https://bugzilla.novell.com/show_bug.cgi?id=896009 https://bugzilla.novell.com/show_bug.cgi?id=896009#c3 Jan Matejek <jmatejek@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|INVALID | Summary|ghc-* and pandoc packages |IPython: turn nbconvert |pulled in for no reason |into a subpackage with | |hard-requirement on pandoc Severity|Normal |Enhancement --- Comment #3 from Jan Matejek <jmatejek@suse.com> 2014-09-12 14:02:00 UTC --- yes, the distro is affected in that we have to lock pandoc out of the DVD. on the other hand, "fixing" this problem by removing pandoc dependency introduces a different problem - namely that actual functional dependency is not expressed. the full functionality of IPython does in fact depend on external tool pandoc, which in turn requires haskell's runtime libraries. This is a question of semantics: IPython is part of the default selection, and this pulls in more packages into the default selection, a relatively big set that many users consider unnecessary. On the other hand, you could say the same about, say, the whole GTK3 subsystem. The problem, if any, is that IPython depends on a tool that doesn't reuse much of the existing default selection (preferring "its own" dependencies). A potential way to solve this would be to spin off IPython's non-core functionality into a subpackage. But then we just move the problem down a level: Should this subpackage be Recommend:ed by IPython or not? The point of the Recommends tag is to suggest additional functionality; the point of --no-recommends switch is to refuse this by default. Having the nbconvert tool as a subpackage with a hard requirement on pandoc does a better job of expressing the dependencies, so I'm reopening the bug with a different summary and "enhancement" severity. Still, whether nbconvert should be recommended by IPython is up for discussion. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.