01:33 | <rbuckton> | It occurred to me while responding to https://github.com/tc39/proposal-pattern-matching/issues/322 that we could potentially simplify a pattern like
Could be simplified to
Though if we did want to support that, I'd almost want it to be a follow on so the community would have time to understand how |
01:34 | <rbuckton> | I should note that this is actually somewhat similar to C# variable patterns as well, since C# variable patterns are actually type identifier , where var is a keyword that means "infer the type". |
01:37 | <rbuckton> | i.e., x is string y and x is string and var y are equivalent in C# 8.0 |
05:30 | <ljharb> | with the recent github discussions i'm wondering if we even want to present pattern matching this meeting :-/ certainly we won't be asking for stage 2 |
11:24 | <dminor> | Please don't ask for Stage 2, we share the concerns about complexity and at the very least need more time to evaluate the proposal as it stands. |
11:25 | <dminor> | I'll open an issue with some of my questions / concerns. We also have a team work week next week, and I'm hoping to do a deep dive on this proposal and get more feedback then. |
12:17 | <rkirsling> | would be interested in knowing whether the issue I had written up resonates for your team |
13:40 | <ljharb> | also i don’t have any slides from TabAtkins yet. anyone object to me removing pattern matching from the agenda? |
13:49 | <rbuckton> | In general I think it would be better to remove all of the syntax that is contentious from the proposal spec text, there's just way too much in there we don't have consensus on. Put the optional or contentious things in its own document to make the spec less intimidating. |
13:56 | <rbuckton> | For example, ? /has /hasOwn look interesting as a follow-on, but maybe not MVP since they each can be solved in different ways. Maybe even RegExp patterns, considering the potential parser ambiguities, and can still be handled via an extractor by pulling the RegExp out of the pattern. Even relational patterns, despite my preference for inclusion, are not necessarily an MVP -- C# did not include them in their initial support for pattern matching either. Put all of these and anything else we lack consensus on into a separate document we can refer to regarding future capabilities, so that we can be sure we're considering them in the design space. |
13:59 | <rbuckton> | I don't think match(v) {} and is are themselves a tough sell, despite cover grammar issues to work out. The pattern syntax is what's expansive and needs to be paired down, so long as we ensure we continue to maintain the pattern matching invariants we've established. |
14:02 | <Jack Works> | also i don’t have any slides from TabAtkins yet. anyone object to me removing pattern matching from the agenda? |
14:03 | <ljharb> | i don’t know if there’s even a point in sharing updates yet - it’ll just invite debate on things we’re likely to still change |
14:13 | <dminor> | i kind of think it would be best to skip the update, we're seeing good conversations on github already |
14:35 | <ljharb> | ok, removing it |
15:12 | <Jack Works> | 🫠 |
15:12 | <ljharb> | maybe we can present at the finland meeting |
15:15 | <Jack Works> | looks like the only way for proposals to get outside reviews is to put it on the agenda asking for an advance |
15:15 | <Jack Works> | this is unhealthy |
15:16 | <ljharb> | i mean in this case we kind of discouraged outside reviews until that time |
17:22 | <gkz> | Hi all, as an introduction - I work at Meta, and there is interest in experimenting with adding support for pattern matching (as part of Flow) internally, and potentially exposing it to our ~10k JS engineers for feedback. So, I'm excited and interested in the design and advancement of this (and related) proposals! |
17:43 | <gkz> | This interest extends beyond pattern matching, to also ADT enums as well (which seem to be discussed above as well), as an extension to Flow Enums |
17:51 | <rbuckton> | This interest extends beyond pattern matching, to also ADT enums as well (which seem to be discussed above as well), as an extension to Flow Enums enum would be https://matrix.to/#/#temp-enum-interests-group:matrix.org |
20:20 | <rbuckton> | Pattern matching humor: that is { question: 0x2b or not 0x2b } |
21:29 | <TabAtkins> | Sorry everyone, I've very suddenly gotten extremely swamped this past week. I was gonna finally have time this afternoon to do the slides, since we were presenting on Thursday, but I guess I don't need to now |