On Sun, Feb 18, 2007 at 02:43:11PM +0100, [EMAIL PROTECTED] wrote: > > 2007/2/12, [EMAIL PROTECTED] <[EMAIL PROTECTED]>: > > > This patch adds an autogen.sh script for libffi. As this patch should > > > probably go upstream but pnet doesn't survive 'make distcheck' without > > > it, I sugest that we apply it anyway, but put it into a > > > "patches-for-upstream" (invent a better name, please) directory. > > > > As far as I know gcc tree does not include autogen.sh scripts. > > So perhaps there is no chance to push it upstream. I think that > > in such cases we should minimize the changes we do to a package > > so we will have less work to merge with their newer versions. Then > > perhaps it might make sense to to run libtoolize, etc in the libffi > > dir from within the top level auto_gen.sh? What do you think?
Yes, good idea. I'll change the patch when I do the next round of reworking. > The autogen files are never distributed because they are used by the > developers/maintainers only. So we have to put our own autogen files > the the subdirs. Well, it depends. Most projects I know distribute their autogen or bootstrap files, although you don't have to. If you do, the user has the possibility to recreate the autotool stuff the way the maintainers intended. When using later autotools, it should even work to just call "autoreconf", but as it doesn't work correctly with all available versions out there and as we gain an additional possibility to check for example for versions, a script is IMHO a good thing. Robert -- Dipl.-Ing. Robert Schwebel | http://www.pengutronix.de Pengutronix - Linux Solutions for Science and Industry Handelsregister: Amtsgericht Hildesheim, HRA 2686 Hannoversche Str. 2, 31134 Hildesheim, Germany Phone: +49-5121-206917-0 | Fax: +49-5121-206917-9 _______________________________________________ Developers mailing list [email protected] http://dotgnu.org/mailman/listinfo/developers
