Friday 20 April 2012

T 1553/06 – Direct And Unambiguous Access


This test case led to an interesting discussion of whether and under which conditions a document that could be accessed via the Internet was indeed available to the public within the meaning of A 54.

The Boards love things that are “direct and unambiguous”. The present decision is expected to revive the notion of “direct and unambiguous access”. 

If you do not want to have all the details, you might jump to the proposed test, which is found towards the end of the extract that follows.


Background

[6.1] The opposition division, in the decision under appeal, considered that document I1 formed part of the state of the art and anticipated the subject-matter of claim 1. In this respect, the board notes that document I1 includes (nearly identically) the wording of claim 1 according to the main request, together with pertinent parts of the description. I1 was allegedly found on the Internet on 15 November 1999 at the URL


Document I2 comprises the whole of document 11, together with an appended text portion, i.e. (nearly identically) the wording of claim 2 according to the main request (without the reference to claim l), together with parts of the description pertaining to that claim 2. The patentee referred to I2 as apparently being an update of 11. I2 was allegedly found on both 12 and 31 January 2000 under the same URL as 11. According to both parties, that URL no longer exists.

In the light of the foregoing the only novelty-related issue for the subject-matter of claim 1 is whether the content of 11, i.e. I1 and the corresponding part of 12, constituted prior art within the meaning of A 54 EPC 1973 at the filing date. Pursuant to paragraph 2 of that provision:

The state of the art shall be held to comprise everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the European patent application.

According to the case law of the boards of appeal, information is “available to the public” if only a single member of the public is in a position to gain access to it and understand it, and if said member of the public is under no obligation to maintain secrecy (see T 1081/01 [5], affirmed by T 1309/07 [3.2.1]). Whether or not a member of the public has actually accessed the information is irrelevant (see T 84/83 [2.4.2]) .

Whether the content of I1 forms prior art because it was communicated to the opponent’s representative

[6.2] The parties admitted that there was a substantial level of co-operation between them and other professional representatives in order to create a test case. For the board, this explains why it was possible that Mr Mooij, the representative of the opponent DSM IP Assets B.V. and also of DSM N.V. (DSM Patents & Trademarks) at whose request notarial records were drawn up, submitted the documents now referred to as I1 and I2 within the framework of third-party observations under A 115 EPC 1973. According to the notarial records I1 was found on the Internet on 15 November 1999 and I2 on 12 and 31 January 2000, i.e. before the filing date of the patent application from which the present patent originates. Mr Mooij’s access to the documents, now referred to as I1 and 12, before the filing date, however, did not put these into the public domain. This is because a non-disclosure agreement (NDA) not to divulge the respective information had been concluded. According to the patentee’s remarks in the oral proceedings (OPs) before the board, which the opponent did not contest, that agreement was binding, among others, for Messrs Cohen (patentee’s representative), Mooij (opponent’s representative) and de Vries (professional representative involved in the preparation of the test case) until publication of the patent application (which took place on 7 June 2000). The then applicant also mentioned in its letter of 10 October 2000 in the proceedings up to grant that “both Mr de Vries and Mr Mooij were bound to [sic] a non-disclosure agreement pertaining to the subject-matter of the present European patent application” […].

Whether the content of I1 existed on the Internet

Technical background and definition of terms

[6.3.1] In their written and oral submissions, the parties have used technical terms such as “Internet”, “Web”, “webpage”, “URL” and “web search engine”. In order to avoid any confusion and to facilitate the reader’s understanding of the technical issues, these and other technical terms of the decision will now be defined and briefly explained. The following explanations refer to the situation at the time of the alleged publication of documents I1 to I4 i.e. between 15 November 1999 and 31 January 2000.

The “Internet” is a global system of interconnected computer networks which communicate via a standard Internet protocol suite (TCP/IP). The “World Wide Web”, also commonly referred to as “the Web”, is a network of interconnected documents or information resources on the Internet, containing text and/or multimedia information. The terms “Internet” and “Web” are thus not equivalent, the Web being only a part of the Internet. Other parts of the Internet include, for instance, the infrastructure to support “electronic mail” (commonly known as “e-mail”).

A “webpage” is a document or information resource which is stored on the Web and can be accessed, for instance, via a “web browser”. In order to access and display a webpage an Internet user can enter the “Uniform Resource Locator (URL)” of the webpage in a web browser or click on a “hyperlink” leading to that webpage, the URL forming a specific address at which the webpage can be found on the Web. Alternatively, in particular when the URL is not known to the user, a webpage of interest can be found by entering a search query (typically one or more keywords) in a “web search engine”. The web search engine then returns a list of webpages which match the entered keyword(s). A webpage can only be found via a web search engine by entering keywords if the search engine knows its URL and has indexed the webpage with keywords. There are basically two ways in which a webpage can become known to a web search engine: (1) someone, usually the creator or administrator of the webpage, supplies its URL to the search engine (it is common ground between the parties that the creator or administrator can also supply keywords to the search engine); or ( 2 ) the webpage is discovered by one of the “web crawlers” of the search engine. Web crawlers are computer programs which explore the Web in a methodical, automated manner. A common way for a web crawler to discover an unknown webpage is by following a hyperlink present on a known webpage which points to (the URL of) the unknown page. Once the URL of a webpage is known to the search engine, the webpage is accessed and indexed based on keywords generated by the search engine from the text (including hidden text known as “meta tags”) of the webpage.

