On Sat, 16 Aug 2008, Jan Engelhardt wrote: > On Friday 2008-08-15 18:03, P˙˙teris K˙˙avi˙˙š wrote: > > > This is a request for comments on the proposal that the following > > filename changes (indicative at this point) be made in the > > netfilter source code tree, together with correcting the > > corresponding references to the files in both source code, the > > build system, and wherever else necessary. > > > > The changes are proposed in order to avoid file name clashes on > > filesystems that do not allow two files to have names that differ > > only in the case of their letters: > > > > include/linux/netfilter/xt_CONNMARK.h => xt_connmark_target.h > [...] > > As I see it, this will already be addressed by merging complementary > code into one source file (e.g. MARK, mark, CONNMARK, connmark > all into a big new xt_mark.c), to reduce (a) amount of files > and (b) build time. [...] That is half of the story: we encode the functionality (match/target) in the case of the (file)name and there are targets which have got no match counterpart and thus no natural lowercase filename to merge into it. So the directory tree still will be broken on case-insensitive filesystems. If as a side effect of some real internal changes netfilter becomes compatible with case-insensitive filesystems, that's fine. However, I object changes just for the sake to be compatible with such horrid filesystems. Best regards, Jozsef the NACK guy of the week - E-mail : kadlec@xxxxxxxxxxxxxxxxx, kadlec@xxxxxxxxxxxx PGP key : http://www.kfki.hu/~kadlec/pgp_public_key.txt Address : KFKI Research Institute for Particle and Nuclear Physics H-1525 Budapest 114, POB. 49, Hungary
