Summary: | sys-power/apcupsd needs sys-apps/openrc (>=sys-apps/baselayout-2) for powerfail initd (mount-ro service is missing in baselayout-1) | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Robert T Childers <robert.t.childers> |
Component: | Current packages | Assignee: | Diego Elio Pettenò (RETIRED) <flameeyes> |
Status: | VERIFIED FIXED | ||
Severity: | normal | CC: | axs, base-system, gentoo_bugs_peep, grzegorz.chwesewicz, ikelos, remy, scott, sniper |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | x86 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
results of paludis -qD apcupsd-3.14.5-r2
Diff on apcupsd-3.14.5-r2 ebuild with added baselayout-2 dependency |
Description
Robert T Childers
2009-05-28 02:39:44 UTC
mount-ro available only on baselayout2 which not stable yet, and apcupsd-3.14 not asking to install it. So, confirm that. Created attachment 192937 [details]
results of paludis -qD apcupsd-3.14.5-r2
Added a list of the dependencies of the apcupsd-3.14.5 build.
There are confused people in the forums wondering about this: http://forums.gentoo.org/viewtopic-t-711759.html Created attachment 193961 [details]
Diff on apcupsd-3.14.5-r2 ebuild with added baselayout-2 dependency
I added in the dependency for baselayout-2 to apcupsd-3.14.5-r2 version. After adding the new dependency and changing my system from baselayout-1 to baselayout-2 the errors for the missing service have gone away.
Sorry, I'm not up to speed with the baselayout-2 stuff, aside from what was mentioned in #c1. If your patch was checked into the tree, would that pull the masses (baselayout-1) to baselayout-2, or would it simply keep us (something akin to masking) from getting this latest version? I'm _not_ going to add a forced dep on baselayout 2; the powerfail script while very useful is not strictly necessary, you can use apcupsd with baselayout 1 with reduced features, but still working. I'll see to find a different solution. The main reason that I went ahead and moved on to baselayout-2 was because I didn't want to downgrade apcupsd and mask off >=3.14. in order to get it to properly shutdown and restart. The problems I was having with 3.14, in that it didn't want to properly shutdown via /etc/init.d/apcupsd stop went away. It was also giving me trouble getting status info out of it. Why I am not sure, but now that I have fixed the mount-ro service, the multimon.cgi script works properly as well. Since this is depending on something that is still marked unstable (baselayout-2) I don't know what to tell you as far as marking stable or masking it in the repository. I just know that now that I have put in the little patch and upgraded to baselayout-2, its working correctly. As far as I've been able to tell (grepping the sources, etc), APCUPSD itself doesn't need a powerfail init script, it's just for the new baselayout or openrc. Until the new baselayout goes stable, can this package install the new initscripts with a has_version conditional on baselayout-2 or openrc, and install old-style init scripts otherwise?? That way us +arch'ers still have a package that installs appropriately for baselayout-1. That's actually what I had locally, just forgot to commit it the other day; should be fine now. (In reply to comment #9) > That's actually what I had locally, just forgot to commit it the other day; > should be fine now. > is there a change for getting a new ebuild? or do i have to apply the diff manually? It already is IN the sync server. (In reply to comment #11) > It already is IN the sync server. > hi flameeyes! according to http://gentoo-portage.com/sys-power/apcupsd the changes are not in the ebuild available in portage tree yet. greets snIP3r I'm not sure what are you saying. The changes _ARE_ in tree for quite a while already. (In reply to comment #13) > I'm not sure what are you saying. The changes _ARE_ in tree for quite a while > already. > hmmm, ok perhaps some misunderstanding. if i check my local portage tree in /usr/portage/sys-power/apcupsd and the ebuild apcupsd-3.14.7.ebuild i get this: ... DEPEND=" cgi? ( >=media-libs/gd-1.8.4 ${WEBAPP_DEPEND} ) nls? ( sys-devel/gettext ) snmp? ( net-analyzer/net-snmp ) gnome? ( >=x11-libs/gtk+-2.4.0 >=dev-libs/glib-2.0 >=gnome-base/gconf-2.0 )" ... but i assume i should look like this: ... DEPEND=" cgi? ( >=media-libs/gd-1.8.4 ${WEBAPP_DEPEND} ) nls? ( sys-devel/gettext ) snmp? ( net-analyzer/net-snmp ) gnome? ( >=x11-libs/gtk+-2.4.0 >=dev-libs/glib-2.0 >=gnome-base/gconf-2.0 ) >=sys-apps/baselayout-2.0.0" ... also please take a loook here: http://forums.gentoo.org/viewtopic-t-711759.html other users have the same issue. what has to be done to get the new ebuild? thx snIP3r Please read what I wrote, the ChangeLog and whatever else: I'm _not_ going to add a forced dep on baselayout 2; the powerfail script while very useful is not strictly necessary, you can use apcupsd with baselayout 1 with reduced features, but still working. THE EBUILD IS GOOD AS IT IS. (In reply to comment #15) > THE EBUILD IS GOOD AS IT IS. I don't see how having my system now reporting: * Caching service dependencies ... * Can't find service 'mount-ro' needed by 'apcupsd.powerfail'; continuing... [ ok ] , every time I run an init.d script is "GOOD AS IT IS". As the original reporter suggested, you are misreporting the dependencies in apcupsd-3.14.5-r2. You should not be including an init.d script with an unresolved dependency in a stable package. I don't really know how apcupsd.powerfail is used, nor do I really care. But, you attitude about this is wrong. Plain and simple. As per the ChangeLog entry: 16 Jun 2009; Diego E. Pettenò <flameeyes@gentoo.org> apcupsd-3.14.6.ebuild: Only install the powerfail init script if openrc is installed; this way it's not polluting the installation for baselayout 1 users. Closes bug #271480. , this should be done in a apcupsd-3.14.5-r3 and -r2 should be removed from portage. If you did that, then you would actually CLOSE this bug. As it is, you corrected a different bug (against apcupsd-3.14.6). |