The notarial records and the parties’ agreement

[6.3.2] The board notes that the opponent, in support of its contentions that I1 and I2 existed on the Internet, filed originals of the corresponding notarial records A1 (for 11) and A2 and the notarial record in T15 (both for 12) during the OPs. These records certify that I1 and I2 could be found on the Internet at certain dates and times after entering certain strings of keywords for each of them (see the details above, under point 11). Each of these originals bears the signature and stamp of a Dutch notary public, and I1 and I2 were attached to the associated notarial record(s). Both A1/11 and A2/12 and the notarial record of T15/12 were bound in one folder each. In addition, the patentee, in reply to a question by the chairman in the OPs, confirmed that I1 and I2 could be accessed on the basis of keywords. The board considers the parties’ statements to be credible given that this is a test case and the existence of I1 and I2 is a precondition for enabling the board to deal with the gist of the test case.

Conclusion on whether the content of I1 existed on the Internet

[6.3.3] In the light of the foregoing the board considers that the opponent has proven beyond reasonable doubt that I1 and I2 could be found on the Internet at the dates and times indicated in A1, A2 and the notarial record in T15 after entering the strings of keywords mentioned in these notarial records. Contrary to the opponent, the board regards the terms “beyond reasonable doubt” and the more colloquial term “up to the hilt” as being equivalent. […] Given that both parties agree that I1 and I2 existed on the Internet at the dates and times indicated in the notarial records, the question as to whether the reported requirements of the French courts for Internet disclosures […] would, in the absence of such agreement, have to be fulfilled (in the present case they are not) for acknowledging proof beyond reasonable doubt may therefore be left an open one. The same applies to the more general question as to the proper standard of proof.

Whether the content of I1 forms prior art
because the notary actually found it on the Internet

[6.4] As stated, it has been proven that a Dutch notary public found I1 on 15 November 1999 and, again, as part of 12, on both 12 and 31 January 2000 on the Internet on the basis of certain keyword strings.

However, the board does not consider that the content of I1 was divulged to the public by the fact that the notary public had seen it before the filing date. This is because from the test nature of the present case it must be concluded that the notary was under a duty to keep this content confidential. Otherwise the test case would be largely moot.

According to the opponent (reasoning against its case) it might be argued that the notary was not “the public”. Firstly, there was no evidence that the notary was a skilled person i.e. someone who would understand the content of what he found. Secondly, arguably, a notary would have a duty of confidence to keep all information obtained in a professional context secret. The opponent also maintained that it would be for the patentee to show that the notary was not a member of the public and, in the absence of pertinent evidence, the notary was.

The board disagrees that the notary was a member of the public. The board wonders, in relation to the opponent’s second point above, how the burden of showing that the notary acting on behalf of DSM N.V. (DSM Patents & Trademarks) had a duty of confidence could conceivably be upon the patentee. The notarised documents I1 and I2 were furnished by the opponent, DSM IP Assets B.V. Given the co-operation between the parties in this test case and the non-disclosure agreement mentioned above, the board considers it highly unlikely that the opponent’s representative, who also acted for DSM N.V. (DSM Patents & Trademarks), would have left it up to the notary’s discretion to decide whether or not to divulge the contents of I1 and I2 supposing arguendo the notary was not bound by professional rules or public law not to divulge them. This would have rendered the test case in relation to I1 and I2 largely moot. The opponent’s contentions in this respect are therefore not intelligible. As the board considers that the notary was not entitled to disclose the content of 11, as included in I1 and 12, the opponent’s first point whether the notary had the technical knowledge to understand that content is of no relevance.

Whether the content of I1 forms prior art
for the sole reason that I1 or I2 existed on the Internet

Introduction

[6.5.1] Regarding public availability of a disclosure the case law has distinguished a number of concepts. In T 952/92 [2.1] the board found :
whatever the means of disclosure (written description, oral description, use, etc.), availability in the sense of Article 54(2) EPC involves two separate stages: availability of the means of disclosure, and availability of information which is accessible and derivable from such means. (emphases added)
The board made this statement in conjunction with a quote from G 1/92 [1.2] : 
A 54(2) EPC [1973] does not make any distinction between the different means by which any information is made available to the public. Thus, information deriving from a use is governed in principle by the same conditions as is information disclosed by oral or written description.
The board in T 952/92 also said:
In other words, the disclosure of a written description is the information which a skilled person can learn by reading it, the disclosure of an oral description is the information that a skilled person can learn by hearing it, and the disclosure of a product which has been used is the information that a skilled person can learn from it, either visually or by analysis for example.
In this regard, the board deciding on the present case draws attention to the fact that the parties and a number of board decisions quoted from in the present decision sometimes use the term “information” as a synonym for “means of disclosure”.

