrip International A/S, have against Defendant 1, Flemming Stanley, Defendant 2, Stantech I/S rep. by Flemming Stanley and Peter Tolboell, and/or Defendant 3, Peter Tolboell, set up the following claims:
1) That Defendants 1 and 2 be sentenced pursuant to Section 19, Subsection 3 of the Marketing Practices Act then in force (now Section 22, Subsection 3), cf. Section 2, Subsection 2.
2) That Defendants 1 and 2 be prohibited, cf. Section 14 of the Marketing Practices Act then in force (now Section 13, Subsection 1) from contacting the Plaintiffs’ actual and potential customers in contravention to the provisions of the Marketing Practices Act and make derogatory and offensive statements concerning the Plaintiffs towards these customers, including allegations that the Plaintiffs have committed criminal offences.
3) That Defendants 1, 2 and 3 be ordered to pay jointly and severally to Plaintiffs 1 and 2 damages of DKK 600,000.00 plus litigation interest from 8 February 1994.
Defendants 1, 2 and 3 have against this asked for dismissal of the Plaintiffs’ claims.
The facts of the case:
Plaintiff 1, Yarc Systems Corporation (Yarc) which is registered in California, was established in 1988 with Trevor G. Marshall as the principal shareholder and CEO.
Plaintiff 1 has since then been engaged in the development and sale of rip-related products to a large number of firms all over the world.
Plaintiff 2, Danrip Int. A/S, which was originally a shell company was upon agreement between Hans Peter Glunz and Flemming Stanley acquired at the end of 1991. From the transcript of the abstract form the companies’ register it appears that the company’s secondary name is Danish Raster Image Processing Technology A/S (Danrip Int. A/S). The company’s shareholder and managing director was Third Party 4 Hans Peter Glunz who with Flemming Stanley and attorney Peter Thykier formed the Board of Directors. In April 1993 Plaintiff 1 acquired 60% of the share capital while the remaining part is owned by Glunz ApS holding 11%, by Third Party 1 Karsten Jeppesen holding 10%, by Third Party 2 Erik B. Hansen holding 9.5% and by Third Party 3 Stig Mikkelsen holding 9.5%.
Defendant 1, Flemming Stanley, was the managing director and owner of Stantext A/S which since 1986-87 and until its bankruptcy in 1992 was engaged in hardware and software for the graphic art industry, including RIP products.
RIP is a relatively complicated computer product consisting of hardware as well as software and the purpose of which within the graphic art industry is to convert from computers high resolution images, text and graphics, to reproduction ready film for subsequent printing.
Defendant 3, Peter Tolboell, was engaged as development engineer with Stantext A/S and following the bankruptcy of Stantext A/S he and Flemming Stanley formed Defendant 2, StanTech I/S, which is engaged in the same line of business as StanText A/S.
The relationship between Plaintiff 2, StanText, and Defendant 1:
In the month of July 1991 Hans Peter Glunz entered into an exclusive distributorship agreement with StanText A/S.
PG’s exclusive distributorship shall concern the sale and distribution of the Post Script RIP described in Appendix a for application on Macintosh platforms.
THE EXTENT OF THE EXCLUSIVE DISTRIBUTORSHIP
The exclusive distributorship shall include all sale and distribution of the product nationally and internationally, subject to the condition, however, that StanText shall still on the terms and conditions stated in §3 be entitled to sell directly both to private label customers and to …………
SPECIAL OBLIGATIONS AND RIGHTS
b) PG hereby places a binding order and so commits himself to buy 10 units of the product for delivery no later than 1 month after the approval of the product, and further 15 units no later than 14 days after the IMPRINTA exhibition in February 1992.
d) PG commits himself to form a limited company, hereinafter referred to as Glunz Computer Systems A/S, with a capital of DKK 300,000.00 and shall transfer to said company all rights and obligations under this Agreement. In addition, PG commits himself until the end of 1992 to use the major part of this working hours on the distribution and sale of the Product.
e) StanText shall at any given time be entitled to purchase up to 67% of the shares in Glunz Computer Systems A/S at a price calculated on the basis of the equity of the company less goodwill, however at least at a price of 100 …………
f) StanText commits itself to provide PG with security on the product. This is done by StanText’s deposit of the source code and any necessary documentation of production with Stanbrook & Partners ………… appended instructions that the code may be handed over to PG for his unrestricted use if StanText fulfils the terms and conditions stated in §§10 a) and b).
BREACH OF CONTRACT
Either party may terminate this Agreement without notice in the event:
a) of the other party’s material breach of this Agreement and this party fails to remedy the breach within 3 months‘ written notice after having been required to do so, or
b) the other party goes into bankruptcy or suspends its payments.
On termination of this Agreement in either the whole of the territory or in parts, PG shall sell and return to StanText and StanText shall purchase from PG all unsold marketable and undamaged products for which PG shall be refunded at the original cost price. Unmarketable and damaged products shall be returned and refunded at the valuation price. All costs in connection with the return of the products shall be shared equally between the parties.
By letter of 6 February 1992 Hans Peter Glunz confirmed in writing that StanText A/S was entitled to use the designation “Danrip” for their product called “Mac-the RIP” and that the terms and conditions contained in the distributorship agreement entered into concerning “MAC-the-RIP” applied correspondingly to the product “DanRIP”.
In the period from 12 January 1992 till 30 June 1992 Danrip Int. bought a number of Danrips from StanText A/S. The last delivery, which comprised 13 units, was forwarded on 14 May 1992. Hans Peter Glunz was in the USA at that time and had Stig Mikkelsen inform StanText that Danrip A/S had not ordered these units and did not wish to take delivery of further rips. Of these 13 units, 9 were returned. The purchase price of the remaining 4 is DKK 302,550 which forms the basis of the Defendants’ claim.
However, problems arose as to the rips supplied which had the result that Danrip Int. A/S did not pay in accordance with the parties’ agreement which got StanText into financial difficulties. A number of meetings were held between the parties and their attorneys. During these meetings StanText claimed that the exclusive distributorship agreement was no longer in force and that StanText would possibly grant the exclusive rights to other distributors. Furthermore, StanText requested that the 13 Danrip systems be returned, 9 of them were returned.
On 13 July 1992 StanText A/S entered into a distributorship agreement with CVS Fotoclark, Hannover, covering parts of Northern Germany.
On 30 July 1992 Hans Peter Glunz on behalf of Danrip Int. signed a draft agreement.
In his letter of 3 August 1992 to Danrip Int.’s attorney Stantext’s attorney commented upon the agreement and forwarded a revised draft. From the letter it appears among other things:
Enclosed please find a framework agreement between Danrip Int. A/S and STANTEXT A/S signed by Flemming Stanley.
The following changes have been made to the agreement:
§ 1: Like the former agreement between DANRIP INTERNATIONAL and STANTEXT this Agreement includes DANRIP based on Mackintosh platforms. In the event that the exclusive distributorship agreement is to be extended to cover other products which might be produced by STANTEXT in the future such shall be subject to separate negotiation.
§ 3: The wording “as compensation for the technical problems arisen so far” has been replaced by the wording “it has been agreed that”. It is not my view that the agreement embodied in § 3 is a compensation for the technical problems arisen so far. The agreement is an indication of STANTEXT’s kindness towards DANRIP INTERNATIONAL in order to help in the situation in which DANRIP INTERNATIONAL has not sold the first version of the DANRIP.
§ 5: It has been added that the distributorship agreement is entered into based on the exclusive distributorship agreement forwarded by me on the basis of the prices fixed therein. I think that this reflects the agreement we made at the meeting on 30 July 1992.
In the last and second last paragraphs certain amendments have been inserted upon request from STANTEXT A/S.
Besides, I assume that these questions as you also stated at the meeting are made the object of separate provisions in the final distributorship agreement.
In continuation of the minutes from the meeting of 2 July 1992 prepared by attorney Anders Torboel and referring to the meeting on 30 July 1992, Danrip International A/S, Hulvejen 20, 2680 Solrød Strand (hereinafter referred to as Danrip) and Stantext A/S, Sindalsvej 30, 8240 Risskov (hereinafter referred to as Stantext) have entered into the following agreement on the sale and distribution of Danrip products:
Upon entering into this Agreement Danrip shall waive its exclusive distributorship in the rest of the world apart from the territory mentioned above and shall supply to Stantext a list of customers/distributors with whom Danrip is in contact in countries other than Denmark, Norway and Sweden.
At the same time Stantext shall hereby assume full responsibility for the continued servicing of existing customers/distributors and shall assume full responsibility for any present and possible future complaints from these customers/distributors.
Accordingly, Stantext may raise no claim whatsoever against Danrip in connection with complaints or the exercise of remedies for breach from customers or distributors/agents.
It is agreed that Danrip shall be entitled to keep 4 Danrips nos. 31, 31, 34 and 37 free of charge, of these 2 have been installed with customers.
In the event that Danrip in the period until the end of 1993 places an order for an additional Danrip-maxi unit, the price shall be equal to the price of 10 units applicable to other distributors, however not exceeding DKK 93,520.00.
Coincidental to the supply of this maxi Stantext shall commit itself to supply a maxi free of charge.
The above agreement shall constitute a framework agreement enabling the parties to enter into an actual exclusive distributorship agreement as soon as possible in order to regulate the specific terms of the exclusive distribution based on the draft distributorship agreement forwarded to attorney Peter Thykier and on the basis of the prices fixed therein. ……
The framework agreement was signed by Flemming Stanley only. Shortly thereafter problems occurred and a number of meetings were held. During these meetings restrictions in the geographical territory and the extent of the agreement were among other things discussed. The parties continued, however, their cooperation.
Concurrent with the discussions on a revised agreement between the parties a change in the price of version 1.1. was discussed as well. This caused Flemming Stanley to write the following letters of 19 September 1992 to Hans Peter Glunz:
At the beginning of July we made an agreement which among other things included you getting 5 boards as “compensation”.
You were of the opinion that you had suffered a loss on version 1.1. of the board and that the only way in which you could have the loss offset was by selling your stock at Mini prices.
Even though I did not share your view I agreed to your wish and went straight home and credited you with DKK 242,048 + VAT under credit note no. 1992-249a of 30 June 1992. At the same time we promised to supply you with one unit free of charge when you bought a new one.
Yesterday I had a conversation with Stig who informed me that now almost 3 months after our agreement no reduction in the prices has been effected by Danrip Int. A/S. Any attempt at selling version 1.1. was made on the basis of the original prices, or even at the new increased prices!
In the light of this I can only note that we are dealing with crediting on a wrong basis and unless you immediately start the sale of your old stock at Mini prices we will invoice you again for the boards credited.
On 5 October 1992 Flemming Stanley wrote the following letter to Hans Peter Glunz of Danrip Int. A/S concerning their cooperative relationship:
Referring to our telephone conversation just now, I cannot keep from confirming your own statements made on the basis of the situation at the “Helsingborg Dagblad”:
To start with the conclusion of our conversation, you found it to be absurd that “Helsingborg Dagblad” does not continue to run the installation we lend them for testing (“it is better that they run the system instead of leaving it in the cupboard”.)
Yes, you’re quite right. But we find ourselves in the peculiar situation that we are the ones who find the customer, who make the installation, and who lend a system - while at the same time over and over again are hearing you say that you are definitely not going to buy anything if you can avoid it (but you still want a distributor’s commission!)
We entered into a cooperative relationship which you breached. We met your wish to establish yourself as a distributor in Scandinavia instead. As part of the agreement we credited you with a considerable amount conditional upon your selling old boards at cheap prices. Still, you have - contrary to our agreement - kept up your prices of the old boards and I have received no response whatsoever to my letter of 19.9.92.
Our relationship has been troubled lately. At the worst, when you complained about an installation process in Sweden and we later found out that Stig - from whom you had your information - thought the whole process to be an absolute success. This is not the most productive basis of a continued fiduciary relationship.
I think we should stop pretending. We have tried to support you in order to create a new structure in Scandinavia but we are forced to realise that even though you know - and are confirmed hereof - that also our latest and so far best product is running perfectly you will under no circumstances take part in the marketing of it unless it is at our cost.
The above is also to be seen in the light of the fact that we understand from Stig that you no longer actively participate in the marketing of the Danrip, but have started other activities.
With regret I have to inform you that if things continue as they are now for just a few more days, we think that it is time to find other alternatives to our distribution problems and in that case we will take the measures necessary to alter the cooperative relationship drastically.
On 23 October 1992 StanText A/S forwarded an invoice for 4 units of the Danrip line nos. 00031.00032. 00034 and 00037 - totalling DKK 302,560. This caused Hans Peter Glunz to sent fax letter of 27 October 1992 from which it among other things appears:
“I have with astonishment received your invoice of 23 October 1992 and hereby reject its legitimacy.
As I explained to you at the beginning of October, it is you who fix the resale prices (the official price list) and not me. I cannot just reduce the prices in some price list as I all the time have to count on being able to buy from you if the customers later or right away need an accelerator or if I have to be able to match your rates of speed (or if you make up a new price list).
When I asked you what reduction in the prices would satisfy your expectations you simply answered. “I don’t know” - shall we continue this game of guessing?
I can inform you - it might set you at rest - that so far I have not sold a single DanRIP at full maxi price. They have been sold either as Minis, sales on which I have not earned a krone, or at an even lower price if it has not been to a distributor.
I have only invoiced a distributor one Maxi at the ordinary price - but it has not been paid for due to continued problems.”
StanText A/S went into bankruptcy proceedings in December 1992 and attorney Soren Pedersen, Arhus, was appointed as liquidator of the estate. On 15 December 1992 Danrip Int. A/S’s attorney wrote the following letter to StanText A/S’s attorney - a copy hereof was sent to Flemming Stanley - concerning the termination of the agreement entered into previously.
My client, Danrip Int. A/S, has informed me that StanText A/S yesterday presented a petition for its winding-up. In this connection I kindly ask you to either confirm or deny this piece of information.
If a petition has been filed, I hereby on behalf of Danrip Int. A/S terminate the exclusive distributorship agreement entered into with StanText A/S on 22 July 1991.
The termination is based on clause 10, item b according to which either party is entitled to terminate the agreement without notice if the other party goes into bankruptcy or suspends its payments.
Under clause 5, item f of the same agreement StanText A/S has deposited with you the source code and the necessary documentation of production.
This material is to be handed over to Mr Peter Glunz (now Danrip Int. A/S) for his unrestricted use and in the light of this, I kindly ask you to send or deliver to my office the material as soon as possible.
Finally, - when a provisional liquidator has been appointed - I ask you to inform me of said liquidator’s name.
For the sake of good order, I draw your attention to the fact that Danrip Int. A/S thereupon intend to further develop the Danrip project, however it is not yet finally decided what shape such further developing will take.
By letter of 23 December 1992 StanText A/S’s attorney informed Danrip Int.’s attorney “that the agreement to which you refer in your letter is no longer in force due to your breach and was cancelled on 30 July 1992.
It appears from the records of a meeting of 6 January 1993 on the examination/assessment procedure of the bankrupt estate of the StanText product Danrip:
Description of the system:
Danrip is a system which is used within the field of electronic handling of text/images. The purpose of the product is to increase the speed of production of photo typesetting of large complex objects at high resolution.
The technical designation for systems of this type is “RIP” (=Raster Image Processor); these processors are found in standardised forms in e.g. laser printers, however with considerably less capacity.
Acquirers of systems of this type are e.g. suppliers of types, printing works and the like.
According to Flemming Stanley 6-9 systems have been sold, each at a price of approx. DKK 75,000.
The system consists of an electronic unit (add-on board) for installation in a computer and a computer program. These elements are to function inseparable.
As the producer of the system Danrip StanText A/S has designed the electronic unit and developed the computer program associated with it. This program is with StanText A/S in the form of a source code. On selling a system the purchaser acquires a copy of the add-on board and the translated (compiled) program. Thus, the acquirer has no possibility of making changes/additions to the mode of operation of the program.
According to information from managing director Flemming Stanley Nielsen the source code of the computer program consists of more than 100,000 program lines. The program has been developed in the programming language C. Furthermore, Flemming Stanley Nielsen says that there is no available description of the program.
On request, managing director Flemming Stanley Nielsen presented parts of the source code which were found in a computer at the firm. No describing comments were found herein.
Under normal circumstances a complete description of the mode of operation of a program is available, from the overall general objectives to the executing component parts. Usually a description of a program consists of a mixture of descriptive text and illustrations as well as the programmer’s comments in the source code itself.
The future value in use of a program is to a large extent dependent upon this description as such a description makes it possible for others apart from the programmer(s) who made the program code itself to use the program. To further develop and maintain an undocumented program of just a few thousand lines of program code is difficult for the originator and for outsiders virtually impossible. Often it is easier to remake the program.
On assessing the value of the product Danrip a decisive factor is the fact that there is no description of the associated program source code. A possible would-be buyer is - talking of software of this size - in no way capable of further developing the product himself. And in the light of the development within the computer business it must be assumed that the system is soon outdated. It is therefore assumed that there are no real would-be buyers of the product.
The product may only be valuable to the persons who originally developed the computer program as this group of persons is the only one that might expect to maintain/further develop the product.
On the information presented, my best estimate is that the product is worthless.
Senior Consultant, UNI.C Arhus
On 13 January 1993 attorney Peter Thykier wrote the following letter to Flemming Stanley:
I am informed of your fax letter of 6 January to Bo Hagelquist of Bregwalls Grafiska.
