[PATCH 2/2] Move code related to finding existing installations to anaconda

Vratislav Podzimek vpodzime at redhat.com
Wed Jan 28 07:22:39 UTC 2015


On Tue, 2015-01-27 at 16:01 -0500, Chris Lumens wrote:
> > +def populate_storage_roots(storage):
> 
> Likewise here - only one new function named differently from all the
> others.
This is a newly added not moved function, that's why I made it properly
named right away. All the other fucntions' names would be fixed in a
follow-up patch set.

> 
> More generally, if you have run these through pylint and all and they
> work, I guess I am okay with it.
Yes, pylint's okay with these changes, installation seems to work okay
too.

>   I'm not real excited by moving storage
> stuff back into anaconda in general, given all the time we spent getting
> rid of it.  To me it makes sense that blivet has a concept of
> autopartitioning and existing installed systems.  I can imagine that
> other blivet-using things are going to want to use that code, and now
> they're going to have to pull in anaconda for the same functionality.
I get your point here and it's definitely a good topic for a broader
discussion. My opinion is that this should go away from blivet's core
because it is really installer specific and it is the biggest reason why
blivet requires pykickstart (the other reasons should be eliminated
too). Maybe python-blivet-autopart or anaconda-autopart or something
like that would be a better place for such code? Suggestions?

-- 
Vratislav Podzimek

Anaconda Rider | Red Hat, Inc. | Brno - Czech Republic



More information about the anaconda-patches mailing list