Bug 312539

Summary:	dev-ruby/rmagick-2.13.0-r1 should update itself after you upgrade ImageMagick
Product:	Gentoo Linux	Reporter:	Stefan <gentoo-bugs>
Component:	New packages	Assignee:	Gentoo Ruby Team <ruby>
Status:	RESOLVED LATER
Severity:	normal	CC:	gentoo-bugs
Priority:	High
Version:	unspecified
Hardware:	All
OS:	Linux
Whiteboard:
Package list:		Runtime testing required:	---
Bug Depends on:	192319
Bug Blocks:

Description Stefan 2010-04-01 02:03:57 UTC

After emerging www-apps/redmine (with +imagemagick) and using it for some days now, i have upgraded media-gfx/imagemagick from 6.5.9.4-r1 to 6.6.0.10. The result was, that redmine doesn't start anymore, although "/etc/init.d/redmine start" threw no errors at all.

I have done some research and found this error, when trying to start redmine manually:
"/usr/lib64/ruby/gems/1.8/gems/rmagick-2.13.0/lib/RMagick2.so: This installation of RMagick was configured with ImageMagick 6.5.9 but ImageMagick 6.6.0-10 is in use. (RuntimeError)"
The solution is obvious then: re-emerge dev-ruby/rmagick after upgrading media-gfx/imagemagick.

Reproducible: Always

Steps to Reproduce:
1. emerge media-gfx/imagemagick-6.5.9.4-r1, emerge www-apps/redmine (with +imagemagick)
2. emerge media-gfx/imagemagick-6.6.0.10
3. /etc/init.d/redmine start



Expected Results:  
After upgrading media-gfx/imagemagick, dev-ruby/rmagick should be marked for re-emerge.

Comment 1 Samuli Suominen (RETIRED) gentoo-dev

2010-04-01 07:50:37 UTC

re-assigning to ruby if they want to dummy-revision bump the package...

but there's a dev-portage@ bug open for this... if $A is emerged then $B would be as well, currently portage lacks the support for that.

Comment 2 Zac Medico gentoo-dev

2010-04-01 16:58:38 UTC

(In reply to comment #1)
> but there's a dev-portage@ bug open for this... if $A is emerged then $B would
> be as well, currently portage lacks the support for that.

See bug 192319.

Comment 3 Alex Legler (RETIRED) archtester

2010-04-01 18:07:23 UTC

Until portage has implemented that feature, there's nothing we can do.
Fake revbumps sounds like a rather vague workaround to me, I don't think we want to do that.

Besides, this issue is easily discoverable. It should appear in log files or when the host application is started manually, so I'm closing this bug as LATER.