Even though your false allegations hardly will cause any major damage as all my clients of course know of the truth of the matter, for the sake of good order I still have to make reservations as to claim for damages against you should your conduct - against what we expect - result in a loss for me client. However, as mentioned, I do not think that this will prove to be necessary as your scribbling speaks for itself.
I must, however, emphasise that you are not entitled to “pursue further the product” as - as you are well aware - my client asserts that it is Danrip that holds the right hereto.
The parties do no agree on this issue but to start with it is for the bankrupt estate to decide what is to happen in relation to the product, and the right to this product depends on negotiations or alternatively, legal proceedings between the bankrupt estate and my client.
In all circumstances this is not a question which in any way concerns or is any of the individual Flemming Stanley’s business.
On 24 January 1993 the same wrote to the liquidator of StanText A/S’s bankrupt estate concerning the exclusive distributorship agreement:
The existing 2 draft framework agreements dated and signed 3/8 and 30/ 1992, respectively, are not, as you will see, of the same wording which obviously proves that we are talking of drafts only. The parties never agreed on a framework agreement. For this reason alone the original exclusive distributorship agreement is still in force. To further substantiate this I can inform you that under Clause 12 of the original agreement Stantext is on termination of the agreement under an obligation to buy my client’s stock which evidently has not happened either. Also, this reassures me that the original agreement, naturally, is still in force. On the phone you mentioned that it was a possibility that documents and the source code would be handed over to my client and that you in addition sold “the project” to a third party.
I do not agree that this is a possibility …………”
In his letter of 2 April 1993 the liquidator encouraged Flemming Stanley to make an offer for the assets of the estate, including the source code belonging to the estate and additional tangible rights and intellectual property rights. Especially as regards the source code objections have been raised by third party against the rights of the estate.
Bids would be opened for on 15 April 1992 and the sale would be effected to the highest bidder.
Attorney Peter Thykier objected in his letter of 7 April 1993 to the liquidator and I emphasised “that the name Danrip belongs to my client, that the source code be handed over to my client and that Yarc Systems Corporation holds the rights to a considerable part of the software which the estate now intends to put up for sale …………”
In a fax letter of 14 April 1993 Flemming Stanley informed Hans Peter Glunz that he would the next morning issue a press release, of which a copy would be forwarded, and asked for Hans Peter Glunz’s comments as he stressed that he did not wish to misinform.
From the press release of 14 April 1993 among other things appears:
“Unusual methods are used to acquire the rights in the estate following the bankruptcy of a computer firm in Arhus
The consortium has, also according to Flemming Stanley, repeatedly committed forgery in order to “prove” that he allegedly had fraudulently abused his position against StanText A/S …………
Introduction to “the StanText Story” (detailed description enclosed):
5. Following the initiation of the administration of the estate a former supplier, Yarc Systems Corp. in California, makes an attempt at persuading two of the shareholders of the former firm to steal the firm’s most valuable asset, the source code, and through the back-dating of documents to enter into a binding agreement to secure Yarc the future development and marketing rights. As this attempt is disapproved, a close cooperation is instead begun between Dansk Ingeniørforenings Pensionskasse (The Association of Danish Engineers’ Pension Fund) and Yarc, partly with the purpose of placing the former managing director of StanText A/S, Flemming Stanley, in an overt position of wilful mismanagement by forwarding documents, prepared for the occasion, to a Danish bank, Den Danske Bank, and the liquidator.
6. As attempts from Scanview A/S rep. by managing director Shamir Lehaff but especially repeated attempts from the Glunz-Yarc alliance are not immediately successful, use is made of premeditated forgery.
Yarc, Stantech rep. by Flemming Stanley and others had made offers for the effects which were sold on 15 April to StanTech rep. by Flemming Stanley as the highest bidder. At the request of Flemming Stanley the liquidator confirmed in his letter of 28 May 1992 that the purchase among other things included “the rights to the source code, tangible rights, and intellectual property rights belonging to StanText A/S in bankruptcy, including the estate’s rights to the Danrip project, name rights, etc.”.
Probably the same day before the effects were picked up by Flemming Stanley some of them were stolen, including the source code and the back-up concerning the Danrip. The theft was reported to the police. Karsten Jeppesen who still had a key to the premises where the effects were stored on his own initiative called the police and told them that he had a key but had nothing to do with the theft. The theft was never solved and during the investigation no charges were brought against any person of which Plaintiff 1 was confirmed by fax letter of 5 January 1994 from the police in Arhus.
On 19 April 1993 Flemming Stanley gave a receipt for the surrender of the following effects:
1 Mac II cx, 1 screen for Mac the Q, 1 keyboard for Mac the Q and the source code.
On 30 April 1993 Danrip Int. and Yarc issued the following press release on the occasion that Danrip Int. was turned into a subsidiary of Yarc:
Today agreement has been entered into that DanRIP International A/S, Roskilde, Denmark is turned into a subsidiary of YARC Systems Corporation, Newbury Park, California, USA.
YARC Systems is the world leader in supplying RISC based 3D graphic art systems for Macintosh computers. In addition, YARC Systems is OEM supplier to the Pro-Press industry, including among others Linotype-Hall, Dalnippon Screen and Agfa Compugraphic.
DanRIP was established in 1991 by Hans Peter Glunz, one of the founders of Glunz & Jensen International A/S (Ringsted, DK), to market the DanRIP product developed by StanText A/S.
Recently, StanText A/S was ordered to be wound up and has since then been unable to support the DanRIP product and since then it has neither been possible to develop further the product.
The YARC RISC board was previously marketed under the name “DanRIP Maxi Super” which by graphic art experts with “Seybold Reports” recently was characterised as the world’s fastest RIP. YARC’s engineers supplied all the acceleration technology which brought the performance of the DanRIP Maxi Super to a level of international class.
“The alliance with YARC will make it possible for us to use their technical know-how. At the same time we are able to support the DanRIP product in a way which is satisfactory to our customers”, says Peter Glunz. “This alliance has also made possible that we supply to our customers the RISC based version at a considerably more favourable price”.
Dr. Trevor Marshall, YARC’s chairman of the Board of Directors and CEO, says that the intention is that DanRIP as well as YARC both are to continue the research and development of DanRIP’s product program.
Erik B. Hansen, 44, has been engaged as administrative and marketing manager with DanRIP. Erik comes from a position as sales and marketing coordinator with Glunz & Jensen International A/S and carries with him considerable experience in the marketing of Pro-Press products. Stig Mikkelsen continues as sales and support engineer and also as to service DanRIP is in the capable hands of Karsten Jeppesen who will assist as regards customer support.
Karsten was previously StanText A/S’s leading software engineer. Karsten will of course be closely linked to the development of future DanRIP products.
Referring to an invitation to visit a fair issued by dan-grama a/s from which it appeared that the product name Danrip was used for a software-based RIP produced by StanTech, Yarc/Danrip Int. stated the following in their letter to dan-grama a/s of 12 May 1992:
At the same time we kindly point out that the description attached to the invitation - besides using our name here as well - is full of incorrect statements. Among other things implementations are mentioned to which StanTech A/S neither has nor will obtain access. The speed at which the RIP referred to in the description is to run implies the implementation of a board which is produced by our parent company YARC Systems Corporation, USA. Due to a major loss YARC Systems suffered as a consequence of non-paid invoices for deliveries of boards to StanText A/S, StanTech A/S will of course not be allowed to buy such boards which are necessary to obtain the speed described.
On 9 June 1993 the firm Stephen Austin, which is one of StanText’s former customers, wrote the following to Flemming Stanley:
Further to our recent communications, I would formally advise you that Stephen Austin and Sons Limited will not pursue further the DANRIP system.
We had agreed to go ahead with DANRIP and for you to install and run on a trial basis for one month. I have to say now, that this company is not confident that your system is adequate for our needs, nor are we satisfied about the situation related to ownership and rights of the DANRIP system, a matter of extreme concern to us.
With such gulf in confidence, clearly there is no future in continuity of this order an I reiterate that we shall terminate the arrangement forthwith. I fully understand your disappointment at our decision but the decision has been taken regretfully but of necessity.
In his letter of 21 June 1993 Danrip Int.’s attorney objected to StanTech’s use of the name Danrip and to the application of software belonging to the parent company Yarc and which is provided with the message “© Copyright Yarc Systems”. He requested further information as to whether software belonging to Yarc was still being used and pointed out that if that was the case such use was not lawful unless an explicit licensing agreement had been made with Yarc which according to information available was not the case and that the acquisitions from the estate did not include the right to use the estate’s license rights.
At the request of the Plaintiffs’ attorney the liquidator confirmed on 24 June 1993 that no transfer of the license rights had taken place.
In his letter of 6 September 1993 Yarc/Danrip Int.’s attorney in Denmark communicated to Publishing Technologies Ltd, England, which is one of the Defendant’s customers and cooperators, that the name Danrip and the product belonged to his client.
The relationship between Yarc and StanText:
From the correspondence and agreements between Yarc and Yarc’s customers, including Agfa, Advanced Micro Devices, Microsoft and Linotype it is evident that since 1990 Yarc has been engaged in the development of RISC processors for accelerators, but also developed a large number of rip-related hardware and software products based on Unix, PC and Macintosh platforms. It appears from a letter of 10 September 1190 that the plan was that Yarc should develop a rip product for application in IBM compatible PCs and Macintosh computers.
StanText had in 1987-88 developed the Skyline product which was based on a Unix platform and was a separate hardware product. From mid 1991 till the end of 1991 StanText developed Mac the Rip version 1.1 which consisted of a circuit board for placing in a standard host computer and was based on a Macintosh platform.
From the end of 1991 till the summer of 1992 StanText A/S further developed a new version of Rip - called Danrip version 1.2.
On 25 June 1991 StanText A/S sent the following fax letter to Yarc:
Referring to a description of your RISC Coprocessor in the Electronic Product News we ask you to send us by fax all information on your product range, including technical info, prices, discounts, etc.
This was followed by a letter of 19 July 1991 from Flemming Stanley to Jody Champion at Yarc from which it appears:
We are in the PostScript business. We have developed and installed UNIX based systems for the high end-users based on VME cards from Du Point Pixel and are currently porting our system to the Macintosh platform.
We have read 6 pages about YARC and your coprocessors and would like to investigate further in a business relationship.
We would appreciate to receive the information for design of daughter cards mentioned on page 3 and any information mentioned on the last page about “PostScript Printing With YARC”.
We would also like to learn about your plans for porting the coprocessor to other platforms (SUN, RS 6000 and the PC).
A quotation on an OEM basis will be most welcome. For the time being our top-priority is the MAC.
Bob Moll asked on 24 July 1991 Flemming Stanley to communicate whether StanText had developed its own PostScript translator or whether StanText had entered into a licensing agreement to this effect.
On 30 July 1991 Bob Moll wrote the following to Flemming Stanley:
As you inquired in your first fax, we do have our boards running on the PC, and in fact we have used the PC platform almost as long as the Mac platform. Quite a few of our OEMs use the PC based systems. As to the other two platforms you mentioned, there has been no interest from our customers up to date for the RS6000, and only a few inquiries concerning the SUN stations.
You also asked for some daughtercard design information. We do have general information available, such as register descriptions, timing diagrams etc., and if you could let us know your area of interest (high-speed printer interfaces. SCSI. and so on) we could perhaps send you an example of one of the boards we have designed for our customers. Please let me know what you need, so that I can send you the relevant information. Also, I would be very interested in any information you can give me concerning the market you operate in, and any of your marketing materials such as specs of your systems would be really appreciated.
On 31 July 1991 Flemming Stanley wrote the following letter to Bob Moll at OEM Development, Yarc:
In our previous system called SKYLINE, the UNIX and WME based system, we used a Task Processor in order to improve speed. We had three boards: The UNIX board, the Task Processor and the RIP.
In our current project we use the Mac and the RIP currently being developed should be able to match the previous system in speed due to the design implemented. However we anticipate that some customers would like to have even more speed. Consequently we are looking for cards like yours to be implemented after presentation of our new system which is planned for the IMPRINTA exhibition in February 1992 in Dusseldorf.
The RIP will drive a Linotronic and then we will implement another 3-5 recorders one of which will be a broadsheet laser printer. After having implemented these recorders we will port the system to the PC, the SUN and the RS6000. In which order we don’t know yet. It might await customer push but I personally think it will be the SUN.
I think the best would be if you called our engineer Peter Tolboell early in the morning to discuss technical matters. If you don’t find him try +45 86 10 11 68.
We will operate in the high end of the graphic arts industry for which we have a proven product that performs very well as well in quality as in speed. BUT we want the concept at a standard platform in order to avoid the usual marketing credibility problem you are faced with if you bring our “own” computer to market.
Our interpreter is heavily dependent on as well floating point as interger computations. We use Think C as the compiler for the Mac.
Hopefully our Mac-based RIP will be available for installations in December and from around February 1992 we expect the whole market to be very aware of the product/concept.
What raised our interest in your material was the fact that it looked as if we could take advantage of the tools offered in order to obtain maximum speed …………”
On the basis hereof Trevor Marshall visited StanText on 30 August 1991. He participated in meetings on the development of the Mac-the-Rip product as well as the design of a nubus circuit board. Based on this, the parties started a cooperation which was confirmed in the following correspondence.
By letter of 18 September 1991 Flemming Stanley informed Harvey Raider, Yarc, that StanText was interested in examining the possibilities of combining Yarc’s hardware and StanText’s RIP/software and that they counted on being able to install “Mac-the-Rip” at their customers’ by the end of November. As soon as the product was made to run, StanText would incorporate Yarc’s hardware in order to improve the speed with a view to presenting it at the IMPRINTA exhibition in Dusseldorf in February 1992.
On 29 September 1991 Trevor Marshall from Yarc faxed to Flemming Stanley:
Last week Mitsubishi approached us about producing an embedded RIP for their new Dye Sub printer. Without an NDA I can tell you it uses an 11x17 inch page, 300 dpi, an 8 bit color depth.
They are considering the Pipeline and Adobe postscripts.
Have your ever given any thought to licencing your technology for an embedded RIP system?
Have you any plans to support postscript Type 2? How do you handle postscript level incompatibility problems with your product?
Does the printer you are working with need to be able to stop and restart the paper feed (or can you keep up with it)? When the image gets very complex can you still keep up with the printer?
How do bit-mapped images affect your operational speed?
On 1 October 1991 Flemming Stanley faxed the following to Trevor Marshall at Yarc Systems Corp.:
Before you call this may be relevant:
Is Pipeline your product?
Yes we have considered licencing our technology but we are very uncertain about the strategy and I feel uncertain at this point of time to commit to anything but will of course be willing to discuss any proposal.
Our new product will be supporting PostScript level 2 around mid 1992. There will be no incompatibility problems but the user will have to decide whether he wants to run Level 1 or Level 2.
We use the disk as memory. We don’t expect to run out of space this way but if we do we will exit with an error.
An average page takes up approx. 8MB. A complex page approx. 20-30MB.
There are two types of imagesetters available. With and without start/stop-function. With complex pages we can’t keep up with the printer so we will have to use a diskbuffer when driving a setter without start/stop.
Bit-mapped images are what we call very heavy. However we expect to be faster than any other product.
Our development is fully in tune with our schedules and we look forward to cooperate with you as planned.
By fax letter of 25 November 1991 Trevor Marshall informed Flemming Stanley that the time schedule was being kept and that the first boards would be installed at customers’ before Christmas but that he was worried about the development of the next version. As he intended to participate in the IMPRINTA exhibition in February, he asked to receive a board before Christmas. He said that interest had been shown in the product which solved a problem at a competitive price.
Trevor Marshall answered by fax letter of 1 December 1991:
Congratulations on getting your first “Mac the RIP” out the door. I know what a problem that can be sometimes, and am sure you are elated at your success.
It is not YARC’s policy to provide our boards on unpaid evaluation. We have a 30 day (no questions) money back return policy, and the Credit Card companies make us live up to it faithfully!
However, in this case, YARC sees a strong symbiosis between our corporations. YARC has been promoting the “Pipeline Associates” Postscript interpreter, and it is proving to be quite slow, and a memory hog, when used in high resolution applications. Our association with Pipeline dates back almost two years, when they were introduced to us by AMD as having a good postscript clone. It was indeed extremely efficient with 300dpi A4 pages. We worked with Pipeline proving joint hardware and software Postscript solutions for a number of companies. We have worked with them on embedded solutions for Mitsubishi color laser engines, as well as the plug-in system solutions. They supply the software technology, we provide the hardware and some porting assistance.
For the last 6 months or so, however, we have been unable to close any sizeable Postscript deals with the Pipeline technology (YARC’s charter is to concentrate on tools, and hardware, and to leave vertical applications to our customers and business partners).
A typical scenario is that we get a lead from AMD to a small (? they define small as less than 100,000 units per year) printer manufacturer that is too small for AMD to consider as a direct customer. We then offer them hardware and porting support services.
I would like to get together some sort of arrangement with StanText that would allow us to promote your RIP product as the Postscript technology of choice. I am sure that a deal like this would enable both our companies to reach markets that we would not otherwise be able to individually address.
There are a few loose ends that need to be tidied up technically, as well. For example, what is the optimum data memory size for “Mac-the-RIP”? How much computer power do you need? How does this balance with cost requirements? Will bus mastering over the NuBus to your current processor be fast enough, or will we need to develop a daughter board connection?
I am currently reviewing my schedule to see if there is any way I could get over to visit you again in the next few weeks. I guess you are not going to be out our way before Christmas?
On 12 December 1991 Trevor Marshall visited Denmark and passed on to StanText A/S various technical specifications, produced in the case as Exhibits 57-60, concerning Yarc’s products.
