On Mon, 10 Jul 2017, Jakub Bogusz wrote: > On Mon, Jul 10, 2017 at 04:57:30PM +0200, Tomasz Pala wrote: > > On Mon, Jul 10, 2017 at 20:02:48 +0900, Jan Rękorajski wrote: > > > > > If you want me to keep this commit and directory then follow up by: > > > > > > a) updating rpm macros > > > > Yes, I was considering this point. Just wondering, what would break (in > > theory: nothing should) and how to perform the validation. Didn't want > > to do such change without more feedback, so now - if you already > > summoned this subject, I'll wait a few days for any comments. > > > > I've already reviewed these and only one (re)definition needs to be > > adjusted (in /usr/lib/rpm/macros.d/pld), remaining macros seem to be > > cascading properly. > > Note that there are some inter-package consistency requirements. > > And just like some packages having hardcoded /usr/libexec, and "require > hackery" to use libdir subdirectory, the others have hardcoded /usr/lib** > for this purpose and would "require hackery" to use libexec. > > Without using libexec consequently, I don't see any profits (single > place for internal binaries).
I see a profit - not doing that hackery. The other hackery is for mixing arch and noarch subpackages. Right now it's either all or nothing - libdir or datadir (see git-core ping-pong), or hacking program to understand both locations. With libexec dir we'll have it the way author wanted and will be able to build noarch subpackages. -- Jan Rękorajski | PLD/Linux SysAdm | baggins<at>pld-linux.org | http://www.pld-linux.org/ _______________________________________________ pld-devel-pl mailing list [email protected] http://lists.pld-linux.org/mailman/listinfo/pld-devel-pl
