<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0//EN" "http://www.w3.org/TR/REC-html40/strict.dtd"><html><head><meta name="qrichtext" content="1" /><style type="text/css">p, li { white-space: pre-wrap; }</style></head><body style=" font-family:'DejaVu Sans Mono'; font-size:9pt; font-weight:400; font-style:normal;">On Tuesday 21 April 2009 02:11:24 Michael Biebl wrote:<br>
> Kel Modderman wrote:<br>
> > On Saturday 18 April 2009 03:46:04 Henrique de Moraes Holschuh wrote:<br>
> >> On Fri, 17 Apr 2009, Michael Biebl wrote:<br>
> >>> Henrique de Moraes Holschuh wrote:<br>
> >>>> On Fri, 17 Apr 2009, Michael Biebl wrote:<br>
> >>>>> I think, one missing piece is a proper interface for updating init<br>
> >>>>> script priorities (if the depencies or the list of runlevel changes) in<br>
> >>>>> a policy compliant way.<br>
> >>>> There is no such interface in this case (if we had one, insserv would have<br>
> >>>> to make it a no-op). You have to edit the initscript metadata directly<br>
> >>>> (because it is embedded in comment headers on the initscript itself) to do<br>
> >>>> such changes, then tell the system to rebuild the initscript dependency<br>
> >>>> tree.<br>
> >>> How do you do that exactly while preserving local modifications?<br>
> >> The local modifications have to be done on the initscript headers, which are<br>
> >> conffiles since the dawn of time. The user is warned that by switching to a<br>
> >> dependency-based initscript system, the old order information is deemed<br>
> >> irrelevant and thus completely ignored.<br>
> >><br>
> >> There is also an override directory that can be used to change the<br>
> >> dependency headers instead of editing the initscript, but we should get rid<br>
> >> of any need to ship files in there as part of the release goal (local admin<br>
> >> can place stuff there as he wishes).<br>
> >><br>
> >> So, you can have local modifications to the *DEPENDENCY* information in an<br>
> >> override directory.<br>
> > <br>
> > I'm pretty sure there is a misunderstanding here. An interface for modifying<br>
> > unmodified script properties (such as what runlevels it starts/stops) is<br>
> > desirable for use in package maintainer scripts when a new version of the<br>
> > package wishes to change the scripts start/stop or sequence/dependency<br>
> > properties. That is what Michael is poking us about.<br>
> > <br>
> > An interface for this was proposed for legacy (aka sysv-rc's) update-rc.d:<br>
> > http://lists.alioth.debian.org/pipermail/pkg-sysvinit-devel/2008-September/002865.html<br>
> > <br>
> > The discussion went cold after Michael posited that the proposed interface is<br>
> > prone to error because it relies on a human to write out a snippet of shell<br>
> > code in a maintainer script. A similar interface could exist for insserv's<br>
> > update-rc.d. I am without any better ideas at this time.<br>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;"><br></p>I had an idea about this issue this week, we should probably discuss this<br>
in more depth when/if we get together in June for the boot performance meeting.<br>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;"><br></p>> <br>
> Yeah, that's pretty much what I tried to say (thanks Kel for the clarification)<br>
> <br>
> For the record, I don't want it to be understood as I'm against insserv.<br>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;"><br></p>It doesn't need to be said, everytime a discussion pops up about Debian boot<br>
system you seem to be in there with stimulating discussion.<br>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;"><br></p>> To the contrary: as we are already 95% there, it would imho be stupid to not<br>
> make use of it and reap the benefits (though I still think we need something<br>
> more flexible/dynamic long term, the improvements by insserv are nice for the<br>
> short term).<br>
> I'd even be in favour of dropping the static priorities interface (and file-rc<br>
> for that matter), as it would make our lives as maintainers much simpler.<br>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;"><br></p>Yeah, I share those same sentiments.<br>
<p style="-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; margin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;"><br></p>Thanks, Kel.</p></body></html>