In connection with the meeting and the material forwarded later StanText initiated new further development of Mac the Rip version 1.1 which resulted in the so-called Dan Rip version 1.2. This development took place in the period until June 1992. In the Dan Rip version 1.2 is included software on disks containing three files - Dan Rip Super, Acc Boot.Aut and Acc Main Out, to which Yarc estimates to hold approx. 14%, 100% and 51% of the rights.
On 16 December 1991 Jim Potter from Yarc faxed the following to Peter Tolboell at StanText A/S:
I’ve seen the Lino output of your PostScript Interpreter, at 1270 DPI. It looks great. Can you do color separations, etc. … to the Lino?
What do the times at the bottom of the image (I-Time: 43.8 sec. Time: 2?.2 sec) represent? How much is the actual PostScript, how much time is taken by the “showpage” to the Lino?
Please send me some more information on your system. I’m interested in the encoding/decoding that Trevor tells me you perform with your FIFO MUX Line-Buffer board. Is your encoding one-dimensional (some sort of Run Length Encoding on a single line) or 2-D, describing a multi-line object?
I look forward to working with you on acceleration your interpreter. I am also sending you a press release from VIDI, showing the speed-up which they realized by using a co-processor. …………”
On 18 December 1991 Jim Potter sent another fax letter:
I’m sending you a new disk of the High C compiler, pass 2 (HC2COM).
On this disk is a copy of our most recent YARC29K Executor. This allication allows you to create “batch” files in your working folder, leaving the Executor itself down in the YARC29K folder. One executes the compiler by double-clicking on the batch file. Alternately, by double-clicking on the compiled executable, the Executor is opened and the executable file is then rum(with optional command line arguments).
In testing this new version of the Executor, I haven’t been able to reproduce your trouble of not finding the YARCBOOT file. It resides as ::YARC29K:BIN:YARCBOOT, and this seems to work at this end. There seems to be no problem if YARCBOOT is in a BIN folder where the BIN folder is in the same folder as the YARC29K Executor was when launched.
One caveat, though – if you open the Executor by double-clicking on it, and then navigate to your work folder by using the “SET DIRECTORY…” menu item, the C compiler will not be able to find your source file when you give the command “HCC SOURCE.C”. You will get an “Interrupted system call” error message from the compiler. The preferred method is to use the supplied batch file; placing the batch file in your working folder and the double-clicking on the batch file will sommon the Executor, and then by running the batch job will get you compiled.
On 15 January 1992 Flemming Stanley in a fax letter asked Trevor Marshall for the design for a FIFO interface which should have been sent together with interface design for Compugraphic. At the same time it was realised that StanText would not be ready for the Imprinta exhibition but would do as much as possible. StanText would avoid NuBus interface and would develop direct interface between two boards even though it would take quite some time.
On 6 February 1992 Trevor Marshall gave this answer:
YARC’s coprocessors are not general application accelerators. It is just not possible to take code compiled for 68000 series processors and make it perform much faster than on a IIfx or QUADRA.
YARC’s RISC coprocessors are complete RISC computers on a board. They are application specific (although , as more packages are ported, we have a number of customers who are running several different applications on the one coprocessor).
We do have support for Adobe Photoshop, in the form of fast 29K JPEG compression technology. This is a “plug-in” which any photoshop user can utilise.
YARC’s coprocessors are extremely fast (according to our OEMs, about 4 times the speed of a Quadra). If you need speed, we can provide it with good price/performance ratio. If your RIP can take advantage of background processor (or multiple processors) and thus allow you to sell a more compact an appealing package, then again the RISC coprocessor is the solution. If you just need a faster MAC, however, you need to ask Apple for that. And as of today, they have no solid plans at all (See my cover article on “Power PC”) in the February 1992 BYTE magazine).
In a fax letter of 1 June 1992 Flemming Stanley informed Trevor Marshall that the Yarc implementation had entered a phase of frustration and that Peter Tolboell even though he got new tools from Yarc continued to encounter problems for which reason StanText did not - as announced - expect to market the accelerator in May. He therefore suggested that Yarc sent a man to Denmark. Flemming Stanley in addition informed that the accelerator was 100% faster than any of its competitors.
On 4 June 1992 Jim Potter reported that he would come to Denmark and would bring along new revised versions of various things as well as a new C compiler from Metaware.
During his visit to Denmark from 5-11 June Jim Potter in a fax letter of 9 June 1992 requested Yarc to send files. The letter begins like this:
Things are going well here at StanText … I think that we may pushing some bitmaps through the FIFO’s sometime this week. Peter and I have been going over all of quirks in a 29K design (const. vs. consth, delay slot NOP’s how the compiler does inlining, etc …) Peter’s daughter board design is very similar to others we have seen - could you please AppleLink to me several files from my PC?
In August 1992 StanText bought 11 accelerator circuit boards from Yarc.
A notice was published in The Seybold Report on Publishing Systems (Seybold Report), which is one of the world’s most respected magazines within the graphic art industry, on 19 May 1993 stating that StanText had been ordered to be wound up and that its rip was being marketed by Danrip Int. which was a subsidiary of Yarc and that Yarc produced RISC boards like the ones employed by StanText in the Danrip Maxi Super.
Upon Flemming Stanley’s purchase of assets from the estate a dispute arose partly as to the right to the name Danrip and partly as to the right to the technology. In a fax letter of 6 June 1993 from Trevor Marshall to Flemming Stanley it says :
You will recall that in any case we retain full rights and access to the Source Code and object code that YARC provided for StanText (following your written request of June 1st 1992 and the subsequent visit by Jim Potter). You must cease and desist immediately from using or claiming rights to this license. It formed a substantial part of the StanText RIP code. You must know that it was YARC technology made the StanText RIP product work correctly, and we also gave it all its speed.
We have now heard you are using an accelerator from Radius, a competitor of YARC. Your use of our code and proprietory technology with this competitor’s product is clearly illegal.
On 10 June 1993 the liquidator, attorney Soren Pedersen, among other things faxed the following to Yarc’s attorney concerning the surrender of the source code:
in the light of the above I therefore have to insist on what I stated in my letter of 15 January 1993 according to which your client has no claim for the surrender of the so-called source code. The reason for this is that the original exclusive distributorship agreement was terminated due to breach and only replaced by a - according to oral agreement - non-exclusive distributorship covering Scandinavia and parts of Germany.
Probably on 12 September 1993 Yarc/Danrip issued a press release concerning Yarc’s release of a high-speed rip from which it appears:
Not to be published before SEYBOLD SE, Oct. 21, 1993.
YARC SYSTEMS RELEASES HIGH SPEED RIP
YARC Systems, the leading supplier of 3-D RISC acceleration hardware, today releases its new high speed RIP with integrated OPI.
The RIP runs on a RISC accelerator card for Macintosh and IBM compatible PCs.
YARC is not a new-comer to the pre-press marketplace, having produced OEM products for Linotype-Hell, Agfa Compugraphic and Dai-Nippon Screen.
In May 1993 YARC acquired DanRIP Intl. A/S, in Denmark, which was formed to market the RIP technology developed by the now bankrupt StanTaxt A/S- DamRIP’s installed base covers most of Europe, with customers ranging from newspapers to service bureaus.
YARc’s new RIP uses two (2) dedicated RISC CPUs, including AMD’s Pentium beating 29050. Its three proprietary ASICs are capable of bursting pixels to the marking engine at data rates which are much faster than any commercially available film recorder.
“We have balanced the computin tasks between the two CPUs and the ASICs so that the overall RIP speed is maximized” explained Dr. Trevor Marshall, YARC’s CEO, “YARC has produced a RIP that is very fast on every image not just on benchmarks”.
In 1993 Yarc and Danrip Int. instituted legal proceedings against Stantech Digital Image Processing Systems A/S and Flemming Stanley. In Flemming Stanley’s defence of 14 October 1993 to the Superior Court of The State of California it is among other things stated that:
Plaintiff has now “again developed” another DanRIP with two processors. Although Plaintiff by this “new innovation” tries to hide away the copied DanRIP board into three “proprietary” ASICs - Defendants have from the above already stated good reasons to believe in the light of the above evidency that it is still Defendants board/assets which Plaintiff has illegally copied.
Not surprisingly Plaintiffs are trying their best to stoop further information being published BUT all the stated issues so far presented form a substantial part of the current juridical procedure taking place at the Maritime and Commercial Court of Copenhagen, Denmark.
On 17 November 1993 the Seybold Report carried the following article:
Yarc readies new RIP
Yarc Systems, which has been supplying accelerator cards for use in RIPs from other vendors, has developed its own full- function RIP. Called the Yarc-RIP, it is the company’s first graphic arts end-user product, having served mostly OEM markets. It runs on two boards in a Mac. There are no plans for a PC version, but Yarc said it could be done.
The RIP wasn’t quite finished at show time, but Yarc demonstrated the functionality that was working. It was driving a Linotronic imagesetter in the booth, outputting complex line-art files. Yarc said it hoped to have the software completed in about two weeks so it could ship it in November.
The two-board based RIP, which supports PostScript level 1, is based on an AMD 29050 RISC processor with up to 32 MB of memory and a 29000 coprocessor on the second board, which handles the recorder interface and some of the rasterization and screening functions.
The Level 1 interpreter is Yarc’s own development. The RIP was designed t accept plug-in screening modules developed by third parties.
The board contains an 8-MB buffer for spooling PostScript input, and, if needed, it uses the Macintosh disk as an additional buffer. After the entire job is input, the RIP begins processing the job. It can set up either to output bands of data to the recorder or to buffer data and send complete pages at a time.
From a press release from December 1993 issued by Simpro Computer GmbH it appears that the Yarcrip was marketed in Germany.
The Plaintiffs have in the case in California obtained an opinion from an American professor, Dr. Walter Karplus, who having been presented with specified data, including Danrip version 1.2 (Exhibit 10), StanText 1992 and Yarc Danrip rev 2.0 ( C ) 1993 (Exhibit 11) in his letter of 18 November 1993 states:
1. The two circuit boards (items #10 and #11) are similar in shape, employ similar input and output connectors, and are designed to serve the same general purposes. However they differ markedly in details of design and excution. Some of the major differences include the following:
a. The two boards have significantly different physical layouts and significantly different components.
b. Item #11 is of conventional design and contains few novel features, while item #10 appears to contain a number of innovative features to enhance its performance. For example, item #10 uses a PLL clock, frequency than the PLL clock of item #11.
c. The two board have significantly different information flows.
2. In no significant respect can item #11 be considered to be a copy of item #10.
3. The “Statement of Defense” in item #2 that “Plaintiff by this ‘new innovation’ tried to hide away the copied DanRIP board into three ‘proprietary’ ACICs is nonsensical from the technical point of view.
Flemming Stanley did not appear in court for the proceedings in California and on 22 December 1993 The Superior Court of The State of California passed judgment that Flemming Stanley and Stantech Digital Image Processing System pay a total of USD 843,000.
The actions for libel:
Upon a request from Flemming Stanley the liquidator stated on 16 August 1993 that “Trevor Marshall has not outright confirmed that he copied the Dan Rip, but at the same time I point out that to me there is no doubt that copying has taken place”.
In the Seybold Report no. 3/1993 from October a letter to the editor from Flemming Stanley was published. Above the letter the editors under the headline “StanText founder on the Danrip controversy” explained the background to an earlier article written on the basis of information obtained from Yarc. In his letter Flemming Stanley gives an account of his view and makes the statements indicated by Plaintiffs 1, 2 and 3 under claim 2 a and b.
In the Defence, dated 15 October 1993, filed by Flemming Stanley in the case in California the statement indicated by Plaintiffs 1, 2 and 3 under claim 2 f is included.
On 16 November 1993 Flemming Stanley sent a letter to StanTech’s existing and former customers. The letter was made in both a Danish and an English version and in it the dispute with Yarc is accounted for. The statement indicated by Plaintiffs 1, 2 and 3 under claim 2 f is included in the letter.
The same day Flemming Stanley sent a letter to Glunz & Jensen A/S, att.: K. Michelsen, Attorney, a copy of the letter to the customers was enclosed. The statements indicated by Plaintiffs 1, 2 and 3 under claim 2 c and d are found in this letter to Glunz & Jensen A/S.
Following the judgment in California Flemming Stanley on 27 December 1993 issued a press release under the heading “Danish Entrepreneur Ordered to Pay Damages and Fined DKK 5.5m for “Libel” - by the Court in California”. In the press release an account is given of the progress of the case and in this account the statements indicated by Plaintiffs 1, 2 and 3 under claim 2 g are included.
On 9 March 1994 the Court in Arhus gave judgment in a case brought by the bankrupt estate StanText A/S against Hafnia Forsikring A/S. Of the transcript of the court records it appears among other things:
Flemming Stanley explained …………
He is of the opinion that Carsten Jeppesen or the firm for which he now works, Rip International, stole it …… He has filed a 10 pages long report with the police in which he puts forward arguments to show that it was Carsten Jeppesen who stole the items. The police have been unwilling to do more about it.
Comments of the Court:
On the evidence and the consultants’ opinions it is held to have been proved that the computer equipment was removed by persons who held a certain knowledge of the equipment and who gained access to the premises either by using one of the keys from the original bunch of keys or by using a clone.
The judgment was appealed against and from the judgment of the High Court, Western Division, of 21 June 1994 it appears among other things:
Police officer Max Burhoej, Arhus police, explained that he was in charge of the subsequent investigation into the matter. He questioned, partly because of the report filed against Carsten Jeppesen, former employees of the firm in question. The investigation revealed nothing on which basis the police could bring charges against any persons. …………
The High Court orders:
On the evidence produced before the High Court it is held that the requirements in § 11 of the security provision of the policy were met. According to police officer Max Burhoej’s statement the police investigation in relation to the report filed against Carsten Jeppesen caused no charges to be raised against any persons and it remains unsolved how the perpetrator or perpetrators gained access to the premises. Accordingly, the Court cannot dismiss that the theft took place by using a key made for this purpose.
After the passing of the judgment Flemming Stanley sent the following letter to a customer, Knud Priess, in Horsens, Denmark on 10 September 1994:
In continuation of our telephone conversation the other day I am sending you the enclosed material which might make you realise that you are taking part - probably unwillingly - in the covering up - excuse me my frank expression - of a criminal act of quite considerable extent.
For your information, the firm in which Karsten Jeppesen is engaged has copied our hardware. In cooperation with the police in Arhus we found in the autumn of 1993 such a copy in Copenhagen………….
The firm of course had to use false “evidence” in the court in California but as I obviously did not attend the “séance”, the court as expected passed a judgment in default against me. But let us now wait and see what will be the outcome in Copenhagen.
For your information our software was stolen from the estate by a person who held a key to the premises. Karsten as one of two had not returned his key!
Do you not think that an unusually clear picture is gradually forming of what Karsten is involved in? To complete the picture, I inform you that he has now attempted to defraud the bankrupt estate, but for that he will, of course, be called to account as deserved.
On 27 June 1995 Flemming Stanley wrote the following to a customer in Sweden:
For your information a meeting with the expert witnesses took place yesterday. I am sure that Mr. Trevor Marshall of YARC Systems Corp. has informed you that he made an extensive report in order to push the opinion of the expert witnesses. I can inform you that the expert witnesses have once again declared that the DanRIP must have been copied by YARC and that the report of Mr. Marshall has had no influence of the opinion of the experts!
According to Danish court rules, mr. Marshall is of course entitled to forward more questions to the expert witnesses. I can assure you that he will do his utmost to procrastinate the court case even further. He has no other options! However, I can also assure you that anyone having just a slight knowledge of the legal rules prevailing know the end result of this issue.
Before Christmas we expect to have finalised the matter. If we are successful we will for sure have issued a writ against you and your company as well.
On 21 October 1995 Flemming Stanley wrote among other things the following to European Receivers of Yarc Systems Corp./Danrip Int. A/S, att. Areka Data, Norway, Bergwalls Grafiska AB, Sweden, Deltagraph A/S, Denmark and Medialine Computer + Systeme GmbH, Germany:
“…………History has shown that criminals always make mistakes. Again I shall direct your attention to the fact, that your supplier himself has declared that he is able to further develop the DanRIP system! …………”
In the case the initial expert survey was made by the experts Christian W. Christoffersen, Graduate Engineer and professor Peter Johansen. In October 1995 Christian W. Christoffersen withdrew as he felt threatened by representatives of the Plaintiffs and Christian Rosenlund took his place. Later also Eskild Fjord Pedersen, project manager, was appointed.
The two experts mentioned first in the above submitted the first Opinion on 26 May 1995 in which it is stated:
Expert Opinion, phase 1 eq. list of questions item 5.
This first phase consists of text and assessments related only to item 5 in the list of questions to the experts.
5.a Description of the tests developed by the Defendants
The defendants have developed 8 different postscript files which exercise certain specific functions/features in the implementation (the postscript interpreter component).
The following is a description of the function of the postscript files.
5.b. Similarities between the software of the three products
Similarities between DanRIP v. 1.2.6 and YarcRIP v. 1.0.9.
Generally the two systems are designed according to the same structure and the same level of menu systems. An example is the identical way in which you get to “Spot Size” through the set-up menu. The applications are likewise designed as two windows popping up on the screen when started. One of the windows is used for logging events. The other is used to give status information.
Differences of general character are the order and the content of the menus.
