Stream: wg-traits

Topic: trait alias expansion


nikomatsakis (Jan 08 2019 at 16:42, on Zulip):

Actually @Alexander Regueiro something about your PR is confusing me:

I didn't notice it before, but the TraitAliasExpander -- why is the stack these TraitAliasExpansionInfo entries (which are small vectors) and not just a single PolyTraitRef?

nikomatsakis (Jan 08 2019 at 16:42, on Zulip):

it seems like they could just be a single predicate

Alexander Regueiro (Jan 08 2019 at 16:43, on Zulip):

Oh... that was because the expansion info (the path taken through expansions), including the spans, was necessary before when I was linting on expanded duplicates. But then we decided we didn't want that.

Alexander Regueiro (Jan 08 2019 at 16:43, on Zulip):

So yes, it can be converted to PolyTraitRef now :-)

nikomatsakis (Jan 08 2019 at 16:43, on Zulip):

Ah, OK.

Alexander Regueiro (Jan 08 2019 at 16:43, on Zulip):

simplified

nikomatsakis (Jan 08 2019 at 16:43, on Zulip):

I thought it might be something like that

nikomatsakis (Jan 08 2019 at 16:43, on Zulip):

OK

Alexander Regueiro (Jan 08 2019 at 16:55, on Zulip):

:+1:

Alexander Regueiro (Jan 08 2019 at 19:42, on Zulip):

how's it all going? :-)

nikomatsakis (Jan 08 2019 at 20:00, on Zulip):

mm OK but I ran out of time and had to run to a meeting; I'm not 100% sure how to integrate the trait alias expansion with the "principal" checking code

nikomatsakis (Jan 08 2019 at 20:00, on Zulip):

I'm debating actually about separating the PR into two

nikomatsakis (Jan 08 2019 at 20:01, on Zulip):

i.e., move that code into a separate PR

Alexander Regueiro (Jan 08 2019 at 22:49, on Zulip):

yeah this is why I had to leave this PR aside heh

Last update: Nov 18 2019 at 01:00UTC