Stream: t-compiler

Topic: pre-design meeting 2019-12-13


matklad (Dec 09 2019 at 15:49, on Zulip):

Hi @T-compiler/meeting -- a few of us from @WG-rls2.0 took a stab at creating a hackmd for the upcoming design meeting this Friday. The topic is our plans regarding rust-analyzer, the "official" choice of IDE, and rustc. The document summarizes our current thinking and also includes a section at the end for the discussion agenda. My hope is that people can review the doc now, before-hand. You can ask questions here if there are things that are unclear or could be clarified, and perhaps leave ideas for broader discussion topics in the document. Then we can prune the agenda on Thursday or so before the actual meeting.

(the above note was written by nikomatsakis and lightly edited by @matklad )

@pnkfelix, @Igor Matuszewski we are especially interested in your feedback on the doc!

pnkfelix (Dec 09 2019 at 19:58, on Zulip):

(was ill today but I'm planning to look at this tomorrow)

centril (Dec 10 2019 at 10:31, on Zulip):

Can you describe what producing events would look like?

Also, I would appreciate not changing the fundamental architecture of librustc_parse without OKing it with myself, @Esteban K├╝ber, and @Vadim Petrochenkov first.

centril (Dec 10 2019 at 10:33, on Zulip):

this bit re. polunious sounds dangerous; we shouldn't be compromising the quality of rustc to make things easier for RA

matklad (Dec 10 2019 at 13:13, on Zulip):

Also, I would appreciate not changing the fundamental architecture

Sorry if "tentative plan" sounded like "hey, we've unilaterally decided to do X". Of course any significant changes to architecture of things are discussed, approved and co-designed by at lest the relevant experts, and preferably the whole of t-compiler.

matklad (Dec 10 2019 at 13:15, on Zulip):

Can you describe what producing events would look like?

As-if one traverses a tree of tokens in dfs order.

Last update: May 26 2020 at 11:10UTC