Of specific points worth mentioning are:
Here are striking similarities between the two dialog boxes even though there are different checking systems. An example is that when measuring the position as to pixels and the size of the illustration window and of text and numerical values for “Frequency/Angle” of the colours, you find that these are identical right down to pixel level which is unusual for two different products.
The numerical values of “Frequency/Angle” are identical, including identical decimal digits which indicates that the same algorithm and terms of calculation are used.
Spot size compensation
Here the dialog boxes are virtually identical, with the exception that Yarc has added its logo. It is especially worth noting “Compensation Active” which is somewhat Danish. An American would use “Enable Compensation” or simply “Compensation” as it is implied that it is active/enabled when the check box is checked and inactive/disabled when the check box is not checked.
Here both the menu systems and the dialog boxes are identical. Screen dumps of the dialog boxes are shown in Appendix B. They clearly show that exactly the same checking items and text have been chosen. In both cases the text is made up of “Calibration curve for” attached with the text from the set up dialog box with the device name, e.g. “Simple Format File” or MacWindow” dependent on which device has been chosen. A practical difference is seen in the organisation of checks which is due to the fact that Yarc has had to make space for its logo. But apart from this the similarity is striking.
5.c. Differences between the software of the three products.
In general, many details are different in the two products but it is irrelevant to emphasise such differences in this connection. It is, however, worth mentioning here that as to user interface the set-up windows are quite different in the two products. There are different sets of drivers for output. An example is that DanRIP has a driver for MacWindow where YarcRIP has not. In contrast, YarcRIP has drivers for Simple File Format and TIFF devices where DanRIP has not.
5.d-e. Are the characteristics mentioned recurring in the three products? Comparison of characteristics from the tests.
Below comparisons and comments upon the individual comparisons made by running of the postscript test files mentioned are listed.
In Appendix A logging from all tests are found. In addition to these tests, for the sake of good order tests were run testing all relevant possibilities of combination between the following software and hardware:
Mac-the RIP v 1.1 (1991) DanRIP version 1.2.5
Danrip v. 1.2 (1992) DanRIP version 1.2.6
YarcRIP version 1.0.9
Danrip v. 1.2. with acc. card Danrip super version 1.2.6
YarcRIP version 1.0.9
YarcDanRIP rev 1a (1993) DanRIP version 1.2.6
YarcRIP version 1.0.9
YarcRIP rev 2.1 (1993) DanRIP version 1.2.6
YarcRIP version 1.0.9
YarcRIP rev 2.1 with acc. card DanRIP super version 1.2.6
YarcRIP version 1.0.9
PowerRIP rev. 1.0/1.1 (1994) DanRIP version 1.2.6
YarcRIP PPC version 1.0.9
The hardware was compatible to such an extent that there were no problems in running the above combinations where Yarc products were used with Stantext software and the reverse.
It became apparent that all products show the same error as regards the clippath operation with circles. Clippath is in this respect perceived as a rectangle and not as an ellipse. This error has been detected in no other RIP products. It is therefore quite telling that the error is found both in the Stantext and the Yarc products.
Unfortunately, the experts did not have at their disposal an image setter to test this. But it is a fact that standard postscript interpreters handle this correctly and shows (ABC) as the result and not (ABE) which is the error in Stantext’s product. It would therefore be rather telling, should a similar error be found in Yarc’s products.
The results of the comparisons between the Yarc and Stantext products with the accelerator card are exactly the same. It is therefore possible to conclude that it is very probable that the same algorithm is employed.
Here checks are made for operations which are not normally found in postscript interpreters. Of 27 special functions 14 recur in the Yarc product. This is quite unusual and indicates that parts of the code are from the same source.
Printdict is seen to print the list in a somewhat special order for all the implementations tested. That is to say, the Yarc products work in the same way as the Stantext products. Standard postscript interpreters make this listing as a normal alphabetical list and not as is the case here by listing the shortest names first.
The Yarc and Stantext products do exactly the same thing showing the 10 units as the result and not a “rangecheck” error as expected.
Here no connection is found between the results. The results indicate that two very different algorithms are used.
Here no correspondence is found between the results. The Yarc products YarcsRIP rev. 2.1 and PowerRIP always show 10000 as the result which does not match that of the Stantext products. There are, however, indications that an error has slipped into the Yarc products as the result is always 1000 which is much lower than what has actually been used of virtual memory.
5.f The probability of identical characteristics in different products
The probability of identical characteristics like the above-mentioned postscript functions with very special names is little. By chance, a few examples might of course occur, but never to the extent seen here. For instance, calling a function “grestoralll” with 3 “l”s is quite exceptional. Likewise, capital letters have been used in some of the Danrip specific functional names. These recur in the Yarc system.
Likewise, it is very unusual to see the same error, like the clippath operation.
5.g. Are the Sero and Yarc Products complete or partial copies of Stantext Danrip version 1.2.6.
As regards Yarc/Danrip software version 1.0.9, it is the expert’s definite opinion that this product partly originates from the same source. The similarity found between YarcRip version 1.0.9 and Danrip version 1.2.6 is of such a special and unique character that it is not likely to appear in two different products, except if copying has taken place. Even the same programmer would not recreate such similarity.
The experts have measured the physical spacing and dimensions of the various checks in the single dialog boxes. Even by the use of gridsnapping two different programmers would not be able to position the items exactly as identical as here.
As regards the postscript interpreter it is a fact that many of the peculiarities found in Danrip version 1.2.6 recur in YarcRip version 1.0.9. To give an example 14 specific Danrip functions recur in the YarcRip software. By testing a third product none of them recurs.
It is the expert’s definite opinion that partial copying of the Stantext software has taken place. Seen as a whole the similarities are too many and too detailed to make it likely that we are talking about two independent developments.
It is striking that Yarc has added its logo to every single dialog box. It is an easy operation to make, but it is quite unusual for applications. Normally, the logo is found in a single dialog window and possibly in the main window.
In our Opinion account has been taken of the fact that the same development devices might have been used, such as devices that automate the development of dialog boxes. What we have assessed does not relate to the development devices but to the development process specifically.
We have also assessed the possibility that Yarc wished to develop a system as compatible to the Danrip system as possible and therefore attempted to clone the Danrip system. This might be out of consideration for customers which were to replace the Danrip system by the Yarc system. This could have been achieved through own independent development. It is possible to obtain a functionally identical implementation. It is, however, unusual that the same errors are developed and that specific similarities like those in the dialog boxes and the postscript operators are found. It is therefore the experts’ opinion that Yarc must have used parts of Danrip’s source code.
The Experts’ Opinion caused the Plaintiffs to contact Ole Lykke, the Institute of Applied Data Technology, and asked him for an assessment of the Experts’ Opinion. From Ole Lykke’s Report of 23 August 1995 it appears:
This assessment will focus on the quality of the Experts’ Opinion only. Thus, no decision will in this Report be made as to whether copying has taken place or not. This Report is based on the standard of quality which the Institute finds necessary in order to ensure that the parties receive a fair treatment.
Background The Institute has prepared this assessment as a consultancy service to Yarc Systems Corporation and has tried to be as objective as possible. It is the Institute alone that has prepared the wording of this document.
Data This assessment has been prepared solely on the basis of the data supplied to the Institute by Yarc Systems Corporation. Unsuccessfully, the Institute has tried to make the authors of the Experts’ Opinion surrender the original test routines. The data supplied is found in Appendix to this assessment. (See outline page 3).
Points of criticism
Incomprehensible The Opinion is difficult to read as it contains several incomprehensible sentences. In some cases one has to guess the meaning or deduce from the text in other parts. It has in this assessment been necessary to give an account of the interpretation the Institute has made of the Opinion’s text. Without such interpretation an assessment of the statements in the Opinion would not be possible. Interpretations of the Experts’ Opinion are found in the section Examination of the Experts’ Opinion.
Lack of documentation Documentation of tests, testing environment and test results is lacking in the Opinion. Only a fraction of the tests are documented with their loggings. Finally, the documentation on which the conclusion in the Opinion is based is insufficient.
Insufficient description The Opinion includes a very brief description of the tests which have been carried out. The test results are listed later in the Opinion. However, it is evident later where conclusions are drawn on the basis of the tests that other tests have actually been carried out, tests of which there are no descriptions and the results of which are not documented. In connection with the description of the tests the reader lacks an assessment of their suitability. For instances are the test input sufficiently varied to ensure that any similarity is not simply a coincidence? Are the tests at all suited for revealing copying?
Imprecise conclusions The conclusions in the Opinion are often imprecise. It is not clearly stated what is meant and it is difficult to grasp from the context what exactly the statement concerns. Often it is left for the reader to draw his own conclusion and whenever there is a conclusion, it is often so incomprehensible that the reader still has to interpret it.
Lack of impartiality Certain formulations in the Opinion are not impartial. Matters which are not documented on the basis of the tests are hinted at.
Lack of care Several factors indicate that the Opinion has been prepared with lack of care. This results in allegations that are directly contrary to the test results. It also raises doubts as to whether the view taken is correct.
Examination of the Experts’ Opinion
Page 5, line 15: Quote: “It is the expert’s definite opinion that partial copying of the Stantext software has taken place. Seen as a whole the similarities are too many and too detailed to make it likely that we are talking about two independent processes of development.”
This quotation makes up the main conclusion of the Opinion. It is the similarities found that lead to the conclusion. At the bottom of page 5 it is mentioned that account has been taken to the fact that the Yarc product might have been developed in order to be as similar as possible to the Stantext product. At no point in the Opinion have the various similarities been analysed based on the possibility that we are talking about deliberate similarity. This makes the conclusion of the Opinion rather blurred. It is quite important to know which similarities indicate copying and which similarities indicate cloning. In section 5.g two similarities are mentioned to support the copy theory:
· Physical dimensions of the dialog boxes
· Same names for special functions.
These are similarities which might well be the result of cloning. The experts find that the similarity in the physical dimensions would not exist if copying had not taken place. Reasons ought to have been included as to this point. Dialog boxes are normally made in a resource editor in which the various constituents (the checks) are typically of the same size (unless something else is requested). If while developing one’s dialog request is made for checks of the same size and with the same positions, the design of the dialog box will be identical.
The other point concerning the special functions with identical names has been evaluated at all on the basis of the cloning theory. Nothing prevents a newly developed RIP from being provided with some special functions with specific names.
Finally, the conclusion is also based on the fact that the two products contain the same errors. Quotation from page 5, line 30: “It is, however, unusual that the same errors are developed and that specific similarities like those in the dialog boxes and the postscript operators are found”. Which errors and specific similarities we are talking about ought to have been mentioned specifically. If for instance the result of the Clippath test is seen as an error in this connection, it is quite important to know whether this actually is an error which was concluded in the Experts’ Opinion at an earlier stage but without examining whether it might be deliberate similarity.
Not satisfactory In the Institute’s assessment the Experts’ Opinion is not of such satisfactory quality to justify its inclusion as evidence in legal proceedings.
The Institute’s conclusion is primarily based on:
· the imprecise conclusions
· the distinct lack of documentation
· the insufficient description
Lack of confidence The distinct lack of documentation and arguments implies that the reader in order to be able to use the experts’ statements must have enormous confidence in the experts’ judgment. Such confidence is lost when seeing other shortcomings of the Opinion such as lack of care and imprecise formulations and conclusions.
Having found a YARCRip at a customer’s StanTech asked Rambøll, Hannemann & Højlund A/S for a comparative assessment. In the opinion of 27 November 1995 it says:
YARCRIP = DANRIP ?
At the meeting with Stantech’s representatives on 14 November where we “explored” a YARCRIP at a firm in Denmark we found a number of similarities between this rip and Stantech’s similar product DanRIP which caught the eye of even I who have no expert knowledge of the graphic art business.
In connection with the comparison we also checked the type of Yarc hardware:
The component in question was a circuit board onto which a daughter board was placed. On the motherboard was among other things written “ZUMASee”.
In addition, the motherboard had the serial number SN14891 and the daughter board had the serial number SN15304. The two board were versions 1.0 and 1.1, respectively.
That another operator should come up with the same names, and even the same (inconsistent) use of capital and small letters, is very unlikely.
The order of this print is quite unique as names are sorted partly according to the length of the names, partly alphabetically.
During the implementation of some standard PostScript functions various errors have slipped into the DanRIP software. An example is that the strokepath function is not working correctly.
The similar function in YARC’s software is neither correct - the error is however shown in a different way.
YARC’s program has the same feature which is not in itself extraordinary. What is extraordinary is that the respective screen pictures to control the function are almost identical in design and mode of operation. - A programmer usually provides a newly developed function with his own distinctive mark, being inspired of course by an example.
Additional possibilities of comparison ….
Practical problems made it impossible to feed certain test procedures to the Yarc installation in question but my impression is that such procedures would give us additional indication of whether the two products are identical.
At a meeting on 23 October 1995 the experts Christian Rosenlund and Peter Johansen were asked by the parties to conduct additional tests, the result of which was the Opinions of 1 December 1993 to YARC and StanTech, respectively.
From the Experts’ Opinion to Stantech is appears:
The purpose of this task is to compare DanRip version 1.2.6 and YarcRip ver 1.0.8 by feeding 20 minor postscript test jobs to both rips and compare the output. The output to be compared is the output which in the postcript is called stdout or errout, that is, not the output on film or paper.
With this test we have examined:
- 1 special function (===) which is found in both rips with the same function.
- 2 rip dependent areas (9 and 10:dictionaries) which are almost identical.
- 2 input related operators (status and file) showing same errors.
- 1 input variable (%stdin) showing same error.
- 7 functions (==, bind, eq, save, [, pathforall, cvs) showing same errors.
- 5 functions (filter, flattenpath, rand, setscreen and eexec) all showing the same error-free result.
- 1 function (exectack) showing different results.
- 1 function (token) which shows different error messages.
Only one of the 11 errors proved by these 20 minor test jobs is different in the two rips. The remaining 10 errors are identical in the two rips.
As to the 5 functions which show no errors, it is hard to say whether the implementation is a copy of the source code or simply a copy of its functionality as the results should in fact be identical. However, the implementation of these 5 functions is very similar to each other, e.g. in the setscreen implementation a feature is found which is beyond the standard function of setscreen and this feature is found in both rips.
Finally, 3 areas (special function === and the two dictionary tests) reveal great similarity in areas in which different rips are often very distinct.
I base my conclusion on the following three observations particularly:
- The many identical errors found in the programs.
- The 3 rip specific areas are very similar.
- The implementations of error-free functions seem to be identical.
This makes the two products seem identical to such an extent that my conclusion is that either
- has Yarc copied parts of the source code directly from the DanRip, or
- Yarc has cloned very meticulously the DanRip, or
- the two rips employ a shared library of routines which may be the case if we are talking about further development of a previously shared version.
Only a comparison of the source codes and/or more tests would make me capable of presenting a more precise conclusion.
From the Experts’ Opinion to Yarc it appears:
1. Rational Tangent Screening:
The angles of Cyan and Magenta may be checked in that the total of the angles must be 90°. Checking of numbers from the Screen dumps shows that both rips “throw away” decimals instead of rounding numbers up where this should actually be done. The result is that the total of the angles of Cyan and Magenta is not exactly 90°. Whether the same errors in both rips are due to copying of algorithms or simply the use of the same standard system call is only to be determined by a comparison of the codes.
It is difficult not to say impossible to determine on the basis of the existing examinations whether the Yarc dialog is further development of the DanRip dialog (further development of the DanRip source code), further development based on own experience or customer wishes and comments or whether the Yarc dialog is simply a cloning of the useful features in the DanRip dialog (i.e. only the experience from the DanRip is re-employed). The two ScreenSetup dialogs have a lot of common features as well. I (CR) have e.g. previously seen screensetup dialogs in which it is possible to alter the degree of fineness of the raster and the angle by “moving the angle of the mouse”.
1.d.1 I (CR) find that by now Spot Size Compensation is found in many especially new rips (e.g. all Harlequin versions). Spot Size Compensation also existed (perhaps under another name) in other rips when StanText launched their DanRip.
1.e.1 Based on the examinations made so far I (CR) am not in a position to determine whether Yarc has copied the code or experience from the DanRip. But both DanRip and Yarc employ solely 10% intervals in their calibration curves. Also PhotoShop (Attachement E) employs 10% intervals but in addition an extra check point in the areas of highlight (5%) and of shadow (95%). This also applies to RWS&H and the majority of the Spot Size Compensation dialogs of which I (CR) have knowledge. The reason for including the 5%/95% points are that the areas of highlight and shadow are the most demanding as to calibration (see among other things Seybold’s first comment to Stantext in Yarc’s appendix B page 15 “…severe loss of highlights…”). This “absence” of checking of highlight and shadow is characteristic of both YarcRip and DanRip.
3.a.2 By “copying” we mean that Yarc has had at its disposal the source code for DanRip and has used parts of it without modifying them. By “cloning” we mean that Yarc has examined thoroughly the functioning of the DanRip and has implemented the same functions in YarcRip by rewriting the program.
10.a.1 Both rips set the clippath as a quadrangle even though “the red book” (level1 book page 129) specifically states that Clippath “sets the current path to one that describes the current clipping path”. The decimals are different (in a few cases the differences are of entire points) but the values change as soon as the resolution or format of the output media (72dpi or 300 dpi, A4 or letter size) is changed.
10.b.1 I (CR) believe that clippath works similarly (and erroneously) in both rips. Attachment K “PostScript Language Document Structuring Convention Specification” describes %%BoundingBox: and %%PageBoundingBox: the comments. These two comments indicate the quadrangle which by the use of vertical and horizontal sides rewrites all support points on a page. The clippath operator is not mentioned at all on the enclosed pages. But Clippath is explained on page 129 in “PostScript Language Reference Manual” (level1 edition) as the operator which “sets the current path to one that describes the current clipping path”, that is, the exact clippath and not the rewritten (possibly rotating) quadrangle.
