Re: file bugs to bugzilla from anaconda

[Chris Lumens <clumens redhat com>]

> I think, that all the hashing code should go into the "between" layer
> instead of Anaconda. Because we may consider changing that in the future
> (it just can happen) and some people will be reporting old bugs with
> even older releases (think RHEL with really old anaconda at that time).
> 
> With the detection code in anaconda it will be:
> 1) another possibility for errors, mistakes and typos
> 2) using older duplicate detection mechanism
> 
> And also adapting the "between" layer to new bugzilla API (which can happen too)
> will be much easier, than converting all ancient media someone has.

Sure, we can do something like this.  I'd take the approach of first
getting in what I've done, then working on some intermediate layer to
handle multiple bug handling systems, then adapting anaconda to it.  If
you look at the code in the patch, you'll see the amount that handles
actually filing the bug is really small.  Adapting later should be
fairly straightforward.

It seems to me we need to spend a lot of time thinking about product.img
and how things like this fit into it.  Fun.

- Chris