On Mon, 21 Jan, Jan Nijtmans wrote: > I'm not trying to discuss whether to allow committing something > or not, that's exactly what the *-glob settings are for.
That's what I do not understand. I do not understand the relevance of the encoding regarding whether a file should be allowed to be committed or not. If I understand your suggestion correctly, the best name would be "non-utf-8-glob" which looks kind of strange to me. And a lot of people would just set this to "*" in order to allow for iso-8859-1, iso-8859-15, Windows-1251, etc. If this however it used to determine whether the web UI can display the content, then for those configurations no file would be displayed in the web UI, as non-utf-8-glob * (or encoding-glob with the same semantics) would mean "not UTF-8, no idea to display" for all files. That's why I suggested that the encoding of a file does not have any relevance of whether it can be committed or not. The encoding of a file just has relevance when the data is being interpreted again. The user knows how to deal with the committed content, only the web UI does not. Therefore the web UI needs heuristics to guess the encoding correctly in order to display it properly. BTW, the UNIX "file" code also has heuristics for "binary" and "DOS line endings", so that you wouldn't have to explicitly set those globs either. My proposal was just to allow any file to be committed, but for display purposes, the "file" heuristics tells what encoding and what line ending a file has, or whether it is even binary. If I am completely misunderstanding what those *-glob are intended for, please explain in more detail to me. > Fossil 1.24 didn't have "unicode-glob" yet, therefore > now is a good time to re-consider the name before > 1.25 is released. Yes, IF those need to stay and IF unicode-glob is a bad name, THEN now is the time to change it to encoding-glob. I'm just not sure that this solves the issue ... unless I misunderstand "the issue". I am not trolling nor making up arguments, I just don't see that renaming unicode-glob to encoding-glob makes things any different. Greetings, Stefan -- Stefan Bellon _______________________________________________ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