10.c.1 DanRip and YarcRip are identical in that they both erroneously return a quadrangle as clippath instead of the correct clippath. I am not able to determine whether Yarc has copied the source code or merely cloned its functionality on the basis of the tests carried out.
12.a.3 The two rips react similarly to this (defective) test. This supports the theory that this function is either based on a shared library or Yarc’s implementation is a copy of DanRip’s implementation. This test does not reveal whether this has been achieved by direct copying of the source code.
Based on the questions put by Yarc and the tests suggested we have found 5 areas (ScreenSetup dialog, SpotCompensation dialog, clippath, dictionaries and the encoding implementation) where the DanRip and the YarcRip are similar to such an extent that it is most likely that they are based on the same source code. We have further found 2 areas (the hardware interface and the status dialogs) which are with certainty different. In additions the differences in 3 areas (VMusage, fontcache and some of the internal operators) are so great that I (CR) do not think that as regards these areas the YarcRip is based directly on DanRip’s source code. However, it is possible that these 3 areas are further development of a previous, shared version.
Only a comparison of the source codes and/or more tests will make it possible for me to come up with a more precise conclusion.
These Experts’ Opinions caused a hearing in court on 5 December 1995 and new questions from the parties to the experts. These questions were answered by the expert Christian Rosenlund.
From the reply to StanTech of 1 February 1996 it appears:
6: Probability of match of errors
Theoretically, the two rips can have been developed independently of each other and without any of the developing engineerings having any knowledge whatsoever of the other rip. There are, however, so many similar characteristics in the two systems that I believe partial copying or very meticulous cloning of the source code to be likely.
See also the conclusion in the Report to StanTech of 1 December 1995. I have, however, not taken into account the possibility that Yarc and StanTech have used a shared library of routines as this was in my view rebutted by both parties during the hearing in the Maritime and Commercial Court on 5 December 1995.
From the reply to Yarc of 6 February 1996 it appears:
2.h Karsten Jeppesen has produced a number of valuable, effective test jobs and shown insight into both YarcRip’s and DanRip’s characteristics. My view is that Karsten Jeppesen would have been of great help in connection with testing, error recovery and possibly customer adjustment of StanTech’s next versions of DanRip.
Discussion 4 Screen Setup dialogs and Screening Technology:
4.1-b The recommended angles for cyan and magenta on a linotronic 300 with a degree of fineness of the raster of 62 lines/cm and a resolution of 2540dpi do not make up 90° when totalled. The reason is that different degrees of fineness of the raster (62-02 and 65.65 lines/cm, respectively) have been chosen for the two colours. If the same degree of fineness of the raster had been chosen for the two colours (either 62.02 or 65.65 lines/cm) and the angles closest to what is theoretically optimum 15° and 75° chosen, the angles 29.74 and 60.26 degrees or 23.2 and 66.8 degrees, respectively, would have been achieved. These angles, however, are so far from the theoretically optimum angles that considerable moiré patterns will be shown in four-coloured prints. It is possible to destroy partly this pattern be changing the degree of fineness of the raster of one of the colours (and only one).
This is the reason for the recommended angles (and degrees of fineness) as regards a Linotronic 300.
My personal view is that the rounding errors mentioned ought to be conspicuous to anyone who knows of the theory behind RT rasters and their simple trigonometric design (e.g. that the total of two acute angles of a right-angled triangle always is 90°).
I have carried through another somewhat modified version of the Clippath test on both the Danrip and the Yarcrip and found the following errors in the two rips:
The reason for the huge difference in values on the clippath test of 1 December was therefore that the set-ups in the two rips were different. A set-up with which the Rip internally handles equal values results in completely identical output from Danrip and Yarcrip.
If, for instance, the two rips are set to print at the same (higher) resolution, 300 dpi on a Linotronic 300, and in the same format (8.3”x11.7”) the last result of the respective log files is received: Same numbers for the two rips (the total test result is shown in Appendix 3). My conclusion therefore is that it is most likely that the implementations of clippath in the two rips are based on the same algorithms.
10.c-d In both rips the log dialog is used to display the status messages that are returned from the PostScript interpreter. The status messages are in PostScript terminology called “stdout” (standard output) and “stderr” (standard error log) as well. The YarcRip includes in addition 4 “status indicators” of the imagesetter itself, and indicates the title of the current job, indicates who printed the job and indicates to where it is transmitted (to the imagesetter, to a disk, or whether it is merely (thrown away”). Who printed the job and the actual title are comments in the PostScript file itself, comments which observe the so-called DSC (Document Structuring Convention). In DanRip status, title and who printed the job are displayed in the “About DanRip” dialog. From the set-up of Yarc’s log dialog it is difficult to see what its main function is. Personally, I would use the upper half, the status part, of the dialog as long as my production is error-free. Not until some sort of PostScript error is shown would I look at the lower half, the part which is identical in the two rips, the half which displays “stdout” and “stderr”.
10.c-g The 2 log dialogs report differently fontload during start-up and the amount of time and memory used after each job.
10.h-p Differences are also found in Danrips’s and Yarcrip’s log dialogs seeing that it is possible to employ standard (Mac) “cut and paste” in YarcRip. In Danrip it is only possible if there is not already text in the pasteboard. During the preparation of the December Report I copied a small segment of a pict image to the pasteboard just before I copied text from DanRip’s log dialog. As regards YarcRip it was always possible to copy from the log dialog through the pasteboard to WordPerfect.
Since our report of 1 December 1995 I have discovered that the original definition of status was actually to manipulate items of the type “file” only, but it was soon developed to make status manipulate items of the type “sting” as well (as far as I have been able to trace it, the Emerald rip was the first rip with this feature. Unfortunately, I do not know which version of PostScript it employed). The oldest rip I have found is LaserWriter II NT employing PostScript version 47, it can manipulate items of the type “sting”.
I have not been able to find a “Qume CristalPrint LaserWriter” in order to feed the test to such a printer. All the printers on which I have run the test (see Appendix 5-2), including LaserWriter II NT which employs the same version of the interpreter (version 47) as the Qume printer, show the same result, but a result different from DanRip and YarcRip.
17.a-c As mentioned in our “reply to StanTech’s questions” of 1 December 1995 StanTech’s execstack test does not show exactly the same results on the two rips, but compared to other rips (the 3 logs enclosed with StanTech’s initial questions and Appendix 5-2 of said reply) DanRip’s and YarcRip’s execstack output are erroneous. Their output are either -null- or nothing where other rips indicate some sort of procedure or operation. So I must say that DanRip and YarcRip are more similar to each other than similar to any of the other rips I have tested.
20.a-b FlxProc, SFlxDict, strtlck and mtc are all 4 names which are found in Adobe’s routine called “OtherSubres”. What was striking of the internaldict test from StanTech was not the content of the internaldict, but the identical content of DanRip’s and YarcRip’s internaldicts and that this content was very different from the ones StanTech had included with the tests (as well as from the other tests I have run, see Appendix 5-12).
21.b Since our 2 replies of 1 December 1995 were handed in, I have found no reason to make changes in our main conclusion.
As to test 15 setscreen: I find that Harlequin level2 and PantherProof show the same test result as DanRip and YarcRip. I know that as to Harlequin this is not computed until an element is drawn which has to be rastered. Why PantherProof shows the same result, I do no know. As to test 16 stdin I have found no reason to change the conclusion of 1 December 1995. None of the other rips have implemented stdin in a similar way to DanRip and YarcRip.
21.c That things are working according to the specifications tells us nothing as to whether they have been copied or not. But when additional functions, and especially errors, are found which are similar in the two products, such errors or functions make up a good basis for comparison of the two products in relation to whether the source code has been copied.
21.d I do not believe that the conclusions relating to the individual tests carried through for the reports of 1 December have been changed in such a way that we for that reason alone have to make changes in the main conclusion. During the hearing in court on 5 December I did, however, get the impression that both parties denied the existence of “a shared library routines” for which reason I now have to conclude that
- Has Yarc copied parts of the source code directly from DanRip, or
- Yarc has cloned very meticulously parts of the DanRip.
The expert Eskil Fjord Petersen presented another reply to Yarc on 12 February 1996 of which it appears:
I disagree with Professor Karplus. It is most likely that we are dealing with a copy.
3.d.2, 3.c, 3.d
The comparison of components in “Yarc-copy1” relates to a comparison between Mac-the-RIP and YARC DanRIP rev. 1a. In this case we are probably dealing with a copy. The diagram in Mac-the-RIP is a neat top-down diagram, easy to read. A diagram that the original engineer would have made. Its functionality is practically identical to YARC DanRIP rev. 1a which is divided into separate units that are difficult to read. This may indicate copying or reverse engineering as it is called.
The complexity of YARC-Rip 2.1 is considerably greater and the “equations” or code in the 3 EPLD are of a considerably different character. It is, however, most likely that certain functions have been copied from Mac-the-RIP.
In a letter of 22 February 1996 Yarc criticised the experts’ reply and found that many questions were left either unanswered or answered unacceptably and new questions were advanced.
The expert Christian Rosenlund’s reply is found in a supplementary opinion of 1 April 1996 from which it appears:
This is my reply to Yarc’s supplementary questions to my reply of 5 February. I have presented to Eskild the part which concerns his reply of 15 February. Unfortunately, at the moment he has got no time for an elaboration. In this reply I will concentrate on explaining any linguistic misunderstandings occurred in relation to my reply of 5 February and where my original reply was unclear, I will of course go into detail in this reply.
The reply as to 2.d+g may be rephrased like this: No, I have not seen any technical proof that StanTech did or did not re-employ the code or concepts from their STANTEX 8000 system (I am not familiar with STANTEX 8000) and that Karsten did or did not participate in the development of DanRip.
Elaboration of 2.i: If the great similarity between the two products is due to the re-employment of concepts and not the re-employment of source code, I find the many identical errors striking, especially having in mind that Karsten Jeppesen through his in-depth knowledge of DanRip must have known of all these errors as well.
Discussion 4 Screen Setup Dialogs and Screening Technology:
4.b As I answered on 5 January, Linotype themselves recommended a single set of rasters in which the total of the angles for Cyan and Magenta is 90°. Therefore, to use a Lintronic imagesetter it is not a requirement that the total of the angles for Cyan and Magenta is 90°.
5.a As mentioned in my reply of 5 January the differences in the reply of 1 December 1995 were not due to rounding errors but to the fact that I ran the tests on the 2 rips on conditions I had wrongly understood to be the same, conditions which I in connection with the reply of 5 February found were actually rather different. Therefore, I did not find the mathematical analysis mentioned to be relevant, but I am of course willing to make it should Yarc still wish so.
5.b The high resolution is of course 1270dpi as indicated in Appendices 3 and 4 and not 300dpi as is written by mistake in the reply (the Linotronic mentioned cannot in fact illuminate at a resolution of 300dpi).
As to the number 0.114e-12: All numbers (except 0.0) are given with an exactness of 2 or 3 decimals. I, therefore, assumed that 0.114e-12 was simply a “rounding error” arisen out of the cvs operator’s conversion of the value from a number representation into a text string representation. What the internal, exact value is of this number, I cannot know. The other numbers are probably also calculated with a higher degree of exactness internally and only rounded off to 3 decimals by the “cvs” operator. I believe that a closer examination of the phenomenon 0.114e-12 is best done by code inspection of e.g. the dcs implementations in the 2 rips rather than by a mathematical analysis.
8.b By the description “run with” I do not intend that I had expected to run film with said combination merely that I had expected a nice error message, an error message of the type I got in the other combinations of YarcRip software and DanRip hardware or lack of hardware.
8.B.X.1-2 I have never seen or myself been able to make this combination of software and hardware display a nice error message at start-up let alone run a PostScript job.
10.d has been answered as a discussion together with the reply to 10.c. In brief: Yes, YarcRip displays part of the DSC comments and YarcRip displays status messages formatted as “Status Messages” (as described in Attachment L). DanRip displays by and large the same information but not formatted according to the same standard.
10.k-m Yes, it is possible to print the content of DanRip’s log dialog. I cannot assess whether this is a “major part” of the functionality of the log dialog. The print dialog is the standard dialog from the printer driver. It is not possible to print the log dialog in YarcRip in the same way.
10.p and s As mentioned in my reply of 5 February there is a bug in DanRip’s “copy paste” function which makes it impossible to copy more text pieces from the log dialog without deleting the pasteboard manually first. I cannot say whether standard Mac system call has been applied for the implementation of copy-paste in DanRip and YarcRip. I can only note that Yarc’s implementation operates as expected, DanRip’s does not.
10.u-v In order to determine whether a program is a copy of another or not, access to the source codes of the two programs is required. As concerns the log dialogs the layouts are not completely identical, part of the standard Mac functions (the copy paste and print functions) are not working similarly either (DanRip’s “copy-paste” does not work, YarcRip’s does), YarcRip cannot print). finally, that part of the text is formatted in a slightly different way in the 2 dialogs. But as the function of the log dialog in itself is very simple it is difficult to determine whether the two dialogs are further development of a “shared ancestor” or whether they are actually independently developed and merely by coincidence are similar to each other because their functions basically are the same.
Addition to Discussion 12: The status test is also erroneous on other rips, see e.g. Appendix 5 (Harlequin’s level1 rip) to the reply of 5 February.
That the status test is erroneous is, as mentioned in the reply of 5 February, actually in complete keeping with Adobe’s initial specifications.
Addition to Discussion 13: As regards LaserWriter Pro 630 I am not able to determine whether it is the first or second parameter that causes the error message.
Addition to Discussion 14: I interpreted the original test from Peter Tolboell as a demonstration not of similar implementations of FID, but of similar implementations of the way in which “==” printed FID. [The last sentence is a translation of the original text of the expert’s report as the text reproduced in the judgment makes no sense.] As these implementations apparently are similar, I cannot use Discussion 14 as proof of no copying or cloning.
Addition to Discussion 15: The version number of the interpreter in the LaserWriter II NT I used for my reply of 5 February is: 47.0 revision 1 as it appears from Appendix 7 to the above reply. According to the initial question from Yarc this is the version number of the interpreter in the Qume CrystalPrint LaserWriter (feeding the same PostScript test to the Qume CrystalPrint LaserWriter (Adobe Systems Inc.’s PostScript level 1 v. 47) you receive …). I have no reason to doubt Yarc’s test result on the Qume printer.
Addition to 16.b It is really irrelevant how often an error in any function will create problems, what is interesting in this case is that the same errors are found in the 2 products. Again, “le”, “lt”, “ge” and “gt” are some of the functions I would not test using standard programs but, on the contrary, a minor “hand-written” test job made for the occasion.
Flemming Stanley has explained that Danrip version 1.2 consists of a Yarc accelerator operating as a computer and a Danrip. The Danrip is an application program with an interface board. By connecting a Yarc accelerator and a Danrip it is possible to transfer the images from Rip to a computer.
During the term of the exclusive distributorship agreement problems arose when Danrip Int. said that they had not ordered the 13 boards supplied in May 1992. Not until StanText referred to the agreement did Danrip Int. take delivery of the boards. Payment hereof was not received, however, and StanText requested payment approx. 14 days hereafter. On 15 June at approx. 4.30 p.m. Hans Peter Glunz called and said that Danrip Int. was not able to pay and had not ordered the goods. Flemming Stanley referred to the agreement and persisted on his claim for payment whereupon Hans Peter Glunz began to make complaints about the product. Having the next day sent a fax to Hans Peter Glunz in which he gave an account of the course of the conversation he left the matter with attorney Anders Torboel. Hans Peter Glunz sent a letter of complaint but all the “defects” that Hans Peter Glunz indicated were not defects but simply wishes for the capacity of the product. In the light of this he considered Danrip Int. to be in breach of contract wherefore StanText would take over the sale and marketing of the product and an agreement was entered into with Fotoclark. StanText did, however, allow Hans Peter Glunz the right to match future agreements and was granted preferential right to the sale should the agreements be identical. At a meeting on 30 July 1992 attorney Anders Torboel tried to make a new agreement with Hans Peter Glunz and attorney Peter Thykier. In advance an exclusive distributorship agreement of the same wording as the one entered into with Fotoclark had been forwarded. Following the adjournment of the meeting Hans Peter Glunz announced that he would concentrate on the Scandinavian market. It was agreed that StanText was alone responsible for the product. They did not sign identical agreements as Hans Peter Glunz made some notes to the agreement which StanText could not accept. Hans Peter Glunz then started selling the product in Scandinavia. The provision that the source code was to be kept with attorney Anders Torboell and handed over in case of breach had lapsed. The reasoning behind the earlier provision to this effect was that Hans Peter Glunz was the sole distributor. The exclusive distributorship agreement was terminated in the summer of 1992 and replaced by a new agreement, identical copies of which, however, were not signed by both parties.
The reason for the return of the Danrip boards in June 1992 was that they wanted the boards back as they did not receive payment. Hans Peter Glunz wanted to return some of the old boards, but StanText wanted the boards most recently supplied. Hans Peter Glunz said that some of the boards had been distributed and as documentation hereof sent the stock list. Only 9 boards were returned, accordingly 4 were still to be returned. In connection with the negotiations in July an agreement was entered into that Hans Peter Glunz could keep the 4 boards and get an additional one on purchase of 5, provided that the price of the boards already supplied was reduced. He was later informed that instead the price had been raised. He made known that this was contrary to the agreement and therefore insisted on payment of the 4 boards.
