Divergence between the UPC and EPO on claim interpretation and description definitions (Agfa v Gucci, UPC_CFI_278/2023)
The recent decision of the UPC Hamburg Local Division in Agfa v Gucci (UPC_CFI_278/2023) is another addition to the case law on claim interpretation and the impact of description definitions. The Local Division found Agfa’s patent for decorating leather valid, yet simultaneously ruled that Gucci’s products did not infringe. The entire case pivoted on the interpretation of the word "achromatic" in view of conflicting description definitions. In contrast to the recent Board of Appeal interpretations of G1/24 finding that description definitions do not override clear claim language (IPKat), the UPC Local Division found that a narrow description definition should be used to interpret the claims.
Legal background: The patent as its own lexicon
The UPC has been steadily building its jurisprudence on claim interpretation under Article 69 EPC and its Protocol. NanoString v 10x Genomics (UPC_CoA_335/2023) established that whilst the patent claim is the "decisive basis" for determining the scope of protection, the "description and the drawings must always be used as explanatory aids" (IPKat). This principle, echoed in subsequent decisions such as DexCom v Abbott (UPC_CFI_230/2023), established that claim interpretation does not depend solely on the strict, literal meaning of the words used (IPKat). Notably, the order of the Enlarged Board of Appeal in G1/24 on claim interpretation used similar wording to that of the UPC, requiring that the description is always "consulted" for claim interpretation (IPKat).
A fashionable dispute
 |
| Definitely not achromatic |
Agfa alleged that several products from Gucci’s "Pikarar Collection", a collaboration with illustrator Angela Nguyen, infringed its patent, including the Padlock Gucci animal print mini bag and Rhyton Sneaker with animal print. In response, Gucci filed a counterclaim for the revocation of the patent in its entirety.
Black and white definitions
At the heart of the dispute was the meaning of the feature "achromatic colour different from black" in the claims. As normally understood, achromatic colours are colours that lacks hue and saturation having no discernible tint of any colour of the rainbow, i.e. black, white, and all the shades of grey in between.
Agfa argued for a broad interpretation of the term achromatic colour. Agfa specifically argued that the term should not be limited to "perfect" achromatic colours but should also encompass base coats with small amounts of chromatic pigments, that may nonetheless appear achromatic to the naked eye including off-white or ivory white. To support this broader interpretation, Agfa pointed to two sections of the patent's description, including a paragraph explicitly describing an embodiment where the base coat contained both achromatic and chromatic pigments to create an "off-white or a pale clay colour", and an example describing a base coat that was "pale yellow".
By contrast, Gucci argued for a stricter and technical interpretation of achromatic based on an explicit definition provided elsewhere in the patent. The description specifically defined achromatic colour as a colour wherein "all wavelengths are present in approximately equal amounts".
Leaning heavily on the principle that a patent may be used as its "own lexicon", the Local Division sided with Gucci. The Local Division in particular considered the explicit and narrowing description definition cited by Gucci as being highly influential. Interestingly, the Local Division also found that other parts of the description, which Agfa had relied upon to argue for a broader meaning, could no longer be considered part of the invention. An embodiment describing an "off-white or pale clay colour" in the description was in particular deemed to have fallen out of the granted claim given that the term "chromatic colour" had been deleted from the claim during prosecution to distinguish the invention from the yellow and blue colours used in the prior art.
The Local Division noted that its interpretation thus created an inconsistency between the description and the claims based on this broader definition. However, for the Local Division, the clear narrower definition provided elsewhere in the description had to prevail for the purposes of claim interpretation. Notably, Headnote 3 to the decision in Agfa v Gucci emphasised that "[s]pecifications in the description that are not consistent with the granted claims cannot serve as a basis of a broad interpretation of a claim". In other words, for the Local Division, description definitions may narrow but not broaden the meaning of a term used in the claims.
The Local Division thus concluded that an ivory base coat, despite being "white", was not within the scope of the patent limited to a narrow definition of achromatic.
