Bug 475564 – app-portage/eix-0.28.5: --installed-slot doesn't work

• Status: RESOLVED FIXED
• Product: Gentoo Linux

[Description Vikraman (RETIRED) 2013-07-02 22:57:13 UTC]

The --installed-slot feature doesn't seem to work in eix 0.28.5. It works fine on 0.25.5.

Reproducible: Always

Steps to Reproduce:
 % eix -cI sys-devel/gcc                                                                     
[I] sys-devel/gcc (4.7.3(4.7)@05/20/2013 4.8.1(4.8)@06/21/2013): The GNU Compiler Collection

Now add --installed-slot
Actual Results:  
 % eix -cI sys-devel/gcc --installed-slot 4.8 
No matches found.
 % eix -cI sys-devel/gcc --installed-slot 4.7 
No matches found.
 % eix -cI sys-devel/gcc --installed-slot 4.6 
No matches found.


Expected Results:  
 % eix -cI sys-devel/gcc --installed-slot 4.8                                                
[I] sys-devel/gcc (4.7.3(4.7)@05/20/2013 4.8.1(4.8)@06/21/2013): The GNU Compiler Collection
 % eix -cI sys-devel/gcc --installed-slot 4.7                                                
[I] sys-devel/gcc (4.7.3(4.7)@05/20/2013 4.8.1(4.8)@06/21/2013): The GNU Compiler Collection
 % eix -cI sys-devel/gcc --installed-slot 4.6                                                
No matches found.

[Comment 1 Vikraman (RETIRED) 2013-07-11 03:34:12 UTC]

diff --git a/src/eix.cc b/src/eix.cc
index 56f1178..8188c91 100644
--- a/src/eix.cc
+++ b/src/eix.cc
@@ -446,7 +446,7 @@ EixOptionList::EixOptionList()
        push_back(Option("name",          's'));
        push_back(Option("slot",          O_SEARCH_SLOT));
        push_back(Option("fullslot",      O_SEARCH_FULLSLOT));
-       push_back(Option("installed-slot", O_INSTALLED_SLOT));
+       push_back(Option("installed-slot", O_SEARCH_INST_SLOT));
        push_back(Option("installed-fullslot", O_SEARCH_INST_FULLSLOT));
        push_back(Option("category",      'C'));
        push_back(Option("category-name", 'A'));

Trivial fix.

[Comment 2 Martin Väth 2013-07-11 05:47:30 UTC]

I must excuse: I was on a lengthy trip where I could not care about
the bug and simply forgot about it afterwards.

Thank you very much for finding the fix which is correct, of course.
It is now (more verbosely) in the master branch (>=eix-0.29.1):
git.berlios.de/cgi-bin/gitweb.cgi?p=eix;a=commitdiff;h=78e355;hp=ff6318b

Ian, would you prefer that I release eix-0.29.1 immediately
(with only this fix), or can we make an -r1 of the ebuild?
I just tested that it would be sufficient to insert the line
(hope bugzilla does not break it)

sed -i -e 's/O_INSTALLED_SLOT/O_SEARCH_INST_SLOT/' "${S}/src/eix.cc" || die

as the first line of src_prepare.

[Comment 3 Ian Stakenvicius (RETIRED) 2013-07-11 13:27:23 UTC]

Up to you, but I'm fine to backport this with a patch into 0.29.0-r1.

If you want to do an immediate release instead, I'll drop 0.29.0 from the tree when i bump, though - no point in having it around in this case.

[Comment 4 Martin Väth 2013-07-11 13:51:50 UTC]

(In reply to Ian Stakenvicius from comment #3)
> Up to you, but I'm fine to backport this with a patch into 0.29.0-r1.

Thanks. I suggest to do this instead of a full version bump:
Experience shows that after such a major change there are more bugs to come
which I would like to close in a "true" 0.29.1 version.

BTW: Except for this bug (which I forgot, unfortunately) all other known bugs
are not regressions, i.e. they shouldn't really hinder stabilization of 0.28.5.

Anyway, especially since users with remote data should have a smoother
experience with eix-0.29.0, it does not hurt to wait yet another month
for stabilization.

[Comment 5 Ian Stakenvicius (RETIRED) 2013-07-11 14:37:36 UTC]

+*eix-0.29.0-r1 (11 Jul 2013)
+
+  11 Jul 2013; Ian Stakenvicius <[e-mail]> +eix-0.29.0-r1.ebuild,
+  -eix-0.29.0.ebuild:
+  revbump to fix bug 475564