> Date: Thu, 14 Feb 2002 23:22:25 -0800 > From: Per Bothner <[EMAIL PROTECTED]> > > Suppose you have nodes named 'Foo~' and 'Foo_' that both map to > Foo.html. Assume that both nodes are actually in very different > parts of the manual, and don't belong together. I..e. there are > other nodes between 'Foo~' and 'Foo_', but they get mapped to other > files. If you use anchors, yes, you can get the navigation to work. > But really, why bother?
We bother about that so that the HTML document will be readable. > Supposed you have a node 'foo~bar', which currently gets mangled > to a filename 'foo-bar' and anchor 'foo%7ebar' (or whatever). > Instead of making the reference be 'foo-bar.html#foo%7ebar', I'm > suggesting having an option where the filename is 'foo%7ebar.html' > and leaving out the anchor, since it is now fully redundant? That would produce files with funny characters, which might cause difficulties on some systems. In effect, you are saying to stop mangle node names when producing file names. AFAIK, it is generally considered a bad idea to have non-Posix characters in file names. HTML anchor names allow more characters, so we put the node name there almost intact. _______________________________________________ Bug-texinfo mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-texinfo
