Public SourceMage Discussion List

[seth AT swoolley.homeip.net]

On Fri, Sep 29, 2006 at 09:13:58PM +0200, Jaka Kranjc wrote:
> On Friday 22 September 2006 00:50, seth AT swoolley.homeip.net wrote:
> > Here's a short little note to say that I've decided, as per Jeremy's
> > recent mail about the git conversion, to go ahead with the next steps of
> > the release process.
> Great. :)
> 
> >
> > to review:
> >
> >/.../
> >
> > The actual candidate for the stable-rc branch may take me the weekend,
> > or part way through next week to cut the first stable-rc from the new
> > integrations for mass prometheus testing.  
> Was any work done on the bug n-plication problem I mentioned "earlier"? I'd 
> like to repeat how much the duplicates annoy me and I think they just waste 
> my time (probably not just m(in)e). I am willing to work on fixing this ...

The duplicates only happen because you broke it very very badly, or a 
very important spell was broken, and even then, only in a way it wasn't 
able to detect.  It already does some duplicate prevention by not 
pulling in known bad spells to build, it's really the end process that 
needs to figure out what actually failed in odd cases.

So, duplicate elimination really isn't a "trivial" fix, and it's not 
really entirely solvable because dependencies can cause brokenness that 
we will never be able to always predict.  That's why the reports go into 
the prometheus section first, so they don't pollute the more true 
non-prometheus sections.

That being said, there are a few duplicates that can be avoided that can 
be implemented before we start this run, but it's not in any means a 
critical flaw that should delay our processes, which already support 
tracking duplicates and resolving bugs through triage already.

If you want to help, #sourcemage-quality is where we go to talk about 
it.  You can work on it since it's in git, just like everything else.  
Just let me know what you are working on fixing so I can avoid 
conflicting with your work.

Seth

> _______________________________________________
> SM-Discuss mailing list
> SM-Discuss AT lists.ibiblio.org
> http://lists.ibiblio.org/mailman/listinfo/sm-discuss