>On Wed, Jun 09, 1999 at 12:07:17AM +0100, Andrew Collier wrote: >> And the >> GPL is flawed anyway - Si might assert that his code to interface to Win32 >> instead of Linux is a differentiable work, in which case the GPL doesn't >> need to cover it. > >Not at all - the rules are very strict on this point. To be able to >claim that, Si needs to distribute something that doesn't contain *any* >of SimCoupe, so that you can then get the official SimCoupe and run them >together to make the final product - and even then that's unlikely to be >allowed because the result could be considered a derivative work (that's >why the LGPL exist).
>From paragraph 2: >If identifiable sections of that work are not derived from the Program, >and can be reasonably considered independent and separate works in >themselves, then this License, and its terms, do not apply to those >sections when you distribute them as separate works. Okay, I've just noticed that they indeed do have to be distributed seperately, but that should then fit the terms. Actually, since the LGPL was changed a few months ago, the infection principle is almost as strong as in GPL (ie. loony RMS has broken the LGPL beyond repair). >> I like Open Source, but I don't like GPL. It is flawed, sometimes rather >> loosely defined, and doesn't give the original programmer enough rights SNIP! > >Whatever it is I don't think "loosely defined" could ever be applied to it. In that case, where does it define "distribution"? Does it count, for example, if a few copies of the binary have been given to beta testers but that the source has not? >Also, the original programmer has any rights he pleases because he owns the >copyright and can release a separate version of the work under a different >licence (what he can't do of course is prevent the original version from >being distributed under the GPL). Why not quote half a sentence and then complain that the first half doesn't make complete sense, eh? Better licenses give the original programmer magic extra rights which are not passed on to subsequent developers, ie that only the original programmer can produce an official distribution. But consider A, who writes a program called Foo v1.0, released GPL. A continues development, and releases Foo v1.1, released GPL. B develops from Foo v1.0 source, and releases Foo v1.1 - which is now totally incompatible with A's program. The GPL doesn't prevent this. But we now have two different, incompatible products with the same name. And of course everybody gets confused, and if B's modification stops working, then as likely as not it'll be A that gets the rude emails and support requests. >> Personally I think that (under most circumstances) the programmer should be >> allowed to distribute his own code under whatever terms he likes. > >No one forces programmers to apply one licence or another to their code. >What are you on about? Si Owen, for example, has no choice of what license is applied to SimCoupe Win32, because it is originally derived from a GPL program. The GPL is infectious - ie any derivatives MUST be distributed according to the GPL. And this is why many commercial companies, for example, won't use their skills to enhance GPL software - because it would necessarily involve losing the benefit of too much sensitive IP. >> The >> sooner the world stops using the GPL and replaces it with a more sane >> scheme, the better IMHO. > >Careful. Xz80 is distributed under the GPL, you know. And? Andrew PS. This is getting *way* off topic. Can we do something more appropriate with it? It's been a while since anyone posted to oxbridge.tat; perhaps... -- | Andrew Collier | email [EMAIL PROTECTED] | Talk sense to a | Part 2 NatSci | http://carou.sel.cam.ac.uk/ | fool and he +----------------+----------------ICQ:38645805-+ calls you foolish | Selwyn College Student Computer Support Team | -- Euripides
