09:59 <Hodgestar> fangerer: Thank you for the hotel information.

10:00 <Hodgestar> I am catching up on things like IRC after being away last week.

11:31 <antocuni> so, today's call went well. I think that most of our and Mark's goals are aligned and we agreed it would be a good thing to converge

11:31 <antocuni> also, Mark said that he has not much time to actively work on this, so his idea was more to start the repo to develop a discussion

11:32 <antocuni> the biggest diverging point if whether to support moving GC or not

11:33 <antocuni> because his current proposal doesn't even mention it, and he's more interested in other design issues

11:34 <antocuni> but it's obvious that at some point the (C)Python community needs to decide whether moving GCs are first class citizens or not

11:34 <antocuni> I seem to understand that Mark's approach is that he would personally be fine as long as it doesn't impact negatively CPython

11:35 <antocuni> I also invited him to join next week's call

11:36 <phlebas> antocuni: I think the most important thing for us to communicate to the CPython core devs is that we want the CPython API and the C extension API to be two different things.

11:37 <phlebas> because I have the feeling that a lot of the discussion and Mark's ideas are really about the concerns and needs of CPython, not extension devs

11:38 <phlebas> whereas HPy couldn't care less how the future CPython C API looks - the whole point is that if they decide to rewrite in Rust next week or include Boehm for GC, C extensions would be unaffected

11:38 <antocuni> yes, we also talked about this point and I think we are on the same line

11:40 <phlebas> great :)

11:40 <antocuni> also, Mark made an important point "pro GC": eventually, it will be possible to integrate WebAssembly's code with the browser's GC, and currently CPython would be unable to make use of it

11:41 <antocuni> this alone might be the best selling point for HPy

11:41 <phlebas> ah, yes, that's a very good point. I wasn't aware that there are plans to integrate wasm with the JS GC

11:41 <antocuni> https://github.com/WebAssembly/gc

11:42 <antocuni> I'm also happy that in this particular discussion I can wear my HPy, PyPy and PyScript hats all at the same time :)

14:29 <cfbolz> phlebas: all the browsers have a prototype implementation now

14:31 <antocuni> cfbolz: ah, I wasn't aware that it's already implemented. Do you have links?

14:32 <cfbolz> antocuni: nope, 'private communication' 😅

14:33 <cfbolz> I chatted with some of the involved people

14:34 <antocuni> cool

14:34 <antocuni> cool that it's happening, not cool that it's private )

14:34 <cfbolz> It's not private I think. But I don't know what links to give you

14:36 <antocuni> sure :)

14:38 <cfbolz> antocuni: here is the chrome issue I think https://bugs.chromium.org/p/v8/issues/detail?id=7748

14:39 <antocuni> thanks!

14:51 <cfbolz> antocuni: if you have any specific questions I can ask Andy Wingo. He's pretty involved in the whole thing

15:11 <antocuni> no specific questions but good to know that I can ask in case I need. I think that eventually we will want to integrate the Python GC with the WASM GC, but it's going to be a long way

17:33 <cfbolz> antocuni: step 1) migrate everything to hpy

17:34 <phlebas> step 2) ??

17:34 <phlebas> step 3) profit!