In the patentee’s view it is the Internet that forms the means of disclosure in the present case, and the Internet must be exposed to deliberately chosen specific external conditions, i.e. a very special set of keywords, must be submitted, to find I1. This exercise constitutes “undue burden” as referred to in G 1/92, which excludes public availability of I1. (This view is analysed at point [6.7.2] below.)

The board notes as a preliminary observation that G 1/92 deals with the availability of information which is accessible and derivable from (the use of) a commercially available product. The availability of the means of disclosure, i.e. the availability of the product itself, however, is not in question in G 1/92.

As to the availability of the means of disclosure, the question of whether I1 forms prior art for the sole reason that it existed on the Internet, i.e. independent of whether it could be found by using keywords or by guessing the URL, logically precedes the question whether I1 became publicly available because it could be found on the basis of keywords or its URL. As a consequence, the board will start by considering that it is not the Internet but I1 itself that forms the means of disclosure.

The board will first explain the case law on the availability of written disclosures which does not rely on G 1/92.

Thereafter the board will analyse whether G 1/92 is also relevant for dealing with the question of availability of the means of disclosure, independent of whether that means is a product (as in G 1/92) or a written description or any other means of disclosure and, if so, what is the impact of G 1/92 on the availability of document I1 and I2, allegedly found on the Internet.

Finally, the board will compare the results of both enquiries and answer the question in the heading.

The case law on the public availability of written descriptions

[6.5.2] According to the EPO publication “Case Law of the Boards of Appeal of the EPO”, 6th edition, 2010, at I.C.1.8., referring to T 444/88, “Board of appeal case law has established that the theoretical possibility of having access to information renders it available to the public” . (emphasis added)

T 444/88 [3.1] dealt with a situation where, in the course of grant proceedings, additional example was introduced in an application document already available to public inspection. The board held that that example shared the fate of that document and was also “made available to the public” was not necessary as a matter of law that any member the public had to be aware that the example was introduced into the document, and therefore was available for inspection on request. It was sufficient that the document was in fact available to the public before the priority date of the patent in suit, whether or not this was known by any member of the public, and whether or not any member of the public actually inspected the document.

In the same vein it was considered in T 1127/00 [9] that “It is established jurisprudence of the Boards of Appeal … [citing the 4th edition of the “Case Law”, at I.C.1.6, 421 that the theoretical possibility of having access to information renders it available to the public.”

Moreover, the decision in case T 381/87 dealing with an article in a scientific journal may be seen as also expressing this view. Headnote 3 starts with the phrase:

If a document in a library “would have been available to anyone who requested to see it” on a particular day, such fact is sufficient to establish that the document was “made available to the public” on that day.

In the decision the board said (see point [4(4)(b)]) :
It is not necessary as a matter of law that any members of the public would have been aware that the document was available upon request on that day, whether by means of an index in the Library or otherwise. It is sufficient if the document was in fact available to the public on that day, whether or not any member of the public actually knew it was available, and whether or not any member of the public actually asked to see it.
Nevertheless, two other decisions, i.e. T 314/99 and T 186/01, in the board’s analysis, did not accept that a mere theoretical possibility of having access to a means of disclosure rendered it available to the public.

In T 314/99 the board was prepared to take the view that D6, a diploma thesis, arrived in the archive before the relevant priority date (see point [5.4.4]) but also said that:
However, in the Board’s judgment D6 did not by its mere arrival in the archive become publicly available, since that did not mean it was as of that point in time catalogued or otherwise prepared for the public to acquire knowledge of it, and because without such means of information the public would remain unaware of its existence. [5.5] … the possibility that the public could acquire knowledge or awareness of the existence of D6 is a precondition of its public availability before the priority date of the patent in suit. [5.6]
In T 186/01 [4] the board held that :
… the accessibility to the library for the public is not proven. D10 [an internal work report by the German Federal Institute for Agriculture (“Bundesforschungsanstalt fur Landwirtschaft”)] did not by its mere arrival in the archive become publicly available, since that did not mean that it was as of that point in time catalogued or otherwise prepared for the public to acquire knowledge of it, and because without such means of information the public would remain unaware of its existence. However, the possibility that the public could acquire knowledge or awareness of the existence of D6 is a precondition of its public availability before the priority date of the patent in suit (see T 314/99 [5.1-6]).
As to the above case law, the board questions that the three decisions cited in favour of a mere theoretical accessibility being sufficient were cases in which the document was only theoretically accessible:
  • in T 444/88, if an application document is already available for public inspection, the public will assume that it may be amended, e.g. by introducing an additional example;
  • in T 1127/00 [8] the document in question was an article published in a scientific journal. The nominal publication date was one day before the (relevant) second priority date of the patent in suit, a Thursday, and members of the public could purchase a copy of the journal from the editorial office on and after the Wednesday of each week in which it was published;
  • in T 381/87, the journal in question was issued regularly, and it was possible for an interested member of the public to ask for the latest issue (see point [4(4)(b)(ii)]).
Opinion G 1/92

(a) Background

[6.5.3] Even though G 1/92 only relates to the availability of information which is accessible and derivable from the (use of) a commercially available product, in T 165/96 the board did rely on G 1/92 in dealing with the question of availability of the means of disclosure. The board held that information was “available” once members of the public could theoretically become aware of it. The case dealt with technical information in an insert in a minor small-ads newspaper (circulation: 24000) distributed in the suburbs of Copenhagen. The patentee had argued that the publication should not be regarded as directly available to the public without undue burden relying on G 1/92. On this argument the board noted that the patentee’s interpretation would introduce into the consideration of novelty precisely that subjective element which in G 1/92 the Enlarged Board (EBA) had sought to exclude. That argument was therefore not valid.

