Hi Enrigue, I'm not a lawyer either; you may know as much about IP law as I do, but I have to tell you that I wasn't commenting so generally as you are. Enrique Perez-Terron wrote: > Hello all, > > As far as I understood this issue, it started out with a request to be > able to list the font names that an application can "ask for" without > being turned down. (Yes, I know fontconfig does not turn programs down, > but considering the bad match it would be to use the last-resort default > font for a requested one that only contains antique Egyptian hieroglyphs > in some non-standard encoding, this is as good as turning one down.) > Sounds like a reasonable request to me; I wasn't commenting on it. > I don't quite see how it would infringe trademarks if this piece of > factual information is made available. I believe not all cases of naming > a trademark constitute an infringement. I am no lawyer and I would > appreciate any clarification from any competent person, but until so > happens I think infringement happens primarily when another product is > offered to the public under the protected name. There are a number of > twists to this issue, like using a protected name in ways that > subconsciously transfers the positive conotations associated with the > protected name to the product being offered, or to the offerer. I doubt > "someone might misunderstand the intent" is a sufficient reason for > considering a quote of a name an infringement. > I do see how it could infringe. And specifically, if you are "naming" something, and the "name" you use or provide is a trademark for which you do not have appropriate permission, I believe you could be infringing on the trademark, even if the thing you are naming has nothing otherwise to do with the association the trademark intends to convey. If you're just recounting names you're not providing, I think what you're saying makes some sense. The difficulty I am imagining is with whether or not you're just recounting names, or assigning some as well. When you're dealing with built-in aliases, I think a line is being crossed. When you begin to talk about products, though, you must then also deal with copyright issues (and possibly patent), not just trademark issues (and maybe not trademark issues at all). > If I receive a document in some foreign format, that specifies it was > prepared with a particular Adobe font in mind, but fails to characterize > this font in any way, I find it far-fetched to consider the specified > font's name a "generic description of style". The substitution of a > suitable available font for the named one will have to rely on > information about that font (i.e., a description of it) residing > somewhere else than in the document, in casu, in one of the > configuration files of fontconfig. The information in this configuration > file would be saying something like "Font X(possibly a tm) is > reasonably, or even best, approximated with font Y(possibly a tm), of > all fonts available on this system". > I can't tell if you're agreeing or disagreeing with me. I wasn't suggesting the general case of using style instead of family names; I was thinking of the built-in aliases. Here's my more basic point: names aren't generally descriptive. So to do semantically meaningful matching, you'd better have semantically meaning descriptions available. Build them yourself if you have to, but if you are just matching on name (names are effectively symbolic and thus not guaranteed to convey semantic information), you are not doing semantic matching. E.g., if I just tell you my name is John, you don't know much about me. If I describe myself, however, you know not only something about me, but have a better idea about how even to assign a meaningful alias to me, if you need or want to do that. > The question the OP wants to ask fontconfig could be rephrased "about > which fonts is sufficient information available on this system so that a > suitable substitution can be chosen?". Again, I believe providing an > answer to this question is a right under free speech, not a trademark > infringement. I further believe this and similar information could > legally be given in a printed (and publicly sold) book, hardcoded into a > program, etc. > "Available on this system" implies "legally permissible to provide", not just where trademarks are concerned, but also where copyrights and patents are concerned. They are different kinds of intellectual property with different kinds of protection afforded under law, but trademark protection specifically is indeed a kind of limitation of free speech, if you want to see it that way. Speaking a bit more generally, the notion of free speech in the US flows from the 1st Amendment, i.e., it was not an original clause of the Constitution. Intellectual property protections, however, flow from Article 1 Section 8 Clause 8 (I believe; this is off the top of my head, so you should check the reference), which is in the original body of the document. I am not, by the way, agreeing with how this clause has been interpreted (it mandates a responsbility to Congress, but serves only to guide them in that mandate); I am only observing that it is not trumped by the 1st Amendment's free speech clause, which, by the way, the courts have ruled is a limited right. > Surely, trademark issues are important, but it would be unfortunate to > limit the scope of free software because of a poor understanding of the > legal issues. > I wholeheartedly agree how important it is for both developers and users alike, and especially users, of open software, to understand at least the basics of intellectual property law. No question. But that means understanding that open software is still, generally, licensed software, and that its scope is limited, if only to a small degree, by that fact. Of the body of software that might make up a Linux system, for example, I'd guess that well less than 10% is "free" in the sense of being in the public domain (PD) and not subject to copyright and offered under the terms of a license. Trademarks, copyrights, and patents are all owned property under law, and the owners have established rights, even if they exercise those rights by choosing to allow others to use that property without compensation. The details and implications of this point are far beyond the scope of this list, however, as interesting a discussion it might be. > Regards, > Enrique > > On Wed, 2004-03-31 at 19:57, John A. Boyd Jr. wrote: > >>I meant "limited to", not "imagined to" (where did that come from?...) >> >>Might I also add, however, that terms used to describe are much less >>likely to be interpreted in a legal sense as trademarks, compared to >>terms which name. Indeed, when names begin to become descriptive, >>like "Kleenex" and "Xerox", it becomes all the more difficult to >>enforce them as trademarks. >> >>-John >>
