"Juanma Barranquero" <[EMAIL PROTECTED]> writes: > On 3/21/07, David Kastrup <[EMAIL PROTECTED]> wrote: > >> Do you have an example for a Postscript file on your system that is >> neither identified by the magic string "%!PS" nor an appropriate >> extension? > > No. But I don't understand your question. I was agreeing with you > (yeah, it happens sometimes).
Basically, we have three "proposals" (partly proposed by previously existing code): a) accept "%!" as magic PostScript override (previous behavior) b) accept "%!PS" as magic override (what I proposed and checked in) c) don't accept any magic postscript override When arguing against c), it is not clear whether you agree with a) being a sufficiently bad idea. I had one example file causing problems with option a), and I already gave examples for files requiring b) (those produced using dvips -i don't have an extension, but in all cases start with "%!PS"). My point of view is that b) nowadays appears like the most pragmatically useful option, judging from problematic cases I have seen. If people can live with that, I suggest we leave it at that. While there are arguments for the other choices, we had them mentioned (Stefan's argument that magic should be kept to a minimum certainly _is_ valid) and, I believe, considered. Repeating them will just waste time. If people feel that they have been weighed wrongly, the way to resolve this is not repeating arguments, but vote on the resolution. Anybody who feels that the current solution b) as checked in by myself is not the best solution? If so, I'd want to either hear new arguments or have a vote. Everything else seems like a waste of time. -- David Kastrup _______________________________________________ emacs-pretest-bug mailing list [email protected] http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