Taking this holding at face value, one might conclude that the mere existence of I1 on the Internet was sufficient to make it become publicly available because, after all, members of the public could also theoretically become aware of it. In order to find out whether such a conclusion is warranted, the board will proceed to an analysis of G 1/92.


(b) The answer given by G 1/92

The referral to the EBA concerned the interpretation of the requirement “made available to the public” in relation to the prior use of a commercially available product. As stated, the decision deals with the availability of information which .is accessible and derivable from a commercially available product as a means of disclosure, not with the availability of the means of disclosure (i.e. the product itself). The EBA answered the questions put by the President of the EPO as follows:
1. The chemical composition of a product is state of the art when the product as such is available to the public and can be analysed and reproduced by the skilled person, irrespective of whether or not particular reasons can be identified for analysing the composition.

2. The same principle applies mutatis mutandis to any other product. (emphases added)

(c) Undue burden

In the Reasons for its opinion, the EBA made some “general remarks on the kind of information which can be derived from the public use of products for the purpose of the application of the requirement ‘made available to the public’ in A 54(2).” The EBA inter alia said:
[1.4] … Where it is possible for the skilled person to discover the composition or the internal structure of the product and to reproduce it without undue burden, then both the product and its composition or internal structure become state of the art.
2. There is no support in the EPC for the additional requirement … that the public should have particular reasons for analysing a product put on the market, in order to identify its composition or internal structure. … It is the fact that direct and unambiguous access to some particular information is possible, which makes the latter available, whether or not there is any reason for looking for it. 2.1 The introduction of such an additional requirement would … obviously represent an element of subjectivity leading to uncertainty in applying the concept of novelty as defined in … A [54(2) EPC]. (emphases added)
In this respect, the board in T 952/92 [2.2] noted that the reference in opinion G 1/92 to the term “without undue burden” was not strictly necessary for providing an answer to the referred questions, and therefore could not have been intended to alter or add to the existing law concerning what constituted “the state of the art”. The concept of reproduction of a product “without undue burden” was traditionally associated with the question of “sufficiency” of a description of an invention in a patent specification. The concept of reproducibility “without undue burden1’ had also been extended by analogy to cases concerning novelty, where a prior document described a product such as a chemical compound which was the subject of a claim in a patent. Such a description of a product did not render the product “available to the public” if a skilled person was unable to make the product, using his common general knowledge and “without undue burden” (in other words, in the absence of an “enabling disclosure”). The board in T 952/92 went on to say:
However, the extension of application of the concept “without undue burden” from reproduction of what has been described in a prior document to the discovery of what is not yet known about a previously sold product (namely, its composition or internal structure) would involve very different considerations, and the Board does not accept either that that was intended by the EBA in opinion G 1/92, or that it is legally correct. …
In the Board’s view, to apply the concept of “without undue burden” to the determination of the composition or internal structure of a prior used product which cannot be ascertained visually (for example by analysis) would introduce a subjective element into the determination of novelty, which was specifically rejected by the EBA in opinion G 1/92 … On the contrary, … in the Board’s view it is the fact that direct and unambiguous access to information concerning the composition or internal structure of a prior used product is possible, for example by means of analysis, which makes such composition or internal structure “available to the public” and thus part of the state of the art for the purpose of A 54(2). If such an analysis is possible in accordance with the known analytical techniques which were available for use by a skilled person before the relevant filing date, the composition or internal structure thereby is available to the public.

… the analysis by a skilled person of a product which has per se been “made available to the public” by means of prior sale for example, using available analytical techniques, can be considered as equivalent to the reading by a skilled person of a written description in a document which has per se been “made-available to the public”. The likelihood or otherwise of such a skilled person either reading such a written description, or analysing such a prior sold product, and the degree of burden (i.e. the amount of work and time) involved in such reading or analysing, is in principle irrelevant to the determination of what constitutes the state of the art. (emphases added)
The board in T 952/92 thus held that the phrase “without undue burden” G 1/92 [1.4] quoted above did not govern the “discovery” of the composition or internal structure, but only its reproduction.

(d) No “element of subjectivity”

As indicated above, in T 165/96 dealing with technical information in an insert in a minor small-ads newspaper the patentee relying on G 1/92 had argued that the publication should not be regarded as directly available to the public without undue burden. The board rejected this argument considering that the patentee’s interpretation would introduce into the consideration of novelty precisely that subjective element which in G 1/92 the EBA had sought to exclude. The EBA had rejected an additional requirement that the public should have particular reasons for analysing a product in order to identify its composition or internal structure because the introduction of such an additional requirement would represent “an element of subjectivity leading to uncertainty in applying the concept of novelty as defined in … A [54(2) EPC].”