Infringement falls out of fashion
With the claim construction settled, the infringement case unravelled for Agfa. The Court found that the attacked Gucci products did not infringe because their base coats were not achromatic. The spectral response data for Gucci's Pikarar bags, provided by both parties, showed a "clear emphasis in the yellow-red spectrum". The reflectance values ranged from below 60% in the blue area to over 80% in the orange/red area. For the Local Division, a difference of "20 percentage points cannot be considered approximately equal in the meaning of the patent". The Court dismissed Agfa's arguments, concluding that the base coat on Gucci's products was ivory, a chromatic colour, and therefore did not feature the claimed "achromatic colour". The infringement action was consequently dismissed.
However, the very reasoning that defeated Agfa's infringement claim also saved its patent from Gucci's revocation counterclaim. Gucci’s invalidity attack was based, inter alia, on public prior use of its own "Flora" products, which it argued disclosed all the features of the patent claims. The Local Division was convinced that the Flora products were publicly available before the priority date. However, the invalidity attack failed on the same point of claim construction. The Local Division found that the Flora products, just like the Pikarar collection, comprised a "chromatic, ivory-coloured base coat". Since the base coat was not achromatic, the Flora products could not be novelty-destroying. The written prior art cited by Gucci also failed to disclose the specific combination of features required by the claims. The patent was therefore found to be valid and the counterclaim for revocation was dismissed.
Final thoughts
When G1/24 was first issued it was hailed by many as an alignment of the UPC and EPO approaches to claim interpretation. However, in Agfa v Gucci, the Local Division explicitly followed the principle of claim interpretation whereby a patent may be used as its "own lexicon", i.e. referencing the US approach to claim interpretation whereby the patentee may act as their own lexicographer (page 18, paragraph 2; Headnote 2). By contrast, Boards of Appeal applying G1/24 have found that, whilst the description should always be consulted, this does not mean that narrowing description definitions should be imported into the claims (T 1561/23 (IPKat) and T 1999/23 (IPKat)). The Board of Appeal in T 1999/23 found, for example, that "a restrictive definition of a term in the description may not be used to restrict the subject-matter of the claim, which would otherwise appear clearly broader to the skilled person"
We therefore currently have a divergence between the EPO and UPC approaches to description definitions and claim interpretation. It will be interesting to see if the other UPC Local Divisions and Court of Appeal will or will not now follow the Boards of Appeal.
Of course, claim interpretation remains a hot topic at the EPO now that the question of adapting the description to the claims has finally been referred to the Enlarged Board of Appeal in G1/25 (IPKat). The decision of the UPC Local Division Agfa v Gucci is also an interesting addition to this debate. The decision particularly arguably adds weight to the argument that description amendments involving removal of language in the description that is broader than the claims are unnecessary. In this case, the Local Division were able to come up with a clear view on the interpretation of the claims despite inconsistencies between the claims and different parts of the description. For the Local Division, a broad definition in the description that was inconsistent with the claims could simply be ignored as clearly incorrect. On the other hand, the decision did also explicitly state that the offending language "should have been deleted" and has therefore been cited as supporting the need for description amendments (page 23).
Whatever your view, given the emerging divergence in attitudes of the EPO and UPC to narrowing description definitions, applicants may now be even more loath to amend the description pre-grant in view of the potential impact of these amendments on claim interpretation before the UPC.
Further reading
- First substantive decision of the UPC Court of Appeal overturns preliminary injunction in NanoString v 10x Genomics (UPC_CoA_335/2023) (21 Mar 2024)
- UPC takes strong stance on therapeutic antibody inventions (Sanofi v Amgen, UPC_CFI_1/2023) (18 Jul 2024)
- Board of Appeal back-pedals on referral in view of "unequivocal" lack of legal basis for the description amendment requirement (T 56/21) (18 Oct 2024)
- The description of a patent should "always" be used to interpret the claims (DexCom vs. Abbott, UPC_CFI_230/2023) (21 Feb 2025)
- BREAKING: Referral on description amendments finally confirmed! (G1/25 - "Hydroponics") (29 Jul 2025)
- First interpretation of G 1/24 (the description must always be consulted) (T 1561/23) (Jul 2024)
- G 1/24 in the spotlight: Description definitions do not override clear claim language (T 1999/23) (Jul 2025)
Reviewed by Dr Rose Hughes
on
Monday, August 11, 2025
Rating:
Follow the IPKat on LinkedIn 
Posts
I commented this decision in my blog.
