It seems to me that pypy is the obvious candidate for blessing as the new reference implementation (because it is self-hosted), but I never see any mention of this being so, with most next-gen effort still being spent on 3.x and CPython. Why is this so?
1) We don't have Python3 support yet, we're working towards it, but we're behind some. Have the reference implementation be a regression over the previous one doesn't make much sense.
2) We don't have enough users.
And honorable mention reason: RPython is an... interesting language to program, something like a cross between C, Python, Java, with error messages from MUMPS.
Yes, it probably is. It's a bit of a chicken and egg situation though, everyone who knows RPython well enough to make the changes can already understand the errors just fine.
Comments
It seems to me that pypy is the obvious candidate for blessing as the new reference implementation (because it is self-hosted), but I never see any mention of this being so, with most next-gen effort still being spent on 3.x and CPython. Why is this so?
2 reasons IMO:
1) We don't have Python3 support yet, we're working towards it, but we're behind some. Have the reference implementation be a regression over the previous one doesn't make much sense.
2) We don't have enough users.
And honorable mention reason: RPython is an... interesting language to program, something like a cross between C, Python, Java, with error messages from MUMPS.
> error messages from MUMPS
Is fixing that tractable? Cryptic error messages is one of the most frustrating turn-offs to new would-be contributors.
/probably one of the few people here who has written MUMPS code
Of all the things that I dislike about MUMPS, error messages aren't anywhere near the top of the list.
Yes, it probably is. It's a bit of a chicken and egg situation though, everyone who knows RPython well enough to make the changes can already understand the errors just fine.
heh, epic alum?
I'm currently at epic, but I know there are other companies still using the language.