On 4 February 2014 18:11, Jean-Pierre Flori <jpfl...@gmail.com> wrote:
> > > On Tuesday, February 4, 2014 6:05:58 PM UTC+1, Jean-Pierre Flori wrote: >> >> >> >> On Tuesday, February 4, 2014 5:56:29 PM UTC+1, Bill Hart wrote: >>> >>> Hi all, >>> >>> from today until 28 Feb (incl.) I will be working exclusively on MPIR >>> (with a couple of very small interruptions). >>> >>> Today I did the following: >>> >>> * cleaned up .gitignore to make git ignore all autogenerated files >>> >>> Great. I've been working on that lately. >> There are some commits on my github branch. >> I'll just rework on top of your updated branch. >> >> Note that from my point of view, nothing coming automatically produced by >> autoreconf should be tracked in the git repo. >> >> * did autoreconf -i so that autoconf/automake work on my development >>> machine (and replaced our original hacked versions of config.guess and >>> config.sub). It looks like the fsf versions of those files were last >>> updated in 2013, but I've made a ticket about perhaps updating them again. >>> >> Yup, I update the fsf files to make them work on Cygwin64. >> Before that they dated from 2003 or 2006... >> > Groumpf, that also updated the yasm subdirectory... > I tried to make it ignore some of the yasm files. I don't mind sticking with vanilla yasm if you prefer. But at least from my perspective git was complaining every time I tried to do anything. So something had to change. Was yasm updated to a later version recently or something? We never used to have these problems. > Not sure if you really want that, though it does not really hurt. > (Last time I ran autoreconf I used the --no-recursive option, so yasm > remained the vanilla version.) > Yeah, I had problems before running autoreconf. But you are right. I probably didn't want to do this. We can revert those yasm files changed by my autoreconf patch if you like. I will *try* to produce commutative patches from now on! > > I'm not totally convinced with the way yasm is shipped right now. > Maybe just shipping a vanilla tarball would be enough. > Me either. But it's so much work to fix. As soon as you let people use the system yasm, everything breaks and we get piles of bug reports. We could of course check the precise version of yasm and have the build system complain otherwise. I'll add a ticket for this. > > > IIRC last time I played with the git version there were also issues with > line endings. > Me too. I don't know where they have gone, but they have gone. I think my gitignore patch fixed some of them. The rest I probably just committed the changes it wanted. I intended to fix this. I don't know if I really have. > I think Brian reintroduced some CR/LF or things like that. > See this commit > https://github.com/jpflori/mpir/commit/0171152c26d69c521305bcb92b646ae74d7bf0ea > Maybe its worth to think about the way line endings are dealt with right > now. > We just need to change git to use eol-native for the affected files. Or am I thinking of svn, I forget. Bill. -- You received this message because you are subscribed to the Google Groups "mpir-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to mpir-devel+unsubscr...@googlegroups.com. To post to this group, send email to mpir-devel@googlegroups.com. Visit this group at http://groups.google.com/group/mpir-devel. For more options, visit https://groups.google.com/groups/opt_out.