ReplyDeleteAgfa based its infringement action on the fact that the granted description still contained a reference to a “chromatic base coat”. However, the claims had been limited to an “achromatic base coat”, as the search had revealed a novelty destroying document for the original independent claim covering both chromatic and achromatic base coats.
When looking at the EP file, following the limitation to “achromatic”, the ED proceeded directly to grant. In other words, the description had not been consulted and amended so as to exclude from the scope of protection a “chromatic base coat”. There was a clear inconsistency in the meaning of Guidelines F-IV, 4.3.
I am not convinced that in the present case, the patent was its own dictionary. For a start the word dictionary is not to be found in the decision of the UPC-LD Hamburg.
If the patent had been its own dictionary, according to the established practice on this topic, the extension of scope to a chromatic base would have been read into the granted claim. This is exactly what the UPC-LD Hamburg refused to do.
In Headnote 3, the UPC-LD said expressis verbis: Specifications in the description that are not consistent with the granted claims cannot serve as a basis of a broad interpretation of a claim.
I cannot see any divergence between the UPC and the EPO. If the ED had done its job properly, there would not have been any inconsistency between claims and description. The Agfa/Gucci case shows best how necessary it is to align claims and description before grant.
The outcome is exactly as it should be, IMHO. Most commentators seem to agree.
DeleteAgfa was frivolous/overly optimistic in asserting infringement, and was corrected. The system worked, where's the problem?
You wrote: If the patent had been its own dictionary, according to the established practice on this topic, the extension of scope to a chromatic base would have been read into the granted claim.
I do not agree. To be its own dictionary, the description would have needed contain an explicit definition of achromatic (well, it did, but not one which Agfa liked). The mere fact that an indirect argument can be set up (an example contains a coloured base, and the claims cover also this example, so achromatic in the claims must include 'colour') is insufficient to be considered a dictionary definition. It is much more likely to be an error, as the LD correctly observed.
Examples are just that. Examples. Reports of actual experiences. They may have been included by mistake to begin with, they may turn out to switch sides during prosecution. Who cares, they were never there to define the invention, only to comply with Art 83.
Art 83, you say, and I agree. But what about Art 84? The claim has to be supported and Examples help with that too, no? Which brings us to the current EPO posture that the only description which can be said to "support" the narrowed claim is one which is exhaustively "conformed" to the narrowed claim.
DeleteIn that vein, I like your expression "change sides". I'm not a chemist, but does it not happen all the time that during EPO prosecution a claim supported by a large number of Examples is narrowed so that Group A of the Examples still lie inside the literal wording of the amended claim but others (Group B) don't any longer? Now suppose that all of Examples A and B evidence the solution of a particular technical problem, and achieve the technical effect, thereby supporting the patentability of the invention. After the claim is narrowed, at the EPO the Applicant must now delete Group B entirely, or else edit the description to announce that all the B Examples are NOT the solution to the problem, not the contribution to the art, no longer the invention which is the subject of the patent application. If a court has to grapple with Art 69, EPC, I'm not at all sure that all that well-meant conform-ising is really going to help it. Are you?
I don't really get your point at the end.
DeleteI am, exceptions aside, firmly in the camp of not amending the description to conform to the claims. The claims are what they are and define the scope of protection. To be interpreted by consulting the description of course. A convincing, positively stated reason should be present in the description story before A69 can be used to deviate from the literal meaning of the claims. (eg a definition in the description). The simple presence of an example outside the scope of the claims is insufficient for triggering A69. There is always inert ballast in the description, such an example is merely a case of such inert ballast.
If the description really contradicts the claims, of course this contradiction should be removed. But merely a potential contradiction in some far-fetched scenario is insufficient to overcome the primacy of the claims, and therefore does not need fixing.
OK, BJ, you don't "get" my last point. Doesn't matter. Truth is, I was writing with readers other than you in mind. I hope they do "get" my last point. Never mind though. Thanks for the comment. It gives me an opportunity to indulge myself yet again.