The present board’s position

[6.5.4] As discussed above under point [6.5.1], in T 952/92 [2.1] the board found that:
whatever the means of disclosure (written description, oral description, use, etc.), availability in the sense of Article 54(2) EPC involves two separate stages: availability of the means of disclosure, and availability of information which is accessible and derivable from such means. (emphases added)
The present board agrees with this statement. Moreover, in the present board’s view, and in line with T 165/96, as the EBA ruled that applying the concept of novelty must not depend on an element of subjectivity, such element must not only be excluded in determining the availability of information which is accessible and derivable from a means of disclosure, such as the composition or internal structure of a commercially available product, but also in the preceding stage of determining the accessibility of such means. In both situations the degree of burden involved is in principle irrelevant to the determination of what constitutes the state of the art.

The board considers that this principle does however not mean that theoretical access to the means of disclosure is sufficient, as stated in T 165/96. In line with the decisions referred to above, under point [6.5.2], what is decisive is the practical possibility of having access to the means of disclosure. More specifically, this view can be derived from decisions T 314/99 and T 186/01 and is not contradicted by T 444/88, T 1127/00 and T 381/87. The board bases its stance on the assumption that if the public, in practice, remains unaware of the existence of a certain means of disclosure or has no access to the means of disclosure, then there is no reason to deny a patent applicant or patentee a patent in return for the divulgation to the public of the information embodied in that means of disclosure, subject to the remaining patentability requirements being fulfilled. In addition, considering theoretical access to be sufficient would make it possible to hide documents on the Web, which would nevertheless belong to the state of the art. Such documents may be referred to as hidden publications. These could be an attractive tool for some to put the success of a patent application or the validity of a patent into question. This is because it would be in the hands of the authors of hidden publications to unearth them at their whim. This would jeopardize legal certainty. The board therefore recognises a public policy interest in preventing hidden publications.

The considerations in the preceding paragraph do not run counter to the analogy made in the second paragraph of the present point [6.5.4] between determining the availability of information which is accessible and derivable from a means of disclosure, such as the composition or internal structure of a commercially available product, and determining the accessibility of such means, with the degree of burden involved being in principle irrelevant in both situations. Indeed, where practical access is not possible, this can be likened to the composition or internal structure of such a product not being discoverable by known analytical techniques and thus not being regarded as “made available to the public” because no “direct and unambiguous access” to it is possible.

For the above reasons the board considers that the requirement of “direct and unambiguous access”, set out in G 1/92 and T 952/92, should apply not only to access by the skilled person to information concerning the composition or internal structure of a commercially available product, but also to access by at least one member of the public to the means of disclosure.

In the case of the World Wide Web, given its enormous size, the fact alone that a document exists there does not go beyond mere theoretical accessibility. In order to conclude that the document is available to the public, it must further be established that direct and unambiguous access to it by known means and methods is possible.

Conclusion on whether the content of I1 forms prior art 
for the sole reason that I1 or I2 existed on the Internet

[6.5.5] It follows from the above that the mere fact that I1 and I2 existed on the World Wide Web is not sufficient to consider them as having been made available to the public. Until it is established that direct and unambiguous access to them by known means and methods was possible before the filing date, the access remains only theoretical and therefore does not meet the requirement of “made available to the public” within the meaning of A 54(2) EPC 1973.

The board will now examine whether direct and unambiguous access to I1 or I2, and thus to the content of I1, was possible.

Whether direct and unambiguous access to I1 or I2 was possible on the basis of their URL

[6.6] The (identical) URL of webpages I1 and 12, i.e. the character string  
represents a specific address indicating where to find I1 and 12 on the World Wide Web. In other words, the URL reveals the access path to I1 and 12.

As explained under points [6.2] and [6.4] above, this character string was not disclosed, either orally or in written form, before the filing date of the patent in suit to any person not bound by confidentiality, i.e. to a person outside the group of people consisting of those involved in setting up the test case and the notary public.

The URL of I1 and 12, however, was known to the web search engine AltaVista. Therefore the question arises whether a member of the public could have had direct and unambiguous access to this URL (and to I1 and 12) via AltaVista. This question will be discussed by the board in section [6.7] below.

Prior to that discussion, the board must also consider the possibility, submitted by the opponent, that the URL could have been guessed with some basic IT skills and a few attempts. If so, the URL might have provided direct and unambiguous access to I1 and 12. The board cannot exclude that there might be URLs which are so straightforward, or so predictable, that they could readily be guessed exactly and thus be regarded as providing direct and unambiguous access to the webpages at those URLs. However, this is conceivable only in exceptional cases. Otherwise, access to the content stored at the URL will only be theoretical, which, as found in section [6.5] above, is not sufficient for considering that content to be publicly available.

The board does not consider it to be straightforward for a member of the public, even one with IT skills, to guess the full URL
Guessing the term “morozov”, at any rate, would by no means be straightforward without knowledge of the terms in the URL.

In the light of the foregoing, the board concludes that guessing the URL of I1 and I2 did not provide direct and unambiguous access to I1 and 12, and thus did not make these documents, and thereby the content of 11, available to the public.

