Bug 327233 – media-sound/rubyripper-0.6.0 - Version bump

• Product: Gentoo Linux
• Status: RESOLVED FIXED

[Description Daniel Pielmeier 2010-07-06 23:38:47 UTC]

rubyripper-0.6.0 is out since 2010/06/18.

Attached ebuild diff tries to fix bug #299210 and uses the new ruby-ng eclass. Rubyripper now can use cdrdao (see attached metadata diff).

@ruby team: Maybe one of you can take a look at the changes.

[Comment 1 Daniel Pielmeier 2010-07-06 23:39:07 UTC]

Created attachment 237811 [details, diff]
rubyripper-0.6.0.diff

[Comment 2 Daniel Pielmeier 2010-07-06 23:39:38 UTC]

Created attachment 237813 [details, diff]
metadata.diff

[Comment 3 Diego Elio Pettenò (RETIRED) 2010-07-06 23:53:56 UTC]

Hrm, ruby_add_rdepend usually gets the whole otherwise-RDEPEND value so

ruby_add_rdepend "dev-ruby/pkg1 dev-ruby/pkg2 foo? ( dev-ruby/pkg3 )"

the problem is that I haven't added support for USE-based dependencies there  yet :/ Not sure if [foo][ruby_targets_ruby19] works...

Also, you should append to both DEPEND and RDEPEND, so

RDEPEND="$RDEPEND virtual/eject ..."

[Comment 4 Daniel Pielmeier 2010-07-07 00:10:34 UTC]

(In reply to comment #3)
> Hrm, ruby_add_rdepend usually gets the whole otherwise-RDEPEND value so
> 
> ruby_add_rdepend "dev-ruby/pkg1 dev-ruby/pkg2 foo? ( dev-ruby/pkg3 )"

Okay.

> the problem is that I haven't added support for USE-based dependencies there 
> yet :/ Not sure if [foo][ruby_targets_ruby19] works...

Recognized this. Portage throws an
invalid atom: '>=dev-ruby/rcairo-1.8.0-r1[svg][ruby_targets_ruby18]'
error, but only if I set ruby_add_rdepend after the normal RDEPEND. Setting it before seems to work, at least portage does not bail out here.

> Also, you should append to both DEPEND and RDEPEND, so
> 
> RDEPEND="$RDEPEND virtual/eject ..."
> 

If I add this I get the invalid atom error even if I set ruby_add_rdepend before normal RDEPEND.


So how to work around about the missing use dependency support in the eclass?

[Comment 5 Diego Elio Pettenò (RETIRED) 2010-07-07 00:27:14 UTC]

I'll try to fix the eclass tomorrow for that case, I was able to procrastinate it up to now.

And yes of course if you do RDEPEND="virtual/eject.." it won't bail out: you're ignoring the dependencies altogether, it would do as much good as commenting out ruby_add_rdepend...

[Comment 6 Daniel Pielmeier 2010-07-07 08:17:16 UTC]

(In reply to comment #5)
> I'll try to fix the eclass tomorrow for that case, I was able to procrastinate
> it up to now.

Thanks for your help!

> And yes of course if you do RDEPEND="virtual/eject.." it won't bail out: you're
> ignoring the dependencies altogether, it would do as much good as commenting
> out ruby_add_rdepend...

Dooh, I see now. First I had it after RDEPEND and RDEPEND="$RDEPEND virtual/eject ..." was not needed because the eclass adds the dependency after previous defined deps. Obviously this does not work without $RDEPEND if I do it before.

[Comment 7 Diego Elio Pettenò (RETIRED) 2010-07-08 04:01:08 UTC]

I've implemented the merge now.

[Comment 8 Daniel Pielmeier 2010-07-08 19:13:49 UTC]

Created attachment 238051 [details, diff]
rubyripper-0.6.0.diff

(In reply to comment #7)
> I've implemented the merge now.
> 

Thanks for making the necessary changes to the class. This diff should work now.

[Comment 9 Daniel Pielmeier 2010-07-09 14:40:08 UTC]

@sound team: Any objections if I commit the changes and add myself to maintainers?

[Comment 10 Samuli Suominen (RETIRED) 2010-07-09 14:48:20 UTC]

(In reply to comment #9)
> @sound team: Any objections if I commit the changes and add myself to
> maintainers?
> 

go ahead, it's only a good thing if rubyripper gets a dedicated maintainer...

[Comment 11 Daniel Pielmeier 2010-07-09 19:08:16 UTC]

(In reply to comment #10)
> 
> go ahead, it's only a good thing if rubyripper gets a dedicated maintainer...
> 

Done.