DeletePerhaps if you read my comment down this thread, posted yesterday at 09:19, about the famous (notorious?) disodium claim and its eventual infringement by equivalent, that might help.
Meanwhile, there is something in your comment that I don't get. You announce yourself to be "firmly" in the camp of NOT amending the description (let us call these happy campers the non-conformists), but then say that "of course" any "contradiction should be removed". That puts you equally firmly into the church of the strict conformists. Perhaps you have an understanding of the word "contradiction" that differs from mine (and that of the EPO)?
The EPO (eg G1/25) speaks of inconsistencies, not of contradictions. Contradictions arise when the description and the claims are irreconcilable, eg when the description says that there should be four or more of component A, and the claims state that there should be at most three of component A.
DeleteThe issue is: the claims should be supported, and as long as they are, it is OK. If there are also non-supporting sections in the description, that is no problem (IMHO), they do not help support the claims, but neither do they diminish the support that there is (as long as they don’t actively contradict the claims).
Where I differ from the EPO, and some commentators, is, I suppose, that the EPO interprets non-supporting sections of the description as counter-supporting, or undermining support, whereas for me they are just inert ballast. A non-supporting section in the description, therefore, does not create an inconsistency, and doesn’t need fixing. (and even if it did create a minor inconsistency in a certain far-fetched scenario, the courts are more than capable of handling that).
I am not even sure that sentences like ‘the invention concerns….’ in the description have any particular meaning. A52 makes it amply clear that according to the EPC there are two groups of inventions: Those that are novel and inventive, and those that are not. From that follows that ‘ the invention concerns…’ does not need to be interpreted as relating to the claims under examination or the claims as granted. Those who do, IMHO, attach too much significance to the description.
There is a point to be made that ‘the invention’ as used by the applicant is merely what the applicant, at the time of filing, considered the invention to be. Whether the entire ‘invention’, or only part of it, ends up in the granted claims, is quite a different matter and determined during examination. There is nothing against ‘the invention’ being defined much broader than the claims. (well, at least legally. It doesn’t help overall understandability of the granted patent for the lay person, and that is unfortunate)
Now I know that changing ‘the invention concerns ....’ to be in line with the claims is a small effort, and usually with no serious consequences, and I am not strongly opposed to the practice IRL. I am, however, strongly opposed to the EPO insisting on it.
I know of the pemetrexed case, and I consider it a judicial error (relatively isolated). Very unfortunate, but not sufficient for EPO to bend over backwards (and arguably deviate from the authority given to it by the EPC) to avoid that in the future judges make similar mistakes. IF action needs to be taken to avoid such errors in future, these should be actions by the legislator, not by the executive organisations like the EPO.
For you, BJ, the term "invention" in the description is what the Applicant thought it was, on the filing date. But for others, it is that which the claim for the time being declares (I refrain from using the word "define" out of respect for the position taken by M. Thomas). So when the claim is narrowed, whereby only one of the four original Examples of the "invention" is still within the declaration, there is a problem under Art 84 EPC for the EPO but not for you, not for me, and not for any others of the world's Patent Offices.
DeleteHow to solve the EPO's peculiar and particular problem? Is the problem with an inconsistency or a contradiction? I don't know. You tell me. I think the choice of word makes no difference to the problem or its solution.
The easiest way to solve the problem is for the EBA to declare that "invention" means what you say (and what everybody else understands it to mean). At a stroke, the number of contradictions and inconsistences in the description would fall precipitously. Meanwhile, apart from all that, if (for example) the claim declares "sodium" but the description defines "sodium" to mean "halogen", the EPO has to call out the Applicant under Art 84 EPC. Like it always has done. Like any other Patent Office would do, when examining for claim clarity, and support by the description.
What I like about the pemetrexed case is the thought experiment: what if Applicant, at the outset, had included a claim at the level of generality (pemetrexed and its salts, I suppose) corresponding to the real-like contribution to the art. Art 69 EPC is supposed to confer a scope of monopoly proportionate to the contribution to the art. You say the courts got it wrong. I am not convinced. Opinion within interested circles seems to me to be divided, about 50:50. Great case therefore, for debate about the pro's and con's of "conforming" the description.
