On Fri, 2020-12-11 at 02:51 +0000, Ken Moffat wrote: > On Wed, Dec 09, 2020 at 03:10:35AM +0000, Ken Moffat wrote: > > On Fri, Dec 04, 2020 at 03:25:20AM +0000, Ken Moffat wrote: > > > On Thu, Dec 03, 2020 at 10:50:50PM +0100, Pierre Labastie wrote: > > > > > > > > > > Now to work out what to build ;-) > > > > > An interim report, mainly to highlight some "That hurts!" "Well, > > don't do that, then." items. > > > > After coming to a dead stop (or rather, dead apart from the progress > bar) on kf5 [ see my post to blfs-dev if you care ] I'm coming to > the conclusion that the pain from using jhalfs to find regressions > in the book is probably not worth the effort for me.
Jhalfs is not good for the kf5 page, nor for the plasma5 page. The instructions in the script concerning the /opt links must be suppressed when building for the first time (but not when rebuilding a newer version). Also, the scripts do not perform the "cd" to BUILD_DIR (nor does it create it), so this has to be added. And third, the "bash -e" and "exit" in the script should be removed, because it does not work (not sure exactly why) in a non interactive shell. Looks like the standard input is lost. Adapting jhalfs to these two exceptional pages (similarly to the special Xorg lib, apps, fonts pages) is on my TODO list, but not done yet. > > In case I don't reply any further, I'll mention here that blocaled > does indeed need to be built in a booted xorg session, otherwise the > locale gets screwed up (gnome-terminal fails to start, thinking it > is in a non-UTF-8 locale, and the keymap was stuck on us instead of > my intended en_GB). > Hmm, I'd say that building blocaled should be possible in chroot, but configuring it (at least the "gdbus" command) should be done on a running machine, _before_ lauching Xorg (otherwise the keyboard settings are not good). Pierre -- http://lists.linuxfromscratch.org/listinfo/alfs-discuss FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