For the sake of completeness, the board observes that the situation in which a URL must be guessed in order to access a webpage is analogous to the situation in which access to a document is protected by a password which members of the public have no means of obtaining except by guessing.

Whether direct and unambiguous access to I1 was possible 
because it could be found on the basis of keywords

How I1 could be found by entering keywords

[6.7.1] I1 was found on 15 November 1999 on the basis of the keyword string “CATHODE RAY TUBE and GRID and THREE ELECTRON BEAMS and PHOSPHOR” in the AltaVista search engine. This is regarded as proven in view of notarial record A1 and of the fact that the patentee does not dispute it (see points [6.3.2] and [6.3.3] above).

The content of I1 was again found as part of I2 on the basis of
  • the keyword string “THREE CATHODES and PHOSPHOR SCREEN and CONVERGENCE” in the AltaVista search engine on 12 January 2000 (regarded as proven; see points [6.3.2] and [6.3.3] above) and
  • the keyword string “CATHODE RAY TUBE and GRID and THREE ELECTRON BEAMS and PHOSPHOR” in the AltaVista search engine on 31 January 2000 (regarded as proven; see points [6.3.2] and [6.3.3] above).
Given that I1 could be found on the basis of keywords, I1 must have been indexed. The parties agree that I1 was indexed […].

The patentee’s analogy to G 1/92

[6.7.2] As mentioned above, at point [6.5.1], in the patentee’s view it is the Internet that forms the means of disclosure in the present case. The patentee denied that document I1 was made publicly available due to the fact that it could be found on the basis of keywords. The patentee contended that in order to reveal the contents of documents I1 and I2 one needed to expose the Internet to a very particular interaction, viz. in the form of submitting a very special set of keywords or a particular character string as the URL of the document on the Web. Here the patentee drew an analogy with the exposure of a product to a deliberately chosen specific external condition which revealed particular properties of that product. In such a situation the information in point was considered in opinion G 1/92 of the EBA not to be available to the public. This opinion indicated that the requirement “made available to the public” was satisfied provided that there was direct and unambiguous access to particular information which made the latter available. In the present case, an analysis under specifically chosen circumstances of the Internet to derive the features of the claims as granted would constitute undue burden.

The board has already elaborated extensively on the criterion “undue burden” above, under point [6.5.3(c)]. In respect of the patentee’s submissions on the exposure to interaction with specifically chosen outside conditions, the board draws attention to the proviso of G 1/92 [3], which relates to “a commercially available product”:

It may be added that a commercially available product per se does not implicitly disclose anything beyond its composition or internal structure. Extrinsic characteristics, which are only revealed when the product is exposed to interaction with specifically chosen outside conditions, e.g., reactants or the like, in order to provide a particular effect or result or to discover potential results or capabilities, therefore point beyond the product per se as they are dependent on deliberate choices being made. Typical examples are the application as a pharmaceutical product of a known substance or composition (cf. A 54 ( 5 ) EPC [1973]) and the use of a known compound for a particular purpose, based on a new technical effect (cf. G 2/88). Thus, such characteristics cannot be considered as already having been made available to the public.

This passage was interpreted by another board as follows (see T 301/94 [3.6]): 


Item 3 of opinion G 1/92 refers to the use of a known compound for a particular purpose, based on a new technical effect as defined’ in G 2/88, and it is stated in this context that such characteristics cannot be regarded as having already been made available to the public when the compound itself is available to the public. The situation considered by the EBA … concerns a claim relating to a new use of a known compound reflecting a newly discovered technical effect and not a claim to the compound itself. It is the new technical effect which constitutes a hidden or secret feature, not the composition itself or one component thereof. (emphases added)
In the board’s opinion, the patentee’s view that there is an analogy between the need to submit a set of keywords, whether “very special” or not, to a search engine and the exposure of a product to an external condition as dealt with in G 1/92, cannot be accepted for the following reasons.

Whether the World Wide web can be regarded as a product in the sense of G 1/92

As a preliminary matter it should be recalled that G 1/92 deals with the composition or internal structure of a commercially available product, such as a chemical product or a mechanical or electrical article. One condition for this composition or internal structure to be state of the art is that it can be analysed and reproduced by the skilled person. In contrast thereto, the World Wide Web (the relevant part of the Internet in the present case) is essentially a global repository of interconnected documents stored in electronic form. In other words, the Web could be viewed as an enormous library of electronic documents.

Assimilating the Web (or a library) to a commercially available product in the sense of G 1/92, is thus already prima facie far-fetched and, in the present board’s judgement, goes beyond the intent of the EBA in G 1/92.

Moreover, in contrast to that of a “normal” commercially available product considered in G 1/92, the composition or internal structure of the Web constantly changes, in fact in fractions of a second, making it impossible in practice to discover at any given time its whole composition or internal structure by using known analytical techniques.

For these reasons, the board cannot accept the patentee’s argument that the Web should be regarded as a product in the sense of G 1/92 and that documents found on the Web by entering keywords would be akin to revealing extrinsic characteristics pointing beyond that product per se. It follows that the patentee’s analogy of the exposure of the Web to a very special set of keywords with the exposure of a product to a deliberately chosen specific external condition has no object.

Can a specific test be defined?