I can give in on the issue of ‘the invention concerns…’ I do feel that a case can be made that ‘the invention concerns…’ means nothing in terms of scope of protection, but I know this is a lost battle. I should not have brought it up, it clouds the issue, let’s forget about it.
DeleteThe central issue is rather that (IMHO) sections of the description which do not support claim, are not automatically inconsistent (thereby counter-supporting). Clear and positive indications are needed before non-supporting sections should be considered inconsistent, instead of merely harmless page stuffing.
'Clauses' are such a thing. Unless the text clearly indicates that they define the invention, they are harmless, just a list of non-claimed embodiments.
If, as you write, there is no problem for you, me or most of the world’s patent offices, isn’t that an indication that it is a self-imposed problem of the EPO? The easiest solution for self imposed problems is for EPO to accept that it was wrong, and get on with its core business.
Another easy solution is simply to accept the primacy of the claims and accept that the description is there to to disclose preferred embodiments, to explain the technical effect(s) vs the prior art, and to comply with the applicant’s side of the patent bargain: give sufficient disclosure to ensure that patent protection is justified.
Let’s follow your thought experiment on the pemetrxed case. If the applicant had done as you say, the wider claim would have been examined, not only for N/IS, but for all requirements of the EPC. Who knows what would have been the result. If granted, then fine, there is a broader scope.
DeleteI do not agree that Art 69 is supposed to confer a scope of monopoly proportionate to the contribution to the art. Such a scope of monopoly is the MAXIMUM scope the applicant may get as a result of asking by means of claims and having these claims examined. The applicant is well within his rights to claim less than the maximum scope, either intentionally or not. The protocol refers to ‘a fair protection’ not ‘proportionate to the contribution to the art’.
How are third parties to know whether the applicant exercised his right to only claim a narrow scope, or was stupid, or was careless, or was merely ignorant of the alternatives because it they are in a different technical field?
Considering that monopolies from patents take away freedom from everyone else, the claimed scope should be narrowly interpreted, and any attempt to broaden protection by invoking art 69 should be really convincing and substantiated. A high standard of being precise in what he wants from society to consider his own for 20 years may be expected from the applicant.
In particular the significant threshold that the applicant did not claim something (presumably intentionally) when he could and should have, needs to be overcome. In case of doubt… sorry, you took a risk and it backfired. A69 is not intended to be a safeguard against sloppy drafting or later changes of mind. Let’s not forget the first sentence of A69.
In the pemetrexed case this is even more applicable. Look at it this way: if the other salts are such clear equivalents that they should be considered to be covered by the claims, were they not equally clear equivalents to the applicant-skilled person when drafting the claims? Apparently they were not clear equivalents to the applicant (otherwise he would have included them), so how can the applicant honestly claim that they are equivalents for others? Is the applicant perhaps to be held to a lower standard of what may be considered equivalent than the rest of the world?
Why should society grant a wider monopoly than asked for? In my view A69 is a safety net in case the applicant could, at the time of filing, not reasonably be expected to have foreseen the alternative (eg because the alternative did not yet exist, or because it is in a remote technical field), whereby more leeway can be allowed the further away the ‘equivalent’ is from the central teaching of the invention. For instance, I am happy to allow equivalents for nails when somebody has a chemical invention and states that reactors A and B are NAILED to each other, but not when the connection method of the reactors is the actual invention.
@ BJ,
DeleteFor a start G 1/25 has not yet been decided. I can agree with you that the referring decision mentions inconsistencies 49 times, and contradiction only 6 times.
Further, you are really playing on words when you make a distinction between inconsistencies and contradictions. If the description contradicts the independent claim, there is necessarily an inconsistency between description and claim, as otherwise there would be no contradiction.
In T 2685/19, quoted in the referring decision, it was held that the description submitted for AR3 was in contradiction with the claims and therefore did not support the claims. If the claims are not supported, there is necessarily an inconstancy between claims and description.
You also consider that “There is nothing against ‘the invention’ being defined much broader than the claims”. If the “invention” defined in the description is broader than the claims, there is necessarily an inconsistency between claims and description. This is exactly what has been queried in the Agfa/Gucci case.
