> On Wed, 2002-07-24 at 10:22, Mark Rafn wrote: > > Perhaps I misunderstood, but it sounded like it would be required for a > > modified work to identify itself as modified, so that documents can > > determine if they're running on "real" latex. This disallows preserving > > the API exactly while changing the execution.
On 24 Jul 2002, Jeff Licquia wrote: > No, no. The *kernel* would do the recognition, not the documents. It > would have a list of "acceptable" values for that macro. > > Thus, "latex" would refuse to use any modules that didn't identify > themselves as Standard LaTeX, while "debtex" would accept modules that > identified themselves as "debTeX" or Standard LaTeX. Still don't get it. You're either requiring modified work to follow a specific API, which is IMO non-free, or you don't get the desired protection against impostors, as a modified work could simply return the latex identifier. > > A different name to humans. A different package name, sure. In some > > cases, a different executable name (This would be problematic if it > > were broad enough). A different name in it's API? I don't think that > > follows. > > Why not? Why does embedding the name in a registration call offend you? For the same reason that limiting the API of any program would be non-free. I also wouldn't accept a C library that disallowed calling a modified function "printf". > > Adding the facility is no worry. Requiring derived works to use that > > facility is non-free IMO. > > Give me some reasons. "I don't like it" is really hard to argue with, > address, or even evaluate. Because freedom to change (or not change) the API is part of the freedom that I believe Debian guarantees to it's users. -- Mark Rafn [EMAIL PROTECTED] <http://www.dagon.net/> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]