Originally StanText had a product called “Skyline” but in January 1991 he and Peter Tolboell agreed on changing the product. The plans were approved by the Board in June/July 1991 and in August 1991 Peter Tolboell had reached the stage where an actual product could be made. In June/July he had seen an advertisement on the Yarc accelerator. He contacted Trevor Marshall and from October/November 1991 Yarc was able to supply a computer on a board which could be installed in a Macintosh to make it faster. Yarc had nothing to do with the development of the postscript rips as Yarc System is a hardware firm. Trevor Marshall provided them with an accelerator board to make them able to test it. They were told to connect to it via an outlet and a buffer. This was a complicated process as Yarc’s tools were not in order. There are 3 versions of the Danrip, namely, mini, maxi and Danrip super maxi super. The first two use 2 disks while the third requires a third disk. Danrip Int. has never bought a Danrip maxi super.
The announcement of 30 April 1993 of the agreement between Yarc and Danrip Int. indicates that they have possession of the source code, if not further development of the product would not be possible. On 7 May 1993 - 3 weeks after StanText’s purchase of the rights - Danrip Int. sent a price list to a firm named “dan-grama a/s” in which StanText’s software was included.
The photo produced as Exhibit 29 shows a board which is not from StanText. It is obvious that original chips have been removed in order to conduct reverse engineering but they have forgotten to put on the original labels.
During a meeting with Jan Møller of Sero - in connection with problems concerning the film - said Jan Møller told that Yarc had copied the board. Karsten Jeppesen had told Sero that they had bought the rights and developed the system and the hardware part. At Sero’s they found a board on which was written Yarc DanRip. StanText thereupon installed their own software in the machine - and the system worked. This must indicate that we are talking about 100% copying of their software.
As to the action for libel:
He has himself written all the statements and that all he has said is true, but he probably should have worded it differently. He acknowledges that he wrote to Seybold Report that Stantech had bought the rights.
The day they had bought the effects from the estate Peter Tolboell got a call at 5 p.m. from Jim Potter saying that Yarc had copied the system. The next day they had to provide security for payment to the estate and at the premises they discovered that there had been a break-in as various components had been taken from the computer. Of the things taken was the main disk on which customer files and the source code were stored while the thief had left untouched other component of the system. It was not a physical break-in at it was evident that the person in question had locked himself in and taken the effects he wanted. Only the Danrip boards were taken whereas the interface boards, etc., were left. At Karsten Jeppesen’s place of work was a tapestation used for making back’ups which was placed in the safe. This was stolen as well. Karsten Jeppesen later explained that he threw it out when it got to “hot”. In the light of this he felt entitled to issue the press release as done.
Karsten Jeppesen became employed with StanText approx. 1 year before Peter Tolboell. Karsten Jeppesen spend most of his time on up-dating StanText 8000 for the graphic art industry.
The draft press release has probably been sent and what he has written is true. Apart from his statement during the case in Jutland, he has not said or written that Karsten Jeppesen did the break-in.
He believes to be able to prove that the accusations of handling stolen goods and forgery in California are true but not that Karsen Jeppesen did the break-in.
Peter Tolboell explained that he has an education in MSc engineering and works as development engineer with StanTech. He has been employed by StanTech since 1986 and cam directly from university. The first thing he worked on was some graphic art systems on StanText 8000. He later worked on Skyline in 1987. Karsten Jeppesen was already employed at that time.
StanText asked for help from Yarc in the form of an accelerator card and a software control program as well as a development system, Compiler, which translated the source code. Jim Potter helped him in getting the development system running as well as in writing an interrupt routine into the software. This did not concern the actual development but only the functioning and the correction of bugs. Claim 1 as to the FIFO technology “is rubbish” as StanTech does not market the product which concerns only Yarc’s accelerator card which they did not buy from the estate. Yarc’s accelerator card has as its entry point
FIFO which he told Trevor Marshall. The user interface part runs separately and is a StanText product. Accelerator “main log out” is a Yarc product. Accordingly, the FIFO technology does not belong to Yarc.
Peter Tolboell himself, Mogens Kjaer and other StanText employees developed the Danrip. Karsten Jeppesen wrote the spooler system in Mac the Q. Therefore, Karsten Jeppesen only participated in making 1 module in the rip, the reception of data from Mac the Q. Karsten Jeppesen was not engaged in the rip technology and has acquired no knowledge hereof. With the knowledge he had acquired from StanText, Karsten Jeppesen would not be able to develop or further develop a rip, but he would probably be able to get acquainted with the technology concerning this matter.
The labels shown in the photo (Exh. 29) of StanText’s Danrip version 1.2 are hand-written while the labels affixed by StanText were printed on a laser printer.
In the afternoon the day they bought the rights from the bankrupt estate he had a talk with Jim Potter who told him that they had mad a copy of the hardware. He believes this to be true seeing that Jim Potter could mention technical details hereof. At least, they must have made a diagram of the board. The boards are designed differently but have the same functions. He designed the board which is a special interface board which is not described in general.
The product was first named Mac the Rip and then Danrip I. He designed a new system on the Macintosh platform. Yarc had nothing except the accelerator component.
He has not been a party to the writing of the statements in the action for libel.
Stig Mikkelsen explained that he is employed as sales engineer with DK Statoil. He was on the staff of Danrip Int. from the beginning of 1992 till 20 December 1995 which involved visits to customers where he carried out the installation of software.
In the summer of 1992 Sero had delivered on trial a Danrip1, Mac the Rip. The product was not accepted and they lost contact with Sero. In the autumn of 1992 another trial installation was made with Sero the result of which was a sale at the end of the year. Bugs kept being discovered in the product which StanTech tried to recover. Yarc could not recover the bugs as they related to the software which was a Danrip product. The hardware was then remade. He could not give an answer to how much of the rip is software and how much is hardware. He contacted StanText to have the problem solved. Karsten Jeppesen was able to answer technical questions concerning debugging.
Danrip Maxi Super consists of the Danrip board and the Yarc accelerator and software. He had had the software sent from Mogens Kjaer and had been given a Danrip Maxi Super in connection with an installation with a customer in Sweden.
The software sent by Mogens Kjaer was the software they were to buy later. He got the software without getting the hardware. They did not pay for it and once or twice he received 3 disks.
The sale never got under way due to the threats of forgery and handling of stolen goods. He could not reply to what impact the press release in April 1992 had but an adverse effect was evident on sales.
He was the one who returned the boards in June 1992. He has no knowledge of the agreement between the parties but it is his understanding that the agreement was changed so that they were now only allowed to sell in Scandinavia. He was told hereof by Hans Peter Glunz:
They have, as far as he knows, bought some licences and were therefore able to do as they pleased. He did not himself take part in the development of systems. He does not know whether Yarcrip was based on Danrip or what components Yarc held at the time of the bankruptcy, but both rips are based on Macintosh. In January 1994 it was ready to be introduced but spoolers and OPI were still lacking and the product was extremely defective. Karsten Jeppesen and Jim Potter were responsible for the development hereof. Jim Potter stopped approx. 1 year ago.
He has employees shares in Danrip Int.
Flemming Arne Jensen explained that he is a senior consultant with UniC which is a publicly owned enterprise engaged in the running of computer systems for the universities, etc. He was appointed by the liquidator of the StanText estate to assess the software part. He has a degree in law supplemented by a degree in computer science.
A rip is a processor which generates an output (print on paper). He did not assess the rip only the software part which was rather complex, i.e. contained a considerable program code. A skilled developer writes approx. 40 lines a day. The program in question contained approx. 100,000 lines. We are talking about a niche production addressing the graphic art business. In order to further develop and maintain the system the source code is needed and anyone claiming not to have had the source code in these instances is not telling the truth. He has no doubts that the product in question was copyrighted. Normally, the text part of a program is extensively documented. He knows of the firm name Adobe which is engaged in postscript, but that is all he knows.
Karsten Jeppesen explained that he is originally trained as a dentist. In addition he has a bachelor degree within the computer field from the USA.
He was employed with StanText in February 1985. Prior to that he had developed systems, for e.g. pharmaceuticals use and computer language. Until Tolboell was hired, they made Danrip vers. 1.0. StanText 8000 which is a rip including word processing, images, hyphenation program, etc. It was not a postscript rip. A rip is able to receive images and transmit them to an image setter, StanText 8000 was able to perform this function as well. He does not remember at what point in the technical development StanText was when contact was made to Yarc. They began by developing Mac the Rip and Yarc assisted in providing a co-processor. Such was needed to make the product faster. In addition, parts of the board had to be re-designed. He was primarily involved in the net work part. He and Peter Tolboell held the know-how and Peter Tolboell the contact with Jim Potter. Mac the Rip was similar to Danrip version 1.1. No alterations were made in Yarc’s co-processor between versions 1.1. and 1.2. Alterations of the board were not made to make the co-processor work in conjunction with the accelerator. Software was supplied from Yarc for this purpose and it was turned into Yarc Danrip version 1. In October 1992 he went to the USA to remedy defects to a Danrip and demonstrate the system to a customer as he believed to hold the technical know-how necessary.
Prior to the bankruptcy in December 1992 he held a meeting with Flemming Stanley and Peter Tolboell on the formation of a new private limited company. His attorney advised him not to accept the draft agreement and nothing came of it. He handed over the source code for Mac the Q in connection with the bankruptcy. All else related to the source code was destroyed at this point or earlier.
He had his own software business from 1985 till 31 December 1994 when it was wound up. He got in contact with Yarc in the spring of 1993 when he got a call from Trevor Marshall and he was hired in September of that same year.
At the beginning of May 1993 Yarc started developing further the Yarcrip. He is responsible for said further development. In January/February Danrip Int. received the first version which was defective. The development lasted from April/May 1993 till January/February 1994. In April/May Yarc was able to offer the Danrip on the market as they had the software from StanTech as well as the Yarc technology, but Yarc did not have possession of the source code. It took 3-5 man-years to develop the Danrip. Yarc has since 1993 sold approx. 30-40 rips. That the two rips show the same errors may be because we are looking at the same universal error which has been recovered according to customer wishes. Had they had the source code, it had been possible merely to re-combine the whole thing in approx. 14 days. The time employed is realistic for the development of the Yarcrip which, however, was not completed. The development of the Yarcrip was based on manuals on how to interpret the language. They started the development of Yarcrip from scratch.
Adobe is an American company describing the syntax of the language, etc. Postscript is easily accessible. Worldwide there are 8-12 producers that develop rips. The common basis of the rips is a description of the postscript language.
Yarc Corporation employs 16 people and has a turnover of USD 1m a year. Only from its cooperation with StanText did Yarc know of the rip technology. Since 1990 Yarc has supplied software to a number of large customers. He is now in charge of the development department of Yarc while Trevor Marshall is the technician with the highest position within Yarc. He is the one who holds the deepest knowledge of the rip and Yarc’s rip technology is part of StanTech’s rip. He does not know whether the files mentioned in the claim are included in other components than the accelerator.
Yarc did not hand over the source code for the survey as they thought it to be unnecessary and difficult as well.
As to the action for libel:
He stated that copying had taken place, but it concerned a functional equivalent which he was not aware of when he said that copying had taken place. He made this statement prior to his employment with Yarc Corporation and he did not know any better.
He did not do the break-in. Having been notified of the break-in, he called the police and said that he had a key to the premises but that he had nothing to do with the break-in.
He has not done any reverse engineering of Danrip version 1.2 (photo exhibit 29), but Yarc has. This probably took place in April 1993 according to information received.
The product supplied to the firm Sero did not work and the start-up file was altered whereupon it worked.
The source code was handed over to Flemming Stanley and Peter Tolboell.
Mogens Kjaer explained that he is self-employed within the software business. He was employed as development manager with StanText from May 1987 to November 1992 and was Karsten Jeppesen’s superior.
In 1991 they ordered from Yarc an accelerator board in order to run the system faster. From Yarc they only received software for the accelerator and not for the rip technology.
From January 1991 till 1992 Karsten Jeppesen’s task within StanText was to develop software for the queue system and the image database, etc. and to support StanTech 8000 which was a make up and text composition system, he was also in charge of the photo-voltaic cell driver. Karsten Jeppesen had no time for engaging in the rip technology.
In order to develop further the Danrip the source code was needed which was not necessary in order to support. In addition, it was needed in order to solve big problems for the customer. The Danrip worked in January 1992 when they ran a test in connection with the delivery. The test proceeded as expected but there were still matters to be corrected. He cannot remember having ordered that a third disk be sent to Stig Mikkelsen. Generally, he cannot remember in detail what happened in this connection. The system was developed prior to his employment.
It is possible to support the Danrip without a source code. Postscript is also a programming language and it contains among other things a show page. If problems arise as to the show page, the source code is, however, needed.
Anne Marie Christensen explained that she was in charge of the daily administration within StanText. In addition, she wrote manuals for among others Flemming Stanley, Karsten Jeppesen and Peter Tolboell. First she wrote the user manual for StanText 8000 and then a user manual concerning Danrip. The last-mentioned manual contained 6 sections of which Karsten Jeppesen had written about the queue system and the image database. She has not written the reference manuals. She believes that labels for the boards were partly typed but cannot remember whether they were printed on a laserprinter.
Attorney Soren Pedersen explained that he was the liquidator of the StanText estate which was completed with distribution of dividends to all preferential pay claims.
He did not hand over the source code to Danrip Int. as he did not find that they were directly entitled hereto. He stated that they would have to file the claim with the Bankruptcy Court which they did not. Due to the relationship between the parties he chose to request bids for the effects of the estate. At the meeting where the bids were to be advanced Trevor Marshall, Flemming Stanley and Peter Tolboell were present. As far as he remembers only Yarc and StanTech made bids for the source code. Prior to this Trevor Marshall had several times offered him money for the source code and he received daily calls and fax letters both before and after the sale. By turns, Trevor Marshall put pressure on him and wanted to cooperate. During a conversation with Trevor Marshall he got the impression that Trevor Marshall had what they needed to develop the program. After the sale of the source code to Flemming Stanley and Peter Tolboell he cannot remember having received further offers from Yarc.
The firm Scanview bought the furniture from the estate. He sold the rights that belonged to the estate but he did not dare to guarantee what rights they were talking about.
The break-in at the premises of the estate happened around the time of the bid making and it was reported to the police. He discovered the break-in the same day as the offers were opened.
Hans Peter Glunz explained that he is the founder and director of Danrip Int.
10 units of the first rip version 1 were supplied on 12 January 1992. The development of Danrip version 1 was initiated in the beginning of 1991 and it was his idea to place in on a smaller platform. Flemming Stanley also thought this to be a good idea. Originally, Danrip version 1.2, which is further development of Danrip version 1.1, should have been tested on 16 December 1992, but the test was postponed until 12 January 1992. It was not possible to make the software part function satisfactorily and his net of distributors was ruined. Version 1.2 was not put on the market until April/May 1992 and he had at that point 16-17 version 1 in stock. In practice, only one version existed and customers were offered the old version if they were contend with that. It was important to clear out stocks.
On a current basis errors were reported by telephone and Flemming Stanley continuously tried to correct them, but new errors appeared and distributors lost confidence.
He was on a trip to the USA when he was informed that another 30 units of Danrip had been supplied. He said it was a mistake and that they had to be returned. This, however, was not correct as he according to the agreement had to buy 32 units of Danrip. They then reconsidered the whole situation and the agreement was modified in such a way that he had to sell 130 units the first year which, however, proved to be impossible. They discussed the matter and agreed that his territory should be reduced to Scandinavia. He was, however, granted an option as concerned Germany; if could match the agreement that StanText would otherwise enter into. StanText entered into an agreement with a German company named Fotoclark, which agreement he did not wish to match. They also discussed whether they should enter into a new agreement. Flemming Stanley wanted the agreement modified. However, the exclusive distributorship agreement was not annulled. In such case he would have received approx. DKK 2m to pay for the surrender of the 16-17 Danrip models which he had in stock, money which StanText did not have. In connection with the draft agreements the question as to the source code was not discussed. The reason for the agreement on the deposit of the source code was partly that in the event of bankruptcy they would be able to provide services to customers, partly that Danrip Int. needed a guarantee that they would be able to continue in case of StanText’s bankruptcy. The cooperation continued hereafter but the prices were changed. They had previously changed the prices but then in cooperation with Flemming Stanley. Danrip Int. was to get 4 units of Danrip version 1.2 to compensate the earlier loss. There were no conditions imposed in this connection, not even that the price of the earlier version was to be reduced. He did not reduce the prices in the official price list. He has no differentiated prices in the price list. In the price list is found a Danrip “maxi” and a Danrip “mini”. He does not think he has sold any Danrip “maxi” at “mini” price. Flemming Stanley told him to sell version 1 at a lower price, but he sold all at a lower price in order to be able to sell any at all. Thereafter he received another invoice from Flemming Stanley.
When StanText went bankrupt, Danrip Int. had 16-17 units of Danrip, including the 4 versions of 1.2 in stock. Following the bankruptcy he demanded to be given the source code. DanRip offered a symbolic amount for the source code but had no use for it until Yarc appeared on the scene. He does not contend that Trevor Marshall made persistent attempts at getting hold of the source code. He was away on business when offers were opened at the liquidator’s and therefore did not participate but he was in telephone contact with Trevor Marshall that day. The envelope containing the bids was to opened at 10 a.m. It did not happen, however, and the bid making was postponed twice till the afternoon, probably till 1 p.m.
