On Wed, Sep 9, 2009 at 12:28 PM, Ganesh Sittampalam <[email protected]> wrote:

> On Wed, 9 Sep 2009, Jason Dagit wrote:
>
>  Anyway, I'm almost 100% certain we can't hit the impossible case.  I seem
>> to
>> recall that David had asked me to hold off and refactoring it simply
>> because
>> my focus was elsewhere previously (one simple change at a time).  Once the
>> type-witnesses are 100% I think we should revisit get_common_and_uncommon.
>> Actually, I feel like a lot of the Depends module needs modernization and
>> cleanup.  We still see bugs there way more regularly than we should and
>> the
>> way the algorithms are expressed there are too many unsafeCoercePs in that
>> module.
>>
>> I think I'll go make a bug ticket for this future refactor.
>>
>
> Actually I did the refactor on the train and am just cleaning it up to
> submit :-)
>

Wonderful!

I made the ticket anyway because this cleanup of get_common_and_uncommon
isn't the only work we need to do in this module.  Feel free to update the
ticket after you submit your patches:
http://bugs.darcs.net/issue1604

Jason
_______________________________________________
darcs-users mailing list
[email protected]
http://lists.osuosl.org/mailman/listinfo/darcs-users

Reply via email to