Re: Fixing The Implementation of Traits.
[Thread Prev] | [Thread Next]
[Date Prev] | [Date Next]
- Subject: Re: Fixing The Implementation of Traits.
- From: Ori Bernstein <ori@xxxxxxxxxxxxxx>
- Reply-to: myrddin-dev@xxxxxxxxxxxxxx
- Date: Tue, 22 Aug 2017 23:01:33 -0700
- To: myrddin-dev@xxxxxxxxxxxxxx
On Sun, 20 Aug 2017 20:32:00 -0700, Ori Bernstein <ori@xxxxxxxxxxxxxx> wrote:
>
> If you've got some trait heavy code, throw it at this and see if you
> can break it!
So, there's one known bug, but I merged it anyways. The trait checking code is
a little lax when checking whether a parameters to a generic impl is
sufficiently constrained by its traits.
For example:
impl foo @a::numeric =
fooify = {a; -> a + 1}
;;
fooify("not a number")
This is relatively easy to fix, mainly needing some additional checking in
parse/type.c:863. Still, it seems to fix enough of the issues that people
were running into to merge it now.
--
Ori Bernstein
| Fixing The Implementation of Traits. | Ori Bernstein <ori@xxxxxxxxxxxxxx> |
- Prev by Date: Fixing The Implementation of Traits.
- Previous by thread: Fixing The Implementation of Traits.
- Index(es):