Your argumentation that the “‘ the invention concerns…’ does not need to be interpreted as relating to the claims under examination or the claims as granted” is flabbergasting. It is also flabbergasting when you claim that too much significance is attached to the description.
As the description is there to support the claims, cf. Art 84, second sentence, and, according to G 1/24, should be consulted systematically, and not only when the claim is unclear, you are actually negating the headnote of G 1/24.
A non-supporting section in the description, therefore, DOES create an inconsistency, and DOES need fixing. You might feel free to refuse adapting the description, but you should then not be surprised when your application will be refused or your patent be revoked. It is then for you to explain to your client why you did not get a patent or you lost it.
As far as the pemetrexed decision of the UKSC, please look at my comments to Max Drei in the blog: “No change: The EPO's guidance to Examiners on how to apply G1/24” of August 1st. The aim of the decision was to introduce in English case law a new definition of infringement by equivalence. In this respect the decision is not wrong.
With Art 84 as it stands, no actions are needed, neither by the legislator nor by the executive organisations like the EPO. However, the EPO has to apply the EPC, and this is what it does. Not to your liking, but this is something different.
“ I am not convinced that in the present case, the patent was its own dictionary. For a start the word dictionary is not to be found in the decision of the UPC-LD Hamburg.” You say dictionary, I say lexicon?
ReplyDeleteIt is remarkable how far certain people go to assert that the law works in a way which fits their preconceived notions. One only needs to read the headnotes of the decision to see that the board indeed acknowledged that a patent can be its own dictionary.
DeleteG1/24 stated that the Enlarged Board finds it a most unattractive proposition that the EPO deliberately adopt a contrary practice to that of the tribunals that are downstream of its patents. The boards seem intent on ignoring this clear direction in G1/24, which is disappointing. There was an opportunity to bring sense and consistency to EPO oppositions and litigation at the UPC. It is a predictable response from the EPO, however, which has never been interested in sense or consistency.
@ Anonymous,
DeleteAs as it is "remarkakable how far certain people go to assert that the law works in a way which fits their preconceived notions", the same applies to Anonymous exactly in the opposite direction. In a comment on my blog apparently the same anonymous even found the board disingenous has it has blatantly ignored G 1/24.
The head note of G 1/24 does not say that the patent is its own dictionary. This is an interpretation of Anonymous, i.e. an oponion, as respectable as any opposition might be, but not the way Anonymous would like to see the law to work in a way which fits its preconceived notions.
G 1/24 requires the description to be consulted, but does neither means nor implies that the what is stated in the description should overule what the claims states directly and unanbiguously.
In the Agfa/Gucci case, the UPC did exactly what is required in G 1/24, i.e. to consult the description. When consulting the description the UPC noted an inconsistency between the clear wording in the claim "achomatic", and what is stated in the description, i.e. the base coat can be "chromatic". The UPC did not accept that the scope of the claim can be extended to chromatic bases.
In their recent decisions, the boards have not deliberately adopted a contrary practice to that of tribunals that are downstream of the patents granted by the EPO. They have consulted the description, and when they saw an inconsistency between claims and description, they have not overruled the definition in the claim with what they found in the description.
That the interpretation of the EPC by a board or by the EPO should preferably not be at odds with that of post grant tribunals is desirable, but not a necessity. Just have a look at how the German Federal Court has dealt with added matter. The latter is more lenient with a patentee than the EPO, but the EPO has maintained its stance. All attemps to bring in the solutions proposed by the GFC have never been successful.
The EPO cannot enforce its interpretation of the EPC to post grant tribunals, and post grant tribunals cannot impose their interpretation of the EPC to the EPO.
G 1/24 fits in this as is acknowleges that the description is to be consulted, but does not clim that the description overrules clear claimed features.
As to conforming the description, I am reminded of that claim directed to disodium pemetrexed, the infringement by equivalent and the debate whether the outcome would have been different in that case, with a finding of no infringement, had the ED at the EPO insisted on conforming the description before the claim issued. It looks to me as if we are heading in the direction of "caveat emptor". The patent buyer at the EPO who chooses to go to grant with an unconformed description does so at their own risk. You know, like with amendments to claim 1 and the Art 123(2)/(3) trap.