[6.7.3] As explained under section [6.5] above, the mere fact that a document existed on the World Wide Web is not sufficient to consider it as having been made available to the public. Unless it is established that direct and unambiguous access to it was possible before the filing or priority date, the access remains only theoretical and therefore does not meet the requirement of “made available to the public” within the meaning of A 54(2) EPC 1973.

Keyword(s) unrelated to the essence of the content of the document

In the board’s view, the fact that a document could be found with the help of a public web search engine by using one or more keywords does not automatically allow to conclude that direct and unambiguous access to that document was possible. This is a consequence of the fact that it is possible to store a document on the Web in such a way that it is indexed by a public web search engine only with (one or more) keyword(s) unrelated to the essence of the content of the document, thus making it impossible to find it by entering only keywords related to the essence of the content of the document.

The following fictitious two examples illustrate such a situation.

In a first example, an electronic image file containing a description of an improved television receiver is stored on a new webpage created at a hitherto unknown URL. The arbitrary character string “lk8zhd94j87hir” is attached as a keyword to the webpage, either as a title or as a meta tag. The creator of the webpage then supplies the URL to a public web search engine. The search engine follows the URL and indexes the webpage with the only keyword available i.e. “lk8zhd94j87hir”. It should be noted that the text of the description of the technical product is not detected by the search engine because it is stored as an image (i.e. as pixels, not as alphanumeric characters) and thus is not recognised as text by the search engine (assuming that the search engine does not perform optical character recognition (OCR) on the image).

In the above example, the description of the improved television receiver could be found via the search engine, but only by using the keyword “lk8zhd94j87hir” which, clearly, has nothing to do with the television receiver. The description of the improved television receiver would therefore be theoretically accessible by entering keywords, but impossible to find in practice.

In a second example, the facts are the same as in the first example except that the word “adoxography” (a late 19th-century word meaning “skilled writing on an unimportant subject”) is used instead of the arbitrary character string “lk8zhd94j87hir”. Although “adoxography”, in contrast to “lk8zhd94j87hirU, is a known word, it still has nothing to do with a television receiver and thus it would be at least questionable whether direct and unambiguous access to the technical content of the description of the improved television receiver was possible if it was only established that the document was found by entering the keyword “adoxography” in the search engine.

Time period during which the document remained accessible

Another important factor which must be considered in the case of a document stored on the World Wide Web is the length of time during which the document remained accessible. A document stored at a given URL may be deleted at any time by its owner/creator. It is thus technically possible to store a document at a given URL only for a very short period of time.

In order for the requirement of “direct and unambiguous access” to be met, it is thus a necessary condition that a document stored at a given URL on the Web remain accessible there for a sufficiently long period of time. Indeed, too short a period of time would effectively make it impossible to access the document. The minimum amount of time required for allowing direct and unambiguous access by a member of the public to a document must be assessed on a case-by-case basis, taking into account all the circumstances of the case.

Conclusions and proposed test

In the light of the above considerations, the board considers that the fact that a document stored on the World Wide Web could be found by entering keywords in a public web search engine before the priority or filing date of the patent or patent application is not always sufficient for reaching the conclusion that “direct and unambiguous access” to the document was possible. That conclusion may be correct in most cases but, under certain (unusual) circumstances, it may not be.

However, where all the conditions set out in the test below are met, the board is of the opinion that it can be safely concluded that a document stored on the World Wide Web was made available to the public. This test is defined as follows:

If, before the filing or priority date of the patent or patent application, a document stored on the World Wide Web and accessible via a specific URL

(1)   could be found with the help of a public web search engine by using one or more key words all related to the essence of the content of that document and 

(2)   remained accessible at that URL for a period of time long enough for a member of the public, i.e. someone under no obligation to keep the content of the document secret, to have direct and unambiguous access to the document, then the document was made available to the public in the sense of A 54(2) EPC 1973.

As to this test, the board draws attention to the following:
  • If any of conditions (1) and (2) is not met, the above test does not permit to conclude whether or not the document in question was made available to the public. In such a situation, in particular where condition (1) is not met, it must be examined on a case-by- case basis whether there were other circumstances possibly providing direct and unambiguous access to the document, such as a written or oral disclosure of the URL, the presence of the URL (e.g. in a hyperlink) on a webpage available to the public, the document being accessible via a public web search engine not using keywords as search inputs (e.g. based on similarities between images), publication of the document in a Web-based discussion forum, etc.

  • The questions of whether keywords are related or unrelated and whether the period of time was long enough will have to be determined in any single case and cannot be described beforehand in abstract terms. 

