On Sat, 2010-04-10 at 15:49 +0000, Aaron Bentley wrote:
> I don't agree that this is the cause. Adding excludes support to
> iter_changes would be one approach, but it would be equally effective to
> filter the iter_changes output, and this would avoid the need to
> implement such filtering on every iter_changes implementation.
Assuming that filtering can be done correctly by layering (I'm not sure
that that is the case), we could build that helper into the core.
On Sat, 2010-04-10 at 15:49 +0000, Aaron Bentley wrote:
> I don't agree that this is the cause. Adding excludes support to
> iter_changes would be one approach, but it would be equally effective to
> filter the iter_changes output, and this would avoid the need to
> implement such filtering on every iter_changes implementation.
Assuming that filtering can be done correctly by layering (I'm not sure
that that is the case), we could build that helper into the core.
-Rob