IRC channel logs
2023-04-05.log
back to list of logs
<jar286>tsyesika, I'm here but I don't think we need these weeklies. I was insecure about going into my first experience chairing the group but feel I sorta have a handle on it now <jar286>everyone else, I'll be on IRC for about the next hour if anyone wants to talk agenda or whatever <tsyesika>okay, sounds good! If you do need me I usually respond to pings on here relatively quickly in european working hours <Zarutian_iPad>jar286: just one question for the agenda, possibly. Should we go the ze way of having a standard bootstrap object like the NonceLocators on each end? <Zarutian_iPad>and should it have a standard metho that returns an object with all sorts of helper things and constructors for things like maps and sets that would not be in the layer 0 data mofel? <jar286>what's the alternative, and do spritely and agoric disagree? <jar286>oh yeah I'm not on top of this but sometimes those types are built in? <jar286>I like to see things spelled out more concretely <Zarutian_iPad>there has been a whole discussion going on in https:github.com/ocapn/ocapn/issues/5 <jar286>yeah the volume is so high I've had a hard time keeping up. <jar286>I would think the thing to do would be to join the discussion. I searched for 'nonce' there and did not find it <jar286>Not sure I'm the best person to talk to about this technically, but if this is good use of meeting time we can talk about that. I think #5 is already a good candidate for discussoin time <Zarutian_iPad>the issue is the old Lost Resolution bug that WormholeOp was made for as an attempt of an solution <jar286>Oh I didn't know lost resolution was related to boostrap object <jar286>I see you're active in that thread <Zarutian_iPad>naah, the idea I have is to have well known constructors accessible from a standarised bottstrap object. To construct maps, sets and such. <jar286>I have to do triage when it comes to the technical subjects. There were arguments against primitive maps and sets sounded right when it went past me but I can't recreate it now. Getting them from a bootstrap object sounds fine to me. I'm not going to be very opinionated about this <Zarutian_iPad>what I have in mind ocapn implementation wise is to have each level/step rather easy to implement. Re our older discussion on HTTP and IRC protocols <jar286>I don't think anyone will say they're against simplicities. If I have a technical position at all it's that we should make it easy for people to write new ocapn endpoint implementations, so I'm certainly with this. Isn't this just a matter of the peopl who understand it working it out in discussion? And maybe not even meeting time? <jar286>My main problem now is getting lost, having a hard time finding discussion of level 0 and so on <Zarutian_iPad>issue #40 was a break out from #5 to discuss the lost resolution bug thing <jar286>OK. MarkM says "Layer 0 is what we need for interoperability." which sounds fine. But his correspondence table a little later shows sets and dictionaries/maps which confuses me a little <jar286>Basically I see this discussion chugging along productively and feel I can leave the hard work of figuring out the right answer to other people <jar286>Looks like the main problem is not enough attention from spritely <jar286>so maybe that's a reason to take meeting time for it <jar286>maybe we're waiting on tsyesika to deliver the spritely documentation. I think it would be nice to have a draft layer 0 spec available for discussion. I don't know if tsyesika's document will be that spec, maybe yes maybe no <jar286>s/for discussion/as a starting point for discussion/ <isd>I mean, the *currently* data model used by spritely is pretty well documented at preserves.dev, which I linked to from that thread, and it's been useful. It's more a question of where they're willing to go. It sounds like dropping maps & sets is on the table. <cwebber>jar286: as said earlier, we have a deadline this week, but we're going to be giving our review of things publicly next week :) <isd>Zarutian_iPad: I think the design of the bootstrap object is something that deserves a separate issue. Does it have one? <jar286>cwebber, yes, I knew you were under a crunch. looking forward to it <jar286>(to hearing your review that is) <jar286>isd, it sounds as if a new issue might be a good idea. feel free