  • As A 54(2) EPC 1973 does not make any distinction between the different means by which any information is made available to the public, the board does not deem it necessary to make an enquiry as to whether a document stored on the Web that meets the conditions of the above test falls under one of the means expressly mentioned in A 54(2), such as “written description”. In any case, such a document will have been made available to the public “in any other way” .
The parties’ suggestions

[6.7.4] The following suggestions by the parties for devising a test on public availability of a document existing on the world Wide Web cannot be accepted.
  • Number of different keywords required   
The patentee suggested that it was necessary that not just one specific search string led to a certain webpage, but that a number of different (sets of) keywords did. The board does not consider it to be necessary to add as a criterion that more than one specific (set of) keyword(s) should lead to the document in question. Such a criterion would not add much to the test and would not do away with the need for assessing the keyword string on a case-by-case basis. The criterion is vague because it does not say how many possible strings would have to be there, in which way the keywords would have to differ, etc. Hidden publications could still be created on the basis of a number of different intricate search strings. Apart from that, the several search strings would have to be used at the same time because the Web content constantly changes.  
  • Intent of the submitter relevant 
As for the intent of the submitter that both parties suggested was an important criterion, in the board’s opinion, it is sufficient to recall that the EBA, in G 1/92, specifically rejected introducing a subjective element into the determination of novelty. Thus the intent-based argument must be dismissed on its face.
Other suggestions by the patentee on how to assess content submitted to the World Wide Web cannot be accepted either for the following reasons.
  • In-tray
If the test set out in previous section [6.7.3] is complied with, there will be no analogy between submitting information to the World Wide Web and sending information to an (albeit electronic) in-tray. Contrary to what the patentee contended, the information will then, by analogy, have to be considered as having been delivered to the public.
  • Permission to publish
Nor will there be an analogy to the permission to publish (the patentee had cited T 842/91 in this respect). If the conditions of the above test are met, then the document will be considered as actually published.
Application of the test to I1

[6.7.5] In the present case, document I1 was found with the help of the public web search engine AltaVista on the basis of keywords.

(a) The keywords were related.

As mentioned above at point [6.7.1], I1 was found with AltaVista on 15 November 1999 on the basis of the keyword string “CATHODE RAY TUBE and GRID and THREE ELECTRON BEAMS and PHOSPHOR” and the content of I1 was found again at the same URL as part of the content of I2 with the same keyword string on 31 January 2000 and on the basis of the different keyword string “THREE CATHODES and PHOSPHOR SCREEN and CONVERGENCE” on 12 January 2000.

The keywords used in each of those two strings were all related to the essence of the content of I1. This fact is not in dispute; it was acknowledged by the patentee during the OPs.

(b) The period of time during which I1 was accessible was largely sufficient

I1 was found on 15 November 1999 and on 12 and 31 January 2000 (I2 including 11) so that the board considers that I1 and I2 existed on the Web at the same URL for two and a half months and nearly three weeks, respectively. Both periods are obviously largely sufficient for a member of the public to have direct and unambiguous access to I1 and 12. It does not have to be decided whether the same result would have been achieved on the basis of the present keywords if it could be established that I1 could be found by search engines for a short period of time only, such as a period of less than 20 minutes on one particular day, as allegedly in the case of document I4.

As for the parties’ submissions emphasising the importance of the point in time after the filing or priority date until which documents must be found on the Internet in order to be considered as reflecting the situation for retrieving a document at the filing/priority date, the board is at a loss to see how this could be an issue in relation to document I1. I1, like the other documents I2 to I4 relied on as prior art in the present proceedings, were all retrieved before the filing date.

Finally, a member of the public (e.g. not the notary or the persons involved in setting up the test case) who could have found I1 on the Internet before the filing date was under no duty to keep the information embodied in I1 secret.

(c) Conclusion

Against the backdrop of the above findings, the board considers that the test established in previous section 6.7.3 is complied with in relation to document I1 so that I1, as a means of disclosure, was made available to the public before the filing date of the patent in suit. Since the information derivable from I1 could be directly and unambiguously accessed by merely reading I1 (i.e. the information was not encrypted or otherwise rendered inaccessible), it was also made available to the public.

Since document I1 discloses all the features of claim 1 in combination, the subject-matter of claim 1 lacks novelty.

The main request can therefore not be allowed.

To download the whole decision or have a look at the file wrapper, click here.

NB: A nice and concise (French) summary can be found here.

3 comments:

pat-agoni-a said...

I suppose that in a not so far future we will see this question presented to the EBA.
The main question is whether the test on the availability of the disclosure should be limited only to the Internet or is it also applicable eg to a disordered public library in a small town in a civil war region?

Anonymous said...

pat-agoni-a, are you refering to a test on the availability to the killed person?

Myshkin said...

For me the main question to be presented to the EBA would be whether the test formulated in this decision is correct. I have some doubts. Misplaced documents in a disordered public library in some obscure town in my view certainly are publicly available, even though they will not be found by anyone looking for the type of information they contain.

The "civil war" condition added by pat-agonia makes it a bit more tricky. If you have to pay a million dollars to see it, that does not take away the public availability of the document. Is it different if you have to risk your life? For the moment I would say that even civil war conditions are not relevant.

The skilled person and his access to the prior art are legal constructs. Novelty is not about whether a "real" skilled person would or could have found a disclosure at the priority date. Novelty is about the question whether it was publicly done or shown before. What does not count is for e.g. a prior use or other type of disclosure that was intended to be hidden from the outside world and effectively stayed hidden. (I have to study G 1/92 again to see if it really excludes taking into account this element of intent. Without rereading it my impression is rather that it excludes taking into account any subjective element relating to the fictional person from the public that has access to the document.)