Description of problem: Installer finds no issue with KDE Dependancies and allows install to proceed which then crashes for dependancy error Version-Release number of selected component (if applicable): 19.24-1 How reproducible: 100% Steps to Reproduce: 1. Boot Installer 2. Choose Software Spoke and choose KDE 3. Return from Software Spoke and Get Software Error 4. Revisit Software Spoke chosing another software and then KDe before returning to main. 5. Software Spoke returns with no error, so proceed with install. 6. Get Sependancy error Actual results: Good return result from Software Spoke when error exists Expected results: Error Additional info:
Fails to meet release criteria - installer must install all release blocking software
Is this using DVD, netinst, live, what? What dependency error are you actually hitting?
This was encountered using F19-Beta-TC3-DVD and error is for perl-DBD-MySQL-4.023-1.fc19.i686.
Created attachment 743839 [details] Main Spoke after First Visit to Software Spoke Error is obviously visible.
Created attachment 743840 [details] Main Spoke after Second Visit to Software Spoke - Again choosing KDE This makes it appear that one has successfully chosen KDE
Created attachment 743841 [details] The errors seen if one proceeds
The error is also encountered using F19-Beta-TC3-x86_64-DVD
Okay, so basically you're considering two problems here: one there's a dependency issue on the DVD, two anaconda behaves badly if you go into the software spoke twice with a dep error. I'm *almost* sure there's an existing report for the second problem; the devs may know for sure. If not, it should be filed separately. Let's make this bug for the dependency problem itself. It looks like we just need https://admin.fedoraproject.org/updates/FEDORA-2013-6896/perl-DBD-MySQL-4.023-2.fc19 to fix this. That update went stable yesterday, so TC4 should fix it. Let's mark this as ON_QA, and we can close it once we confirm it's good in TC4.
This is essentially a dupe of bug 959610 which was closed in https://bugzilla.redhat.com/show_bug.cgi?id=959610#c3 (though maybe it shouldn't have been).
I think we'll be fine just marking it as a dupe of that. If somehow the fix didn't make TC4, we'll catch it with the test script again anyway. *** This bug has been marked as a duplicate of bug 959610 ***