Home | Docs | Forums | Lists | Bugs | Planet | Store | GMN | Get Gentoo!
Not eligible to see or edit group visibility for this bug.
View Bug Activity | Format For Printing | XML | Clone This Bug
'emerge xorg-x11' ->>>> Calculating dependencies ... done! [nomerge ] x11-drivers/nvidia-drivers-173.14.05 USE="acpi gtk (multilib) -custom-cflags" [nomerge ] media-video/nvidia-settings-171.05 [nomerge ] x11-libs/gtk+-2.12.10 USE="X cups jpeg tiff -debug -doc -vim-syntax -xinerama" [nomerge ] x11-base/xorg-server-1.4.0.90-r3 USE="dri hal ipv6 sdl xorg (-3dfx) -debug -dmx -kdrive -minimal (-nptl) -xprint" INPUT_DEVICES="evdev keyboard mouse -acecad -aiptek -calcomp -citron -digitaledge -dmc -dynapro -elo2300 -elographics -fpit -hyperpen -jamstudio -joystick -magellan -microtouch -mutouch -palmax -penmount -spaceorb -summa -synaptics -tek4957 -ur98 -vmmouse -void -wacom" VIDEO_CARDS="nvidia -apm -ark -chips -cirrus -cyrix -dummy -epson -fbdev -fglrx (-geode) -glint -i128 (-i740) -i810 (-impact) (-imstt) -mach64 -mga -neomagic (-newport) (-nsc) -nv -r128 -radeon -rendition -s3 -s3virge -savage -siliconmotion -sis -sisusb (-sunbw2) (-suncg14) (-suncg3) (-suncg6) (-sunffb) (-sunleo) (-suntcx) -tdfx -tga -trident -tseng -v4l (-vermilion) -vesa -vga -via -vmware -voodoo (-xgi)" [ebuild N ] sys-apps/hal-0.5.11-r1 USE="X acpi crypt -apm -debug -dell -disk-partition -doc -laptop (-selinux)" [ebuild N ] dev-python/pyxf86config-0.3.34-r1 [ebuild N ] x11-base/xorg-server-1.4.0.90-r3 USE="dri hal ipv6 sdl xorg (-3dfx) -debug -dmx -kdrive -minimal (-nptl) -xprint" INPUT_DEVICES="evdev keyboard mouse -acecad -aiptek -calcomp -citron -digitaledge -dmc -dynapro -elo2300 -elographics -fpit -hyperpen -jamstudio -joystick -magellan -microtouch -mutouch -palmax -penmount -spaceorb -summa -synaptics -tek4957 -ur98 -vmmouse -void -wacom" VIDEO_CARDS="nvidia -apm -ark -chips -cirrus -cyrix -dummy -epson -fbdev -fglrx (-geode) -glint -i128 (-i740) -i810 (-impact) (-imstt) -mach64 -mga -neomagic (-newport) (-nsc) -nv -r128 -radeon -rendition -s3 -s3virge -savage -siliconmotion -sis -sisusb (-sunbw2) (-suncg14) (-suncg3) (-suncg6) (-sunffb) (-sunleo) (-suntcx) -tdfx -tga -trident -tseng -v4l (-vermilion) -vesa -vga -via -vmware -voodoo (-xgi)" ('ebuild', '/', 'sys-apps/hal-0.5.11-r1', 'merge') depends on ('ebuild', '/', 'dev-python/pyxf86config-0.3.34-r1', 'merge') (hard) ('ebuild', '/', 'x11-base/xorg-server-1.4.0.90-r3', 'merge') depends on ('ebuild', '/', 'sys-apps/hal-0.5.11-r1', 'merge') (hard) ('ebuild', '/', 'dev-python/pyxf86config-0.3.34-r1', 'merge') depends on ('ebuild', '/', 'x11-base/xorg-server-1.4.0.90-r3', 'merge') (hard)
As hinted in red letters, "Gentoo Bugzilla" is not the correct product to file bugs regarding ebuilds. Please use "New -> Gentoo Linux -> Component: Ebuilds" in future.
Ok. It's my first report :)
*** Bug 229127 has been marked as a duplicate of this bug. ***
xf86config is not listed as a dependency of HAL on its website, and is not checked for in its configure.in. There was no rationale given in HAL's changelog as to why this deviation from upstream exist. If there is a valid reason for this dependency to exist, please explain.
(In reply to comment #4) > xf86config is not listed as a dependency of HAL on its website, and is not > checked for in its configure.in. There was no rationale given in HAL's > changelog as to why this deviation from upstream exist. > > If there is a valid reason for this dependency to exist, please explain. > How about you read the changelog and commit history. There's no reason anyone should have to take time out of their day to explain this. Consider this a user exercise.
(In reply to comment #5) > How about you read the changelog and commit history... I've stated that there was no rationale given in the ChangeLog, hence the request for explanation.
*** Bug 231388 has been marked as a duplicate of this bug. ***
Seems that it was added in 0.5.11, but I don't know why it's needed, and couldn't find anything in ChangeLog
I plan to remove the pyxf86config dep on hal unless someone explains me why its needed ?
