Public SourceMage Discussion List

[Eric Sandall <eric AT sandall.us>]

Quoting Seth Alan Woolley <seth AT positivism.org>:
> On Thu, Jun 30, 2005 at 01:05:31PM -0700, Eric Sandall wrote:
> > Quoting Seth Alan Woolley <seth AT positivism.org>:
> > <snip>
> > >I think what I meant was to check in PREPARE and fail out at that stage
> > >so that it doesn't accidentally create bug 8682 for old upgraders.
> > >
> > >Just a different way of fixing it I guess ... not sure if that's the
> > >"only" way to fix it, but I think that's what we decided on one time in
> > >IRC.
> >
> > I thought that was the fix already in place and to which I was referring
> > to as the fix that is working.
>
> I think you mean:
>
> 2004-07-06 Arwed v. Merkatz <...>
>        * DEPENDS: added dependency on gcc so they're built in correct order
>          on a system-update
>
> That's the correct patch for people who upgrade the "correct and
> supported" way, meaning sorcery queue cast --queue at the same time as
> grimoire is updated so that you're always working with the spell set
> that spell authors will expect.
>
> My proposed change, however, I don't see applied to any grimoire, which
> would make it safer for people who break the rule and update the
> grimoire just to cast a spell (which is not uncommon) and who were going
> from unsplit to split gcc-related spells.
>
> By now everybody "should" be on the new split spells, so the fix is less
> important, but I still think it's a good idea to have because it has
> major repercussions (and let's say you are converting over from another
> distro with a non-split gcc package the same problem might come up if we
> get convert2smgl.sh in major production).
>
> On the one hand, I'm able to add patches myself, but some people think
> QA team shouldn't be fixing problems -- instead they should be reporting
> problems.  So on stuff that nobody's taking any effort to do like
> ncurses, I do it myself; on stuff that people should be able to do who
> are more familiar with the spell, I try to get them to do it.

Perhaps I'm misunderstanding the problem you are talking about, but the
fix I'm talking about is:

$ gaze PRE_BUILD g++
if [ "$(installed_version gcc)" != "$VERSION" ]; then
 message "${PROBLEM_COLOR}Installed version of gcc is
$(installed_version gcc), not $VERSION!"
 message "Update gcc first.${DEFAULT_COLOR}"
 return 1;
fi
default_pre_build                &&
cd $BUILD_DIRECTORY              &&
unpack $SOURCE2 ${MD5[1]}        &&
mk_source_dir  $SOURCE_DIRECTORY.bld

So g++ will not cast if gcc was not updated first. This fix is in devel,
test, stable/0.0, and stable-rc/0.1.

-sandalle

--
Eric Sandall                        |  Source Mage GNU/Linux Developer
eric at sandall.us PGP: 0xA8EFDD61  |  http://www.sourcemage.org/
http://eric.sandall.us/             |  SysAdmin @ Inst. Shock Physics @
WSU
http://counter.li.org/  #196285     |  http://www.shock.wsu.edu/

----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.