In the autumn of 1992 they got in contact with Sero and sold a Danrip version 1. It was, however, defective which was reported to StanText that tried to correct the defects. They received many corrected versions. It only solved the problem temporarily as new problems arose. The interface board was simply replaced by a Yarc board but the same software was applied. He does not know whether new defects emerged.
He got to know Trevor Marshall in February/March 1993 through a StanText board member. Trevor Marshall told that he had worked on a rip. In March/April 1993 they drew up a letter of intent on cooperation. Until Yarc had developed its own rip, they decided to use the 24 rips he held in stock. He had bought these rips and was therefore entitled to use them for whichever purpose he wished, including for alterations and reverse engineering. He started selling the Yarcrip in December 1993 or January 1994. It was not possible to sell the Yarcrip as it - even with an accelerator - was slower than the Danrip without an accelerator.
In the press release of 30 April 1993 it says that what is developed is Danrip’s Product Line not the Danrip. When the firm Simpro in its price list from September 1993 mentions a Danrip, they must be referring to the altered version. However, he does want to contend that a Yarcrip as the one mentioned in Yarc’s press release of October 1993 may have been launched.
This case has had an outrageous effect on himself and the company. Flemming Stanley wrote to his Swedish distributor that he felt offended. Flemming Stanley’s letter to Simpro also resulted in considerable loss which had nothing to do with the defective functionality of the Yarcrip. He estimates the total loss at at least 2.3m as stated in his letter of 2 April 1994 to his then attorney.
Erik Brock Hansen explained that he was employed on 28 April 1993 as sales and marketing director with Danrip Int. He is not trained within the computer business and is not longer employed with Danrip.
He made the statement of Danrip Int.’s loss. They had a written order for the delivery of 10 units with Simpro in Germany, but having received information from Flemming Stanley that Danrip Int. had no title to the product, Simpro did not want more than the 4 units they had already been supplied with and resold to German end-users. Sales went fairly well on the Scandinavian market until Bergman in Sweden received threatening letters from Flemming Stanley concerning actions for damages. He did, however, manage to convince Bergman that there were no problems. In August 1993 he and Stig Mikkelsen were in Norway to negotiate, but the customers in question in Norway thought they were from StanTech and therefore, the Norwegians did not wish to do business with them. Even though they explained that they represented another company, such uncertainty had been created that nothing came of it. In England they tried to establish contact to an English distributor of graphic articles but this failed as well due to Flemming Stanley’s interference. The customers were not satisfied with the Danrip because of the defects. He cannot remember whether the lacking orders were also due to this. Neither the Yarcrip functioned a fully 100%.
The loss suffered in the USA is calculated on the basis of previous turnover. The loss suffered in Japan is based on an assessment by Yarc on previous business. He does not remember whether the losses suffered in Norway and Sweden are based on actual orders.
It has been necessary to pay travelling expenses in order to keep in contact with customers as Danrip Int. because of Flemming Stanley’s statements could not continue the marketing employed so far.
Sero’s replacement of the Danrip with the Yarcrip was probably caused by the defects to the Danrip. The Yarcrip did not function at a 100% at that point. He does not remember, however, whether the orders they received were due to the non-functioning of the Danrip.
The experts Christian Christoffersen, Peter Johansen and Christian Rosenlund accepted the Opinions prepared.
At a meeting on 26 June 1995 the Report was examined and points were found that were due to misunderstandings, but these points did not cause them to change the conclusion. Yarc was encouraged to comment upon the similarities found but never did.
The tests that were made with the purpose of finding errors and characteristics show that Yarc’s rip and Danrip are different from standard rips and that there were many common errors and characteristics in the two rips. There is no doubt that partial copying of StanText’s source code has taken place. They had to examine the functionality seeing that they did not have both source codes. Initially, they suggested that they look at the source codes in order to be able to work as unrestricted as possible as they viewed the source code as a sort of distinctive feature of business. Yarc did not produce the source code or documentation apart from what appears in the supplement to Yarc’s questions to the first Experts’ Opinion. Production of the source code would have made it easier for the experts to reach a conclusion. They were not denied access to Yarc’s source code but it would have been very difficult.
Adobe’s manuals are freely accessible and all material may be forwarded.
As to discussion 5 in the replies to Yarc in February and April 1996, it may be added that the possibility exist of a rounding error when conversion for printing is made, but there are no errors in StanText’s source code.
Comparison of Screen Dumps for the two products shows different positions of the windows but this is insignificant as you decide on your own positions. It is the design of the dialog boxes which is decisive of the conclusion.
Different hardware is needed for the two rips, for which reason they have different interface.
In the readline test there is a difference which goes against the copying theory but the error handling is similar, differences are found as well in the way vmusage is handled. Vmusage is hardly copied directly from StanText’s source code but is probably cloned.
Having reviewed the 8 different tests, they probably would not draw quite the same conclusion holding the knowledge they do now. Later examinations compared to the Opinion from May 1995 have shown that they were wrong. A number of factors indicate copying and the last examinations have supported this conclusion even though some of the initial factors have been dropped. The position of the dialog boxes could indicate a modification of another program which weakens the conclusion.
The start-up functions are similar to a certain extent as they show the same output but they are formatted in different ways.
The laserwriter test mentioned under discussion 13 in the reply to Yarc of February 1996 gives no answer to Yarc’s question and shows nothing as to copying.
The tests on which the reply to Yarc of 28 February 1996 is based show that copying, cloning of the source code or use of feature may have taken place.
As to discussion 17 in the reply to Yarc of 5 February 1996, it is the similar reporting of content which indicates copying. According to the tests made there is a difference between on the one side Yarcrip/Danrip and on the other side other rips. As to discussion 18 b, c and d of the same reply, they agree with Yarc.
The internaldict test, cf. discussion 20, is not described in Adobe and shows in itself nothing in relation to copying and has nothing to do with the source code. As to discussion 21, it has been noted that Yarcrip and Danrip on many points amplify the manuals, amplification which is only seen in these two products that also show the same errors.
Christian Christoffersen explained that he withdrew as expert because of pressure from Karsten Jeppesen who during a telephone conversation said that if the Experts’ Opinion was not withdrawn, Karsten Jeppesen would spread rumours about him.
The expert Eskild Fjord Petersen accepted the Opinion made by him.
In his report Professor Karplus seems to be making a comparison between versions Danrip 1.2 and Yarcrip 1.a which he has made as well, but he does not know whether he and Professor Karplus were asked to answer the same question.
He is not sure whether Professor Karplus had version 1.a. Had Professor Karplus compared with Yarc version 1.a, he probably would have reached another result.
Yarcrip version 1.a seems to be a copy of Danrip 1.2 but it contains many more items. It therefore seems to be correct what Flemming Stanley states in his letters to Professor Karplus.
As to the reply to question 3.C.1-5 from Yarc: A number of functions exist which cannot be copied to ASICS but they may be found somewhere else. Certain functions may be copied to ASICS, but outside ASICS important functions may be found as well. That certain functions may be copied from Danrip to Yarcrip is evident because of the similar functions in ASICS.
The question is whether we are talking about copying or whether they simply make the same as to functionality. This is a difficult area as we are dealing with functions of a very high level. The functions are special in relation to developing a rip and we are not talking about standard functions.
The validity of the exclusive distributorship agreement
Under the exclusive distributorship agreement the source code was to be deposited and in case of StanText’s bankruptcy handed over to Danrip Int. The rights to the source code therefore passed on to Danrip Int. upon StanText’s bankruptcy.
The Plaintiffs deny that the exclusive distributorship agreement terminated at any time prior to the bankruptcy. Only draft framework agreements exist, but nothing in writing that the exclusive distributorship should have ceased. In the summer of 1992 merely a technical alteration of the agreement and a re-negotiation were made. The alterations made concerned prices and discounts only and such alterations had been made earlier. In addition, an alteration was made as to the geographical territory, which involved an option in relation to Germany. During the negotiations no other matters were discussed. The Defendant has failed to prove that a new oral agreement was entered into as both parties at that point were represented by attorneys, it is unlikely that the written agreement was replaced by an oral agreement. If the agreement had been terminated in the summer of 1992, StanText should have bought back the goods supplied which StanText was not capable of financially. The rights belong to Danrip Int. which the bankrupt estate has been informed of in writing.
Yarc’s and Karsten Jeppesen’s rights won on another basis:
Yarc has since its formation in 1988 continuously developed accelerators which is evident from Exhibits 33-45 to this case. In this way as well as through its cooperation with StanText Yarc became familiar with the technology and Karsten Jeppesen developed the driver in the Danrip and the queue system and sold the rip for which reason he must be familiar with it. It is therefore alleged that a joint venture relationship developed between the parties. They followed the same direction as concerns the platform and Stantext received from Yarc a number of manuals to use for the development hereof. Through their cooperation a joint development of the product took place, a product which was unique because of its speed which was achieved thanks to Yarc and without the participation of Yarc, the product would not have existed. We are therefore talking about a shared product and they therefore hold the same rights to the product. Accordingly, we are not dealing with copying and the copying theory is built on Flemming Stanley’s theory as to the break-in. In addition, material parts of the technology applied for the product is freely accessible.
The Plaintiffs’ alleged copying:
The break-in was committed some time between 6 April and 15 April 1993 and having examined Karsten Jeppesen’s affairs the police discontinued the investigation. Nothing indicates that Karsten Jeppesen should have made this break-in with a view to later copying to the advantage of the Plaintiffs. As Karsten Jeppesen already had access to the source code, there was no reason to break in. Furthermore, Trevor Marshall would hardly be likely to offer the liquidator money if he already had the source code. Karsten Jeppesen was not employed with Yarc until September 1993 and had therefore no reason to acquire the source code as he had nothing to use it for. He did in addition take part in discussions on the formation of a new firm together with Flemming Stanley. According to the liquidator the break-in did not take place after the bid making. If Karsten Jeppesen did not do the break-in, when is the copying then supposed to have been effected. Karsten Jeppesen previously had back-up copies of the source code, he has, however, destroyed them. Thus, it has not been established that copying took place or that the source code was removed to the advantage of the Plaintiffs. The Defendants allege that from a technical point of view it would not have been possible for Yarc to develop the Yarcrip in such a short time without the source code, but the press release of 30 April 192 does not concern the Yarcrip which was not ready until November 1993 when it was still defective. Accordingly, the product was finished much later than stated by the Defendants. If the Plaintiffs had copied the source code, they could have put the product on the market already in May 1992. On the contrary, the Plaintiffs had to use the Danrips already purchased to “fill the gap” until their own rip was finished.
The expert Eskild Fjord Petersen did not compare the correct rips and has acknowledged as well that no copying can be made of Danrip’s board to Yarcrip’s ASICS but he still persists on his conclusion.
The rip that was sold to Sero was a Yarc/Danrip version 1 which was held in stock at the time of the bankruptcy. The sale is therefore conducted in complete compliance with the provisions of the Copyright Act, and the correction of errors on the rip was likewise warranted. Nothing odious is therefore found in the fact that the products are identical.
The Plaintiffs deny that they slowed the survey which has been unusual. It is likewise unusual to make reference to tests which the experts say are the same but which are not the same. The question of copying may only be determined by a comparison of the codes but this has not happened even though the experts were offered by Yarc to have the source code handed over, which they refused. The tests on which the subconclusions are based are produced by the Defendants to prove that copying took place. Yarc has demonstrated a number of irregularities in the tests and in conclusion we may establish that errors are not found in both rips seeing that the test results are not consistent. All conclusions in the first Experts’ Opinion were thus changed following Yarc’s supplementary questions and the conclusions diverge. The fact that subconclusions are changed in all the reports indicate that the reports are inaccurate and faulty. Furthermore, when judging the Opinions, it must be taken into consideration that the parties’ technology is based on a shared library of routines.
We regret if the Plaintiffs’ statements have been perceived as threats by the expert, but the Plaintiffs have only reacted to Flemming Stanley’s abuse of the Experts’ Opinion.
Protection of rights:
The Plaintiffs do not deny that computer programs may be protected. A rip, however, is no breakthrough as the ideas and principle behind a rip are well-known and not protected by legislation. It, thus, contains a number of functions which are dictated by its use. Flemming Arne Jensen assessed alone the software part, for which reason his Opinion has no value. The requirement as to the level of originality implies that a program is protected if it is original in the sense that the object must have been produced by the originator’s own intellectual creation. According to Christian Rosenlund’s explanation nothing original is found in StanText’s rips except set screens and many rips incorporate the same functions as the rips in this case. The requirement as to level of originality has not been met in this case as more must be required to obtain protection under the very few authorities in this field. Furthermore, the dual development view implies that StanText’s rip cannot be protected.
The Plaintiffs’ personal liability:
Liability for the acts of a company must lie with those persons who have the commercial interest in the company. In this case those persons who the Defendants want to hold liable only own a small amount of the capital. The Defendants’ claim in relation to Karsten Jeppesen must be dismissed as he - in the opinion of the Defendants only - developed the spooler system.
Claim of DKK 302,560 according to invoice
The credit note was issued by StanText to compensate the loss on version 1.1, and evidence has not been produced to the effect that conditions - such as the reduction in prices - were linked to the issue of the credit note. The Plaintiffs then deny that the claim was assigned from the estate. The issue of the invoice reflects an act of harassment from the part of the Defendants and, besides, the claim was advanced at a very late stage in this case.
The Plaintiffs’ libel claims
The statements under claim items a and b include charges of false marketing and handling of stolen goods as well as illegal copying, the statements under items c and d are inaccurate compared to facts and the statements under items e to h include charges of criminal offences and are grossly offensive. The statements are made to discredit the Plaintiffs and Third Parties and it is an aggravating circumstance that the statements were put forward to a large group. We are, in addition, dealing with repeated accusations, also after the completion of the investigation. Furthermore, the statements under item g clearly show that Flemming Stanley intentionally made the charges which are grossly offensive in both form and substance. Pursuant to Section 269 of the Danish Criminal Code such statements are not punishable if justification is produced, but this has not been done. The charges have been made without any support and Flemming Stanley has had no reasonable grounds to believe them to be true, for which reason they are subject to Section 268 of the Criminal Code concerning libel. The Plaintiffs dismiss that we are dealing with retaliation from the part of the Defendants as it was Flemming Stanley’s repeated scribbling which gave cause to the statements from Trevor Marshall.
Damages to the aggrieved parties must be fixed in view of the gravity, including the fact that Flemming Stanley contacted the Plaintiffs’ customers, thereby imposing on the Plaintiffs a considerable loss.
The statements concerning Karsten Jeppesen are made after the completion of the investigation for which reason Flemming Stanley has no instances in support of these statements. Third Party 1 is in the light of this entitled to damages for injury to reputation, cf. Section 26 of the Danish Act on Liability.
Furthermore, the Court should impose a considerable fine on the Defendants.
The validity of the exclusive distributorship agreement
The Defendants allege that the exclusive distributorship agreement was terminated at the beginning of July 1992 and that Plaintiff 2 no right had to the source code on the occasion of StanText’s bankruptcy.
According to the concordant explanations from Flemming Stanley, Stig Mikkelsen and Hans Peter Glunz business relations were changed in mid 1992 which Hans Peter Glunz has agreed to as he signed one of the draft agreements. The parties adhered to the modified agreement which is partly evident from the agreement with Fotoclark. None of the draft agreements include terms as to the surrender of the source code which was Danrip Int.’s guarantee of continued sales. As the Plaintiff no longer held a sole distributorship this “guarantee” could not be preserved. Following the meeting in July 1992 the exclusive distributorship agreement had in reality lapsed and therefore, the Plaintiff has no right to the source code. The Plaintiff has not set up claim for the surrender of the source code.
Yarc’s and Karsten Jeppesen’s possible contributions to the development of the Danrip
The Defendants deny that Yarc or Karsten Jeppesen participated or contributed to the development of the Danrip in such a way that Yarc or Karsten Jeppesen hold independent rights in relation to the Danrip.
StanText’s approach to Yarc only concerned accelerator cards with a view to increasing the speed of the rip product developed by StanText, and Yarc was not engaged in the rip technology. Neither later approaches concerned the rip technology. Jim Potter’s visit in the summer of 1992 related to the files for Yarc’s coprocessor only, the files as to which the Plaintiff has now withdrawn his claims. Since their bankruptcy the Defendants have bought no accelerator cards from Yarc.
Only subsequently have Yarc and Karsten Jeppesen obtained knowledge of the technology. According to the statements given by Mogens Kjaer, Anne Marie Christiansen and Karsten Jeppesen, Karsten Jeppesen participated in the development of StanText 8000 which is a word processing system and which contains nothing on rip technology. Neither the queue and communications systems contained rip technology. Therefore, from his employment with StanText Karsten Jeppesen had no in-sight worth mentioning into the rip technology and would not himself have been able to develop a rip. Even if Karsten Jeppesen had participated in the development of a rip, the right to such a rip belongs to the Defendants in pursuance of Section 42b, now Section 59, of the Danish Copyright Act.
The Plaintiffs have not tried to document that Yarc participated in the development and the experts were not asked to describe in what fields Yarc assisted.
The rights transferred by the bankrupt estate and the legal protection hereof
The Defendants allege that the bankrupt estate was entitled to deal with the rights to the Danrip, that the Defendants have lawfully acquired the sole right to the Danrip, that the Danrip is of the required level of originality to make the software part protected by copyright and the hardware part protected by a combination of the Copyright Act and the Marketing Practices Act.