You know that thing about how the maintainers know more about the software than you do, and thats why they are the maintainer and you aren't? This is one of those times. It's compnerd's thing, he knows what its for, he is in the process of moving from an apartment to another. You will need to ask him and not simply remove it.
pyxf86config is required for migrate-xorg-to-fdi.py (contained in a patchset) to convert xorg.conf keyboard rules to a fdi file. As it is not the case that a user would always have the relevant section defined in xorg.conf, it would be more appropriate to have the user run this utility manually and give appropriate instructions.
*** Bug 243858 has been marked as a duplicate of this bug. ***
(In reply to comment #9) > I plan to remove the pyxf86config dep on hal unless someone explains me why its > needed ? > You will hose every Gentoo user upgrading to a HAL-ified X server from a non-HAL-ified X server. pyxf86config is used to migrate a user's xorg.conf settings into HAL, where the X server now reads the settings from.
> How about you read the changelog and commit history. There's no reason anyone > should have to take time out of their day to explain this. Consider this a user > exercise. um.. there is zero mention of this in hal or pyxf86config's change/commitlogs. am i looking in the right place? anyways, any ideas how we can fix this? it's blocking me from doing any gcc-porting work on packages that pull in xorg right now. :/
Why is migrate-xorg-to-fdi.py done by xorg-server.ebuild instead of hal.ebuild? Maybe an option would be modify xorg-server ebuild for installing the script in /etc/hal/fdi/policy/10-x11-input.fdi and execute it instead of having hal ebuild doing this work
('ebuild', '/', 'sys-apps/hal-0.5.11-r4', 'merge') depends on ('ebuild', '/', 'dev-python/pyxf86config-0.3.34-r1', 'merge') (hard) ('ebuild', '/', 'dev-python/pyxf86config-0.3.34-r1', 'merge') depends on ('ebuild', '/', 'x11-base/xorg-server-1.5.2', 'merge') (hard) ('ebuild', '/', 'x11-base/xorg-server-1.5.2', 'merge') depends on ('ebuild', '/', 'sys-apps/hal-0.5.11-r4', 'merge') (hard) Same on ~amd64 box. Anyone knows workaround?
Just emerge xorg-x11 without hal use flag USE="-hal" emerge -av xorg-x11 And then emerge -auDN world to enable hal on it
(In reply to comment #15) > Why is migrate-xorg-to-fdi.py done by xorg-server.ebuild instead of hal.ebuild? As of hal-0.5.11-r1.ebuild migration is done by hal, and that's crap too, see http://forums.gentoo.org/viewtopic-t-723323-highlight-xorg+hal.html Why not install the migration script and docs with hal unconditionally, get rid of the X dependency, and move the actual migration stuff to xorg-server? Or, if anyone is worried of cluttering his system with three unnecessary files, why not create a x11-misc/hal-xorg-migration package that xorg-server depends upon via hal USE flag? This circular dependency is totally unnecessary.
I agree that this is messy. I'll take the bug with a view of fixing this for hal-0.5.11-r8. Right now I'm rolling a 0.5.11-r7 with the low-hanging fruit. Moving the script back to X.Org seems better from a dependency point of view, but the real fix might be to PDEPEND on a migrate package from X.Org I could use some X11 herd opinion on that.
Maybe the script should be shipped with xorg-server when USE=hal? Donnie, what do you say? Thanks
Sure. We'll need a hal revbump so we can depend on a new revision that no longer provides the migration script. I vaguely recall that the original reasoning for putting it with hal was that older hal revisions (read: everything outside of gentopia overlay) didn't provide the x11-input fdi, there were weird issues going on with who provided it and such.
(In reply to comment #21) > Sure. We'll need a hal revbump so we can depend on a new revision that no > longer provides the migration script. Certainly. Will drop the python script from the "4" patchset and the dependency for -r7.
+*hal-0.5.11-r7 (04 Feb 2009) + + 04 Feb 2009; <chainsaw@gentoo.org> metadata.xml, +hal-0.5.11-r7.ebuild: + Avoid circular dependencies between HAL and X by dropping migration script + and py86 dependency, closes bug #225091 by Ster0n <er0trashbox@gmail.com>. + Properly ignore Dell recovery partitions, patch supplied by Pacho Ramos + <pacho@condmat1.ciencias.uniovi.es> in bug #236784. Stop hiding unmounted + partitions, as suggested by Kirill Igorevich Cherniy + <kirill_i_cherniy@rambler.ru> in bug #247025. Upstream patch identified by + Rafael <rktspm@gmail.com> fixes detection of partitions over ~1.2TB in + size, closes bug #253815 by Yar Odin <yarodin@gmail.com>. Cease using + deprecated keys in our installed examples, as pointed out by Sebastian + Günther <samson@guenther-roetgen.de> in bug #256304.
(In reply to comment #20) > Maybe the script should be shipped with xorg-server when USE=hal? Looks like we'll want to ship the script in a separate package that xorg-server PDEPENDs on. As mentioned on IRC, xorg-server can't depend on pyxf86config being installed by pkg_postinst (where the script would run) because pyxf86config depends on a library provided by xorg-server.
*** Bug 233757 has been marked as a duplicate of this bug. ***
We decided to break the circle by dropping the migration script entirely. HAL now no longer deps on X at all (the X useflag only installs example FDI files for Xorg input handling). We'll write a proper Xorg/HAL configuration guide instead. Closing fixed.
