[Agda] type-checking x=x

[Guillaume Brunerie]

Hi Sergei,

This is well-known indeed, see
https://github.com/agda/agda/issues/4165 or
https://github.com/agda/agda/issues/42.

Best,
Guillaume

Den tors 20 aug. 2020 kl 13:38 skrev <mechvel at scico.botik.ru>:
>
> On 2020-08-19 22:37, James Wood wrote:
> > Hi Sergei,
> >
> > This sounds like a fairly serious problem (the termination checker not
> > being run at the right time, or something). Do you have a minimal
> > working example?
> >
>
> Most probably this issue is known, is indexed,
> probably other main developers would recall it.
> If you write that it is difficult to find it, then I would probably try
> to reduce the example.
>
> Regards,
>
> ------
> Sergei
>
>
>
> > On 18/08/2020 19:00, mechvel at scico.botik.ru wrote:
> >> Probably this has been discussed earlier, but for any occasion:
> >> I see that
> >>     begin ...
> >>           p ∙ (q ∙ (p' ∙ (b ∙ q'a)))
> >>     ∎
> >>     where  pp' = p ∙ p';  qq' = q ∙ q';  q'a = q'a
> >>
> >> is not type-checked in Agda 2.6.1 in a reasonable time.
> >> Needs it to report something like
> >> "Cycle in assignments: q'a = q'a "
> >> ?
> >>
> >> Regards,
> >>
> >> ------
> >> Sergei
> >> _______________________________________________
> >> Agda mailing list
> >> Agda at lists.chalmers.se
> >> https://lists.chalmers.se/mailman/listinfo/agda
> >>
> > _______________________________________________
> > Agda mailing list
> > Agda at lists.chalmers.se
> > https://lists.chalmers.se/mailman/listinfo/agda
> _______________________________________________
> Agda mailing list
> Agda at lists.chalmers.se
> https://lists.chalmers.se/mailman/listinfo/agda