The liquidator’s reservation as to any third party’s possible prior right is a technicality that is without significance to this case. The rights, including the rights as to reverse engineering, belonged to the estate and were freely transferable and they consist of 98% software and 2% hardware.
According to the opinions from Flemming Arne Jensen and the experts we are talking about a sophisticated and highly complex product of approx. 100,000 lines for which development approx. 25 man-years will be required. At the time of the bankruptcy it was the world’s fastest product. That the product in addition was of a sufficient level of originality is supported by Flemming Arne Jensen who based his assessment hereon and also the experts find that we are dealing with a complex product which differs from other products on the market. Furthermore, independent, creative performance exists on the part of Peter Tolboell.
The Plaintiff’s illegal copying of the Danrip
The Defendants allege that the products marketed and sold since 30 April 1993 by the Plaintiffs and partly by the Third Parties represent or is based wholly or partially on the copying of the Defendants’ Danrip. Accordingly, the Defendants’ deny that the Plaintiffs were or are capable of developing their own rip from scratch.
The most important proof is the Experts’ Opinions which the experts have accepted and they have as well confirmed that copying took place. The 20 new tests also supported the experts’ first conclusion. The differences in the products are not what is interesting in this case, but the characteristic similarities are - similarities which are not determined out of necessity.
The Plaintiffs have obstructed the survey and made threats against one of the experts. In addition, the Plaintiffs have opposed the survey and refused to hand over the source code, probably in order to have time to work on the copied rip. Peter Tolboell’s test which was based on the Sero board alone and not on the new Yarcrip was sufficiently effective to prove that reverse engineering had taken place. The Experts’ Opinion confirms that the purpose of the test was to discover identical errors in the product and not defects in the functionality and is neutral. The experts have not been able to see any alternative methods to this effect. Had the Plaintiffs received this test material before, the Plaintiffs would have been able to recover the errors. The Plaintiffs, in addition, accepted the method and only raised objections when the result was known. The experts have also encouraged the Plaintiffs to explain the similarities. Thus, there is no reason to doubt the neutrality of the tests. On the basis of a Yarcrip purchased by the Defendants the experts were able to develop the 20 new tests. According to the expert Peter Johannesen the result of the 20 tests was that copying had been made by the Plaintiffs.
Whether we are dealing with a complete or partial copy is difficult to decide but seeing that neither Yarc nor Karsten Jeppesen had any knowledge of the rip technology and seeing that Yarc despite financial difficulties was very soon to launch a product, the Defendants take into account that a complete copy was made. Yarc has not tried to deny this claim as they contends that they had not the time for it which must have a prejudicial effect on their case.
The experts refused that the reason for the identical errors is a shared library of routines. The experts refused that we are dealing with cloning as it is unlikely that the same errors would be made if copying had not taken place.
The Plaintiff has admitted reverse engineering which has also been demonstrated by the experts. No evidence better than the reference to board no. 31, which is one of the 4 boards that Danrip Int. never paid for, and which were later found with the firm Sero may be produced. The Plaintiff has thus sold lots of copied products, probably more than the 30-40 stated, all of which has happened at the expense of the Defendants. The illegal copying consists in the appropriation and subsequent illegal use of the source code. The Plaintiff has acted contrary to Section 37 of the Copyright Act and as concerns the hardware contrary to Sections 1, 2 and 10 of the Marketing Practices Act and the Plaintiff continues to violate these provisions.
The Plaintiffs’ claims for damages and punishment of the Defendants
The Defendants allege that the Plaintiffs’ production and marketing of rip products based on complete or partial copying of the rights belonging to the Defendant and the Plaintiffs’ market-related conduct towards the Defendants in other respects unlawfully have inflicted a loss upon the Defendants, a loss for which the Defendants are entitled to damages. In addition, the Defendants allege that the Plaintiffs’ violations of the Copyright Act and the Marketing Practices Act, respectively, are of such gravity that these violations should lead to punishment.
14 days after the Plaintiffs had acquired the rights from the estate the press release from Danrip Int. was issued. That it says Danrip products does not change matters as everyone knew that what it was about was the Danrip. In addition, they wanted to develop further the Danrip even though they had no rights hereto. The Plaintiffs’ approach to the Defendants’ customers inflicted loss upon the Defendants as the Defendants subsequently had to make all the sales work themselves and were forced to explain to customers how matters were. Due to the Plaintiffs’ conduct the customers, furthermore, thought that defective products came from the Defendants which was not the case. A specific statement of the loss is difficult to make but we are talking about a considerable loss which exceeds the Defendants’ claim.
As to the claim set up concerning punishment reference is made to Section 22, Subsections 3, 4 and 6 of the Marketing Practices Act and Section 76, Subsection 1 and Section 80 of the Copyright Act.
The Third Parties’ personal liability
The Defendants allege that the Third Parties may be held personally liable for the violations of the Copyright Act and the Marketing Practices Act and the Defendants also allege that personal liability is required in order to put an effective halt to the violations.
Identification should be made between the companies acting as Plaintiffs and the Third Parties who are joint-owners as regards 9-11% of the capital. The Third Parties took the initiative to form the company and limited liability is not intended as a shield from liability following from criminal offences. Karsten Jeppesen had a copy of the product in his possession from before the bankruptcy and it has not been proved that the source code was destroyed. According to his own explanation Karsten Jeppesen was the prime mover in the development of the Yarcrip and in the liquidator’s opinion Trevor Marshall had possession of the source code.
Hans Peter Glunz was aware of the plans and sold 60% of Danrip Int. to Yarc. Hans Peter Glunz was personally committed and still is as joint owner and manager of the company.
Both Stig Mikkelsen and Erik B. Hansen participated actively and personally in the marketing and they visited a number of distributors even though they were familiar with the true facts of the case. Therefore, they have undertaken personal liability.
In view of the gravity of their actions the Third Parties must be jointly and severally liable with the Plaintiffs, also as concerns the costs of the case.
The Defendants’ claim for payment of DKK 302,560
The Defendants allege that under the invoice drawn up Plaintiff 2 is liable to pay DKK 302,560 to the Defendants in their capacity as lawful transferee of the bankrupt estate’s claim for such payment seeing that Plaintiff 2 breached the agreement on which earlier crediting of the amount was based.
Crediting of the boards took place upon Hans Peter Glunz’s demand and was accepted by Flemming Stanley. Danrip Int. did not reduce the price of the product as agreed which was a clear condition of the crediting. Flemming Stanley pointed out that the terms of the crediting had not been met and once again forwarded the invoice. That the Defendants were subrogated to the bankrupt estate’s claim is undisputed.
As to the Plaintiffs’ libel claims
The Defendants deny that they may be sentenced neither pursuant to the Criminal Code nor the Marketing Practices Act for the statements indicated both because the Plaintiffs have behaved unduly towards the Defendants and because the Plaintiffs have retaliated against the Defendants.
The Plaintiffs who claim that their honour be rehabilitated and have set up claim for compensation have themselves committed gross infringements, cf. the Marketing Practices Act and the Copyright Act. They have made threats against the expert, etc. and are in no position to claim rehabilitation of their honour.
Flemming Stanley alone made the statements and justification of their correctness have been produced.
At the break-in on 14 April 1993 the most important components concerning the Danrip were removed. Shortly thereafter, Karsten Jeppesen said that he had had a copy of the product, a copy which he asserts has been destroyed. Even though Trevor Marshall was not successful in acquiring the rights from the estate, the Plaintiff put a product on the market shortly thereafter. Trevor Marshall later tried to obtain a declaration from the estate in order to legitimate the marketing, but he failed. The Defendants therefore had to inform the market as done.
The Plaintiffs have disputed the statements in Seybold and in addition brought an action for libel in California. The expert in California was asked to compare “apples and pears” for which reason judgment was given against the Defendant. The judgment was widely distributed and the Defendant must be entitled to communicate that the expert was misled.
Flemming Stanley’s statements in the Court in Arhus were made in response to the judge’s questions. Flemming Stanley, who was summoned as a witness was forced to answer. That no charges were brought against Karsten Jeppesen were because the police did not wish to make provisions for the investigation.
The Defendants therefore allege that Section 267, Subsections 1 and 3 and Section 272 of the Criminal Code are applicable and that Section 268 of the Criminal Code is not applicable as the statement was not disseminated in bad faith by the Defendant.
If Flemming Stanley is found guilty, the sentence must lapse in pursuance of Section 84, Subsection 2, cf. Subsection 1, no. 4 of the Criminal Code. The Defendant’s statements are alone due to the Plaintiffs’ behaviour and the claims made for damages have not been substantiated. Any loss the Plaintiffs may have incurred is alone arisen out of the Plaintiffs’ own unlawful behaviour.
Comments of the Court:
In re H 185/93
Re: Defendants 1, 2 and 3’s independent claim against the Plaintiffs and the Third Parties:
The Court is satisfied on the statements given which are in keeping with the draft framework agreements that StanText and Danrip Int. in connection with negotiations in the summer of 1992 agreed to changes in the exclusive distributorship agreement both in relation to prices and in relation to the geographical territory of the sole distributorship. Despite the fact that the framework agreement was never signed in identical copies it was adhered to both as concerned the geographical change and the issue of the credit note. The initial exclusive distributorship agreement is therefore held to have lapsed de facto. Taking into account that the reason for the deposit of the source code was the protection of Danrip Int.’s exclusive distributorship agreement, Danrip Int.’s right to the source code is held to have been forfeited as it is not mentioned in the framework agreement. Therefore, the estate was entitled to sell the source code to StanTech.
The court finds it undisputed that DanRip Int. did not participate in the development of the Danrip. When the cooperation between StanText and Yarc was initiated, StanText had worked with rip products for the graphic art industry for a number of years. StanText’s approach to Yarc was alone concerned with the supply of accelerator cards. On the evidence, the Court is not satisfied that Yarc’s employees participated in the development of the Danrip, neither through the correspondence that took place between the parties nor through visits to Denmark. On the information the Court has on Karsten Jeppesen’s work with StanText, it is not assumed that he prior to his resignation in connection with StanText’s bankruptcy obtained such knowledge of the rip product that he within a short time would be able to be responsible for the development of a rip. The Court therefore finds no grounds on which to establish that Yarc and Karsten Jeppesen contributed to the development of the Danrip in such a way that they have acquired independent rights hereto.
From the Opinion made by Flemming Arne Jensen in connection with the assessment made for the estate it is evident that there was no documentation of the source code and that any would-be buyer would not be able to maintain and further develop the product. From the Experts’ Opinions prepared it appears that several similarities are found between the Yarcrip and the Danrip also on points where such similarities from an objective point of view must be characterised as errors. In spite of the fact that the Plaintiffs in connection with the survey and verification in court by the experts were reaffirmed that the result of several of the tests, seen in isolation, could not be held as proof of copying, the experts insisted that the tests on an overall assessment demonstrate that Yarcrip and Danrip originate from the same source which differs from other rips on the market and that this is not due to freely accessible libraries. In accordance with the Experts’ Opinions the court finds it not risky to establish that the Yarcrip is based on the copying of the Danrip.
By having produced and by having sold a rip based on technology belonging to the Defendants, Yarc and Danrip Int., respectively, are held to have violated Sections 1 and 10 of the Danish Marketing Practices Act. On the evidence the Court is not sufficiently satisfied that the Defendants’ rip is protected by the Copyright Act. The Defendants’ claim 1, item 1 is therefore allowed.
The remaining part of the Defendants’ claim 1 is of such unspecified and inaccurate nature in relation to a possible enforcement and may actually lead to a considerable restriction in marketing that the Court finds no authority to impose such restriction on the Plaintiffs for which reason the Defendants’ claim in this respect is dismissed.
As the way in which Yarc informed the Defendants’ customers that a dispute had arisen between the parties is held to have resulted in considerable costs for the Defendants for the re-establishment of the market and the decline in sales the Defendants’ must have experienced, damages are in the discretion of the Court fixed at DKK 750,000 plus interest as claimed.
On the evidence, the Court finds no reason to impose any sanction on the Defendants.
As the Court is not satisfied that the Third Parties independently participated in the copying and sale of the Danrip, the Court finds no reason to impose any sanction on them or order them to pay damages due only to their positions within the company, for which reason claims against them are dismissed.
Re: Defendant 2’s claim against Plaintiff 2:
In § 3 of the draft framework agreement it says that Plaintiff 2 free of charge could keep 4 items of the Danrip. Neither from the wording of the framework agreement nor from the comments attached thereto from Stantext’s attorney does it appear that the free surrender was subject to conditions. Hans Peter Glunz denied that the surrender was subject to conditions and the Court is not satisfied by Flemming Stanley’s statement to this effect that the surrender was subject to conditions, for which reason the Defendant’s claim is dismissed.
As to legal costs the Court finds as stated below.
In re 32/94:
Re: The Third Parties’ claim against Defendant 1:
As to para a: As this statement was made by the Defendant on request of the judge in the case in which he was summoned as a witness and therefore under an obligation to give evidence, the Third Parties’ claim is dismissed.
As to paras b and c: The accusations against which claims have been set up are, based on their contents, to be understood as charges of theft and fraud and they were put forward in a letter to a customer at a time when Defendant 1 knew that no charges would be brought against Karsten Jeppesen. As the Defendant has not produced justification of the statements indicated and on the above facts is held to have lacked reasonable grounds to believe the statements to be true, he is found guilty of violating Section 268 of the Danish Criminal Code. In view of the fact that the charge was not put forward in public and that it formed part of the peculiar statements made to customers the Court sentences the Defendant to 10 day fines of DKK 1,000 each, alternatively 10 days’ lenient imprisonment.
The Plaintiff’s claim for the revocation of the statements made under paras b and c which are all held to be unfounded is pursuant to Section 273, Subsection 1 of the Criminal Code allowed.
In addition, Defendant 1 is ordered to pay DKK 2,000 to cover the costs of the publication of the ratio decidendi and the findings of the Court, cf. Section 273, Subsection 2 of the Criminal Code.
As the Court finds the statements made to potential trading partners to be offensive to Karsten Jeppesen, Defendant 1 is ordered to pay damages in the amount of DKK 10,000 pursuant to Section 26 of the Danish Act on Liability in Damages.
As to legal costs the Court finds as stated below.
Re: Plaintiff 1, Plaintiff 2 and Plaintiff 3 vs. Defendant 1:
The statements made by Defendant 1 concerning the Plaintiffs’ rights to the technology on which the Yarcrip is based were made at times when the parties put forward various statements towards customers, in scientific journals and to the court on the legal position of the other party in the pending proceedings. The contents of the statements from both parties differs substantially from the way in which the market is normally informed of disputes as to the rights to a product. In view of this and the conclusions of the Experts’ Opinions the Court finds that the statements indicated under paras a, b, c, f and g do not fall within Sections 227 and 268 of the Criminal Code.
The statements under paras d and e contains accusations of false evidence and theft/handling of stolen goods but are made at a time when the investigation had not been concluded. As the Defendants believed to have established that copying of the Danrip had taken place, the Court finds the statements to fall within Section 267 of the Criminal Code only. According to the above stated as to the conduct of the parties before, during and after the times at which the charges were made, the Court finds the Plaintiff to have retaliated, for which reason the sentence shall lapse in pursuance of Section 272 of the Criminal Code.
On these facts the Court finds no reason for which to award the Plaintiffs damages.
As to legal costs the Court finds as stated below.
The judgment of the Court is:
In re H 185/93:
THAT Plaintiff 1, Yarc Systems Corporation, and Plaintiff 2, Danrip Int. A/S, shall discontinue any direct or indirect production, marketing and sale in and from Denmark of rip products based on the know-how utilised by StanTech.
THAT the Plaintiffs shall pay jointly and severally damages to the Defendants in the amount of DKK 750,000 plus interest from 24 August 1993 till payment is effected.
THAT the Court dismisses the other claims set up by the Defendants against the Plaintiffs.
THAT the Plaintiffs shall pay jointly and severally to the Defendants legal costs in the amount of DKK 200,000 plus the experts’ fees and costs.
THAT the Court dismisses the claims set up by the Defendants against the Third Parties, Karsten Jeppesen, Erik B. Hansen, Stig Mikkelsen and Hans Peter Glunz.
THAT the Defendants shall pay to the Third Parties legal costs in the amount of DKK 25,000.
In re H 32/94:
Re: Third Party 1, Karsten Jeppesen vs. Flemming Stanley:
THAT the Defendant Flemming Stanley shall be punished for violation of Section 268 of the Criminal Code by 10 day fines of DKK 1,000 each.
THAT the below statements made by the Defendant shall be held unfounded:
(b) “For your information our software was stolen from the estate by a person who held a key to the premises. Karsten as one of two had not returned his key: Do you not think that an unusually clear picture is gradually forming of what Karsten is involved in?”
(c) “To complete the picture, I inform you that he has now attempted to defraud the bankrupt estate, but for that he will, of course, be called to account as deserved.”
THAT the Defendants shall pay DKK 2,000 for the publication of the ratio decidendi and the findings and shall pay damages in the amount of DKK 10,000 to Karsten Jeppesen.
THAT the Defendants shall pay to Karsten Jeppesen legal costs in the amount of DKK 5,000.
Re: Plaintiff 1, Yarc Systems Corporation, Plaintiff 2, Danrip Int. A/S and Plaintiff 3, Trevor G. Marshall vs. Defendant 1, Flemming Stanley:
THAT the Court finds for Defendant 1, Flemming Stanley.
THAT each party shall bear his own costs.
Nils Foss Palle Mikkelsen
This is certified to be a true copy
For the Registrar of the Maritime and Commercial Court
this 28th day of November 1996.