Steve Langasek <[EMAIL PROTECTED]> wrote: > severity 410903 important > thanks > > On Wed, Feb 14, 2007 at 12:37:42PM +0100, Frank Küster wrote: >> I think that it is clear that these files actually are configuration >> files. This has been discussed for example in #326637. Reading through >> this bug, one gets the impression that everyone who spoke up agreed that >> it makes sense, and is actually done, to customize these files (or maybe >> rather, to add customization files in these directories, see below). > > Sorry, a file is not automatically accorded config file status just because > someone wants to edit it.
I agree, and I've argued for that view myself, e.g. in #379089. The reason why I reported this as RC is that I think there are in fact configuration files among the moved files. If moving them back where they belong is too dangerous for etch, why not tag it etch-ignore? > I think that at least the .dir files ought to find their way back to > /etc/X11/xkb to let users add their own configs, but I don't think this > should be considered RC -- and that we are in fact better off not changing > this for etch, given the risk of error. (that sounds rather like you're granting etch-ignore, doesn't it?) The risk of error is a point, whether it is wise to keep a regression compared to sarge for a complete release, and only fix it in lenny is an other one. As I already said in the original bug report, adding documentation may be an alternative, or may justify and etch-ignore tag - *at*least* NEWS.Debian, but better a README.Debian which describes in detail how to work around that regression (I've heard rumors about /etc/X11/Xmodmap being supported already?). Regards, Frank -- Dr. Frank Küster Single Molecule Spectroscopy, Protein Folding @ Inst. f. Biochemie, Univ. Zürich Debian Developer (teTeX/TeXLive)
