And for them I think it works really well, since they get their quantity through those questions, and the quality (high voted, HNQ, etc.) ones through design and general implementation questions...exactly like what we do
@lyxal No, I mean more like, "this Rust code that's supposed to be a MyNewLanguage compiler isn't working. Am I using ANTLR wrong?"
I think there's kind of three places we could draw the line: 1. No physical code questions 2. Only physical code questions involving frameworks/DSLs specifically for langdev 3. Any physical code questions about implementing proglangs
I prono'd that in my head as "maytuh question for beh-ta" I am tired
Quick preemptive argument for why we shouldn't be merged with CS if any CMs get funny ideas while I'm asleep:
I've been a CGCC user for five years, a site that kind of only sort of fits with SE's format. We have to make a lot of compromises as a result of being a part of that bigger thing. I think that's exactly what would happen with PLDI if we were made a part of CS or another site.
We'd be held back by having to stay within, or only moderately change, that site's rules. They might be fine with more theoretical questions, but what about questions dealing with physical code? Using parser generators? Our freedom to expand to topics we think would help the programming languages community would be hampered, and questions that could be answered by the same knowledgable experts would be fragmented across half a dozen sites.
An overwhelmingly common sentiment in the responses to our advertising posts for this proposal, and our commit comments, has been that the programming languages community needs someplace like this, where Q&A about it can be centralized. There's a lot of people from very different backgrounds involved in our scope, and I don't think they're unified nearly often enough.
Anyway, I'm going to go finish folding my laundry and go to sleep now o/
@RydwolfPrograms Even within academic CS, language design work is a hard sell in more generalist venues (as opposed to more formalised PL work like type theory, or optimisation, which have a more empirical patina), so I would not expect a good community fit or reception within CS.se even for the non-code questions. Most language-design questions on the existing site are really in those areas (or just bad questions).
(This is endorsing what you said, just noting that it represents a pre-existing dynamic in the field and it's if anything likely to be an even wider gap than just actual-code questions)
yeah while cs intersects with ld, ld is cs like ux is cs just because academia considers human-computer interaction to be a subfield doesn't mean a q&a platform should merge them either
...though i guess hci isn't exactly ux
but yeah in any case the bulk of cs is nothing to do with ld and the bulk of ld is of no interest to cs
type theory is an interesting case, where the theory itself is squarely cs but i can't imagine a cs platform would love to talk about the practical advantages and disadvantages for implementing a type system :P
Applied PL work is often really dealing with HCI questions, too ("how can a person best communicate about X with a computer"), but I wouldn't expect UX.se to be a match either
A common question is when will our site go into private beta. The true answer is it really depends and things can get in the way such as holidays or certain staff members being absent.
The general rule is about 1 to 2 weeks after the proposal hits 100% on a Tuesday some time after 12:00 PM EST.
I have to say I am not an SE employee. I cannot predict the future. All the things I said is assuming nothing goes wrong between now and then. I had to make it bold.
The most important thing to everyone who wants to build the community is when the site comes out start using it, invite new users, ask, answer, vote, review, participate, etc... You need to show the SE CM team it's a perfectly healthy community to prevent it from being shut down.
Proposal: Proof Assistants
According to this answer, we do not really get to choose the name of the new site. Instead, the StackExchange overlords will create one for us. (They have had some negative experience with people quarreling over the naming of their site.)
As I understand things, unless ...
art — the ISO 639-5 code for constructed languages and prefix code for conlangs not assigned a code.
ref. List of ISO 639-5 codes
Programming Language Design and Imple…
Programming Language Design and Imple…