At 04.42 +0200 2002-07-31, Thomas Bushnell, BSG wrote: >Lars Hellström <[EMAIL PROTECTED]> writes: > >> The problem with GPL'ing is that anyone who recieves a PS file using a >> GPL'ed font could then claim that the PS file in its entirety must be >> GPL'ed and thus request to get the (.tex or similar) sources for the PS >> file, since these would be "the preferred form for making modifications". > >If the font is really separate: that is, if the encoding is done in >such a way that it's easily extractable, then it clearly seems like a >case of a mere aggregation.
It odd to see such a conviction that "this is aggregation, which is harmless" here on this list, considering that it was recently claimed that a tarball (!) must be considered to be single work until proof of the contrary has been obtained, without any objections from the regulars. Can anyone think of any use other than aggregation for a tarball? But perhaps there are double standards at work ... I don't believe that interpretation of the GPL aggregation clause In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. is plausible enough to rely on in my case, but it could be interesting to examine the matter more closely. Webster's New Illustrated Dictionary defines "aggregation" as "a collection of particulars" and a "particular" as "an individual thing or point or quality". Is it then the case that anything which is a collection of things with individuality becomes an aggregation? The "it's just an aggregation" argument against the GPL would then be that as long as you can tell, for each piece of code in a program, whether it is generated from GPLed source or non-GPLed source, then the program as a whole is merely an aggregation and the condition in the GPL that a combined work must be GPLed would not apply. I doubt this argument could work. However if it did then it certainly would provide a technical solution to the (obnoxious?) GPL incompatibility problem: just design the linker so that it pads the executable with markup saying "beginning/end of material that is part of the work XXX", and then claim the file is an aggrevation of different works, which just happens to be interpreted as an executable program by the OS. Theoretical excursions aside, it seems to me that the Design Science License (as suggested by Martin Schröder), with an extra clause (as suggested by Walter Landry) that inclusion in documents is fine, will be appropriate for what I had in mind. In particular a renaming clause certainly is a *good* thing when it comes to fonts and the like. Lars Hellström
