James Cheney

I tagged this as a bug because I think the original behavior reported is wrong (compared to what the equivalent HTML+JavaScript would do). But I don't know offhand that this...

This reminds me of something I've been wondering about, incidentally. There's been a line of work on something called "type error slicing", for polymorphic languages, which mostly aims to provide...

All gripers take a `Unify.error` as an argument. `Unify.error` is a variant with two cases: `Msg` and `PresentAbsentError`. Most gripers ignore this argument. The only exception is `record_pattern` which pattern...

The main problem with externalizing unification errors is that they are printing variables as integers as illustrated in #447. The `refresh_tyvar_names` mechanism needs to be used in `unify.ml` to avoid...

Hmm. I'm not sure whether this is a bug per se, in the sense that record extension isn't included in the formalizations of how this is supposed to work, but...

Yes, this seems inevitable since XML can be freely used as an expression, without some kind of disambiguating quasiquote introducer like `xml {foo}` (that might make sense to have anyway...

Snoozing until post-0.9.0

Hi @jmatsushita and @agentm, @rudihorn mentioned this discussion to me. (I'm one of the other authors of the paper Jun mentioned). I hope you don't mind my jumping in with...

Sorry, just wanted to see if that was the only reason the CI was failing or if there were other problems.

This issue seems to have been worked around by #1138. I think it should be closed, and another issue collecting ideas for a more principled design created. Some of the...