ReplyDeleteOwn risk, you suggest. Let's take the "gathered" case. Find no inconsistency (because gathered is such a vague term), take the definition set forth in the description as the true meaning of gathered, in context, and go on to find the claim invalidated by the prior art. Basta!
ReplyDeleteAgfa v Gucci should be taken with caution, as the problem appears to stem primarily from inconsistent statements within the description itself. And I agree with previous comments that it is not a case of a patent being its own dictionary. The description does not contain authoritative definitions comparable with the set of definitional statements of EP3076804 which gave rise to referring decision T 439/22 for G 1/24.
ReplyDeleteNonetheless, going beyond the headnotes, Agfa v Gucci contains a statement of interest in relation to the adaptation of description. In reason 2 (bb), it states that several passages of the description inconsistent with the claims should have been deleted.
A more recent decision, LD Paris 363/2024 (SJ Diffusion v. Gisela Mayer) issued on 1st August, deals with the assessment of equivalents. In this case (EP2404516), which relates to a bonnet for a wig, there was no request for revocation, and no divergence between the parties as to claim interpretation.
A claim amendment at the 71(3) stage had excluded one of the disclosed embodiments (« superposition », the other one being « edge-to-edge ») from the claimed subject matter. The court held at § 60 that it does not matter if this embodiment is still disclosed in the description, what matters is the extent of protection as delineated by the claims.
However, the court added that the restriction in the claims to the « edge-to-edge » embodidment should have triggered the deletion from the description of the « superposition » embodiment no longer covered by the claims. The court’s reminder, while an obiter dictum, is highlighted in Headnote #1, which also cites the Agfa v. Gucci decision. It is however undermined by the court itself holding that the disclosure of the « superposition » embodiment in the description of the patent does not matter and only the claims matter. If so, an adaptation of the description appears irrelevant to the assessment of equivalents.
There is something highly unusual in this case : the B1 document does not appear to include the amendments to the description aimed at removing inconsistencies, listed in the 71(3) notification. Could this be the reason for the court’s remark ?
It is to be noted that the same court LD Paris, in the decision 358/2023 (HP v. Lama) issued on 13 Nov 2024, dismissed (point 41) an adaptation of the description excluding embodiments from the scope of the claims as irrelevant to claim interpretation, on the basis that such embodiments as they are shown in Fig 4-8 of the application are to be taken into account. Here, claim interpretation was the issue, unlike the context of the SJ Diffusion v. Gisela Mayer decision in which the issue was only the assessment of equivalents.
Getting back to EP3076804 mentioned above, there is a recent development of interest. The patent owner reacted to G 1/24 by filing on 3rd July an auxiliary request to delete from the description paragraph [0035] which broadens the definition of the claim term « gathered sheet ». The opponent replied that the request was inadmissible under the RPBA and also that it raised a new matter issue as an intermediate generalisation. Oral proceedings have ben scheduled on 11 December.
Whatever possessed Agfa to go out and assert this heap?
ReplyDeleteTo recap, the decision revolves around this passage of the disclosure to clarify the meaning of the claims, and set out their "metes and bounds", like the Muricans say:
[0021] A chromatic colour is any colour in which one particular wavelength or hue predominates. For example, blue and green are chromatic colours, while white, grey, and black are achromatic colours, as they have no dominant hue, meaning that all wavelengths are present in approximately equal amounts within those colours.
This passage is in my opinion nonsensical, from a physical, human perception, and colourimetry standpoints (and probably a couple more if I set my mind to it).
Take a notional 5600K incandescent source (ie., the sun), normally perceived as "white" ("achromatic" in the parlance of the disclosure) by observers, and often identified as such on a CIE 1931 chart on its Planckian locus. Its black body equivalent spectrum, or actual measured spectrum in either outer space or at the surface of the earth is anything BUT uniform in the (implied) visible ~380-700nm wavelength range. (Try entering "5600k spectrum" in your search engine of choice.)
