On Sun, Aug 19, 2007 at 06:29:16PM +0200, Tommy Pettersson wrote:
> Looks like I need the 'optimize --reorder' bugfix in stable
> after all (for issue494), so I've backported it. I may have
> gotten it wrong. Especially I'm not sure if
> 
>   | IsEq <- invert x =\/= y = Just z
> 
> is equivalent to
> 
>   | eq_patches (invert x) y = Just z
> 
> And I omitted all the FL things although some of that stuff
> already is in the stable repo. Do I miss out on some
> optimization because of that? Then I could take a deeper look
> and redo it.

I think this should be fine.

> This also creates a five level deeply nested merger, which could
> be a real killer (in the bad since of the word). Just try
> Obliterate and say No to one of the patches, and watch darcs
> spin...

I'm not sure.  I guess what to do would depend on how much we expect to
change stable before merging with the experimental stuff.  We could also do
a branch that we don't expect to merge together ever again.  :(
-- 
David Roundy
http://www.darcs.net
_______________________________________________
darcs-devel mailing list
[email protected]
http://lists.osuosl.org/mailman/listinfo/darcs-devel

Reply via email to