Core team meeting 2015-06-03
Attending
aturon, brson, acrichto, huon, steveklabnik, pcwalton, wycats
Agenda
- rustcamp status (aturon)
- Community subteam status (steveklabnik, aturon)
- Language terminology (aturon) https://github.com/rust-lang/rfcs/issues/1140
- Whistler topics (aturon) https://etherpad.mozilla.org/rust-whistler15
- Nantes updates
- security bugs (wycats)
RustCamp status
- CFP is now open: http://cfp.rustcamp.com/
- Accepting proposals until June 21
- Everyone (including core team) should submit proposals, and we will put together a balanced program.
- Reminder: evaluation and feedback is blind for the first couple of rounds
Community subteam
- Steve is in the process of setting up the team, invites have gone out and most have responded. Should have more news soon.
Language terminology
https://github.com/rust-lang/rfcs/issues/1140
- eddyb raised the point that UFCS is a bit of a misnomer
- There are probably a few other cases (OIBIT, "unboxed" closures) where jargon during RFC process ends up living on.
- We didn't think about this too carefully in the run up to 1.0 since it's not a backcompat issue for the compiler, but with several books in the works, etc., these terms are soon going to be "committed", if they're not already
- aturon: Several of these terms could be improved for clarity and simplicity. Is it worth trying to do this? Would be an epic bikeshed.
- General feeling was that many of the relevant names are about feature deltas that don't matter in the long run (e.g. UFCS, OIBIT) and the names will naturally be phased out over time. We can instead use simple, direct phrases in docs and error messages (e.g. qualified or disambiguated path, default trait, etc.)
Whistler work week
- Probably going to continue the recent move toward breakout sessions, scheduled ad hoc as the week evolves
- Brainstorming for topics here: https://etherpad.mozilla.org/rust-whistler15
Security bugs
-
wycats: There is a lack of clarity around whether segfaults in safe code should be considered security issues and reported as such. (They're currently just being opened as issues). Likewise not clear what security claims we want to make about the language.
-
Live transcript missed, but basic gist was: this still happens enough that we should accept them as normal bugs, and we need to put out a clear statement about the contexts Rust is, and isn't, ready for use in with respect to security.