Furthermore, three approximately monochromatic sources (e.g.: a red, a green and a blue LED, or the individual VDU pixels you are looking at right now, regardless of their underlying technology) can be readily combined to produce a white luminous sensation when perceived by the three types of vision cone cells within a given angle offset from the observer's fovea, but then, "all wavelengths are present in approximately equal amounts" is by no means true...
Perhaps the drafter had an electronics or communications science background, and had somehow muddled his thinking with notions of thermal noise aka Gaussian noise aka Johnson noise aka white noise, very many octaves away from the visible light region. However the power spectral density of such a noise is uniform in FREQUENCY, not WAVELENGTH. (The expression for Johnson noise and Planck's radiation law both involve Boltzmann's constant. I suspect that the former could be obtained from an asymptotic expansion of the latter. Mais je digresse.)
Maybe the drafter meant something else altogether... But how far should the reader have to go in assessing the logic and meaning of the text? (Note to self: find a way to plug Umberto Eco's 1979 essay "Lector in fabula. La cooperazione interpretativa nei testi narrativi" - EN: "The role of the reader". Reply to self: DONE! Counter-note to self: You'll have to develop this, eventually.)
The original disclosure goes at lengths to explain that the base coat can be pretty much any colour as long as it's "different from black" (how different is "different"?)... But then the applicant flips during prosecution and now argues that the base coat MUST be "achromatic", and that this somehow procures some unique, inventive, incredible, unforeseeable advantage. Was this selection even originally disclosed "directly and unambiguously", or is the public now presented with new information?
[end of part 1]
Had D1 (US2010233441A1) disclosed "white" instead of "yellow" (cf. D1 @ [0099]) would the applicant then shifted instead to a "chromatic" colour? Is a "technical" problem solved? Or is the background colour chosen for esthaetic considerations?
ReplyDeleteIf the base coat colour was in fact critical to the definition of the invention, why wasn't it specified in the disclosure using some generally accepted matching system, such as Pantone or RAL? Try going to the Baumarkt, and study the clerk's reaction as you demand a bucket of "achromatic" paint...
I note that glossy magazines and corporate annual reports are printed on coated papers which usually provide a pure white background on which the ink is applied. But this knowledege is absolutely totally resolutely and irremediably irrelevant if you replace paper with leather, and kaolin clay with polyurethane, lest you commit the heinous crime of hindsight and ex-post-facto analysis. The skilled person in the very rarefied and exquisite field of leather printing was fitted at birth with extremely narrow blinders...
The applicant argued as follows w.r.t. inventive step on 24.01.2019:
The technical effect conferred by this difference is that colour inconsistencies and some surface defects in the surface of the crusted leather are masked, while increasing the colour gamut for enhancing the luxury effect of leather as photographic image quality can be obtained, and also providing economic benefits in that less complex inkjet printers can be used that are printing with an inkjet ink set containing fewer inkjet inks. This effect is clearly shown by [0029].
Call me skeptical... Wasn't this alleged masking effect already achieved by D1's yellow or blue coatings? If anything, a white ("achromatic") base coating will not dissimulate as well the imperfections of the leather, as, say, a dark blue, while being more sensitive to scratches and dirt... (Note to self: make a point by comparing the few colours in which men's underwear is typically offered. Reply to self: I'd rather not.)
I understand the UPC's approach in taking the proprietor words at literal value. That was probably the least messy way to tackle this, even if this meant that the patent was ultimately maintained unamended. They probably didn't relish diving in the inventive step rabbit hole. The defendant, who had demonstrated prior use, now has ensured FTO (albeit at the expense of legal proceedings), and only needs to make the base coating off-white, eggshell, ivory or cream, or pink, mango, or fuchsia, to be in the clear. Having to forego pure achromatic backgrounds will only cost them the US white shoe lawyer or Corsican pimp markets, neither of which one would want to be associated with anyway.
[end of part 2]
[to be continued]
I am not sure that the EPO would come to a different conclusion here: you have a narrow definition and a broad embodiment in the description, so there is some ambiguity. In the end you give the term achromatic its commonly understood meaning in the claims. Same result, maybe slightly different route.
ReplyDelete