5.3 KiB

Raw Blame History

Node Foundation TSC Meeting 2015-09-16

Present

Rod Vagg (TSC)
Brian White (TSC)
Steven R Loomis
Fedor Indutny (TSC)
Bert Belder (TSC)
Colin Ihrig (TSC)
Trevor Norris (TSC)
Shigeki Ohtsu (TSC)
Chris Dickinson (TSC)
Jeremiah Senkpiel (TSC)
Ben Noordhuis (TSC)
Alexis Campailla (TSC)
Domenic Denicola
Michael Dawson

Agenda

Extracted from tsc-agenda labelled issues and pull requests in the nodejs org prior to meeting.

nodejs/node-v0.x-archive

Deprecate Array#values() in 0.12.x #25877
Deprecate smalloc in v0.12 #25784

nodejs/node

Inspecting Node.js with Chrome DevTools #2546
util: Remove p, has been deprecated for years #2529

Minutes

Due to IBM’s acquisition of StrongLoop, IBM is over the TSC 25% company limit rule. Michael Dawson and Steven Loomis have stepped down from TSC, but they are here as observers.

Standup

Rod Vagg: Released 4.0.0, also 3.3.1 maintenance release. Getting the build system ready for building/publishing 0.10 and 0.12 binaries.
Brian White: not much, triaging issues and PRs, some work on dns parser
Steven R Loomis: tsc resignation for IBM/SL merge, some issue triage, working on a more robust download for ICU (ICU server was overloaded, fixed to ICU again redirects to sourceforge, discussed longer term discussions)
Fedor Indutny: v8’s ArrayBuffer fixes, reviewing PRs, fixing issues
Bert Belder: very few reviews/issue comments. Some libuv work on rwlocks.
Colin Ihrig: not much, some issue tracker work
Trevor Norris: Trying to make things (including Buffers) faster, some issue/PR review
Shigeki Ohtsu: Just a few reviews of issues.
Chris Dickinson: Working on shoving static npm analysis data into a database, slowly
Jeremiah Senkpiel: NodeConf.eu, issues and PRs, 4.1.0 release
Ben Noordhuis: only PR review and responding to bug reports, a patch or two upstreamed to v8
Alexis Campailla: ARM cross-compile for CI. Misc fixes for VS 2015 / Win10
Domenic Denicola: some issue and PR review, talking with v8 team
Michael Dawson: landed the build updated for AIX, also updating the tests for AIX where necessary. Looking into PPC failures. Working on getting us hooked into v8 security notifications.

Previous meeting’s agenda

Jenkins merge jobs always overwrites PR-URL and Reviewed-By #179
doc: update COLLABORATOR_GUIDE #2638
deps: update v8 to 4.5.103.30 #2632
Node.js v4 Release Timeline #2522
Inspecting Node.js with Chrome DevTools #2546

Deprecate Array#values() in 0.12.x #25877

Deprecate smalloc in v0.12 #25784

Jeremiah: Are these things that we think we can do? Should we go ahead with that? Is it a good idea to deprecate these things?

Ben: How are we going to deprecate? Literal util.deprecate?

Jeremiah: Probably?

Trevor: They’re already floating a patch

Rod: I’m fine with deprecating smalloc

Trevor: Is there a deprecation cycle between LTS? I’m not sure how this would work? (i.e., if we deprecate now, LTS -> LTS upgraders will not see a deprecation warning.)

Rod: Maybe punt to LTS WG? It could be quite a lot — it could be that we decide between stable versions to remove things (or we’re forced to), do we backport the deprecation to LTS?

Jeremiah: I think it should be only for things we’re forced to remove

Action: Punt to LTS WG

Inspecting Node.js with Chrome DevTools #2546

Trevor: Have an 80% understanding of the work needed: WebSocket support in core (though not necessarily exposed publicly), API injection points. Are we ok with including web sockets support one way or another?

Rod, Domenic, others: +1

util: Remove p, has been deprecated for years #2529

Discussion of Buffer “raw{s,}” encoding as well — undocumented, deprecated, inconsistently implemented in core.

Action: Remove util.p (Jeremiah to respond.)

Bert: Perhaps we should bulk remove?

Rod: util.p is relatively trivial, has a deprecation warning, etc.

Jeremiah: Brought this up because these things have been deprecated forever. util.exec has been deprecated for a while, as well.

Trevor: raw/raws are deprecated in name only, no warnings, but C++ layer does not know about them (falls back to utf8) — maybe should be removed as a bugfix, because it does not work consistently.

Action: Chris to list “least used stuff” (with apologies for how longs its taken!)

Next Meeting

September 23

5.3 KiB Raw Blame History