search

nodejs / node

Node.js JavaScript runtime ✨🐢🚀✨
https://nodejs.org
Other
107.61k stars 29.6k forks source link

Release proposal: 2.3.0 #1939

Closed Fishrock123 closed 9 years ago

Fishrock123 commented 9 years ago

I think it would be prudent to have a 2.3.0 release before 3.0.0, as we've accumulated quite a few commits.

https://github.com/nodejs/io.js/compare/79bb5e10fb45934e51ae336b6e8cc1368def06dd...91d0a8b19cceadf5d05f1d3a0f1616b9d0727be7 (changelog-maker broke with the rename of things..?) (there's some semver-minor stuff in there.)

cc @rvagg

thefourtheye commented 9 years ago

Can https://github.com/nodejs/io.js/pull/1870 and https://github.com/nodejs/io.js/pull/1892 make it in 2.3.0, if they land in near future?

jbergstroem commented 9 years ago

I'd like to get #1934 fixed since it's colouring our jenkins bots red which may distract attention from other bugs.

rvagg commented 9 years ago

Sorry @Fishrock123 I think I neglected to npm publish an update for it. Give it another go. Thanks for pushing through this release, we're overdue.

ChALkeR commented 9 years ago

@rvagg

[91d0a8b19c] - (SEMVER-MAJOR) win,node-gyp: enable delay-load hook by default (Bert Belder) iojs/io.js#1433

Is this correct?

Release proposal: 2.3.0

Edit: Ah, that's in 2.x and already was in 2.0.

rvagg commented 9 years ago

yeah, for the last release (at least, perhaps more?) I removed the SEMVER-MAJOR from that so as not to confuse, it's a floating patch now.

Fishrock123 commented 9 years ago

Oops, accidentally deleted that from my phone.

The semver-major is a false positive.

@rvagg were you also traveling? I'd like to get this out on like Friday if possible, but I'll be at nodeconf so maybe we should wait until Monday?

Trott commented 9 years ago

At the risk of being alarmist over something trivial, might #1761 be semver-major?

(Meta-question: Should I actually be talking about this in #1761 or here? Or is that dealer's choice?)

It changes swallowing tabs in readline if there's no tab-completion functionality enabled behavior to tabs are now permitted in the user input if there's no tab-completion functionality enabled.

The old behavior wasn't documented (as far as I know), but it does seem like it was a choice and by design.

And it seems at least conceivable that it is relied upon by some people in a way that will break things for them. (And they may see the existing behavior as a feature and not a bug.) Something like "My code took the user input and stripped out all the space characters but now it comes in with tabs and so there's whitespace and when I pass the string to my database call it blows up." Granted, that's a scenario where someone is Doing It Wrong, but a breaking change is a breaking change.

Fishrock123 commented 9 years ago

If it is deemed semver-major I suggest reverting and relanding on next.

On Jun 11, 2015, at 7:41 AM, Rich Trott notifications@github.com wrote:

At the risk of being alarmist over something trivial, might #1761 be semver-major?

(Meta-question: Should I actually be talking about this in #1761 or here? Or is that dealer's choice?)

It changes swallowing tabs in readline if there's no tab-completion functionality enabled behavior to tabs are now permitted in the user input if there's no tab-completion functionality enabled.

The old behavior wasn't documented (as far as I know), but it does seem like it was a choice and by design.

And it seems conceivable that it is relied upon by some people in a way that will break things for them. Something like "My code took the user input and stripped out all the space characters but now it comes in with tabs and so there's whitespace and when I pass the string to my database call it blows up." Granted, that's a scenario where someone is Doing It Wrong, but a breaking change is a breaking change.

— Reply to this email directly or view it on GitHub.

Trott commented 9 years ago

Heh. Hope I'm wrong then.

On Thu, Jun 11, 2015 at 8:17 AM Jeremiah Senkpiel notifications@github.com wrote:

If it is deemed semver-major I suggest reverting and relapsing on next.

On Jun 11, 2015, at 7:41 AM, Rich Trott notifications@github.com wrote:

At the risk of being alarmist over something trivial, might #1761 be semver-major?

(Meta-question: Should I actually be talking about this in #1761 or here? Or is that dealer's choice?)

It changes swallowing tabs in readline if there's no tab-completion functionality enabled behavior to tabs are now permitted in the user input if there's no tab-completion functionality enabled.

The old behavior wasn't documented (as far as I know), but it does seem like it was a choice and by design.

And it seems conceivable that it is relied upon by some people in a way that will break things for them. Something like "My code took the user input and stripped out all the space characters but now it comes in with tabs and so there's whitespace and when I pass the string to my database call it blows up." Granted, that's a scenario where someone is Doing It Wrong, but a breaking change is a breaking change.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHub https://github.com/nodejs/io.js/issues/1939#issuecomment-111169744.

brendanashworth commented 9 years ago

AFAIK tabs were the only character not accepted by readline, which was caused by an incorrectly implemented feature that made all tabs disappear, regardless of whether or not the feature was on. It was marked as BUG in the code - but swallowing tabs isn't really something someone should rely on, or would imho. It definitely wasn't intentional.

rvagg commented 9 years ago

@Fishrock123 thanks for the heads-up, I think it's important enough to get this out ASAP so I'll ship it this time tomorrow.

Question about semver-major moved to #1761, if a leaning towards semver-major arises then we'll move it to next and revert off master for this release.

shigeki commented 9 years ago

The bug fix version but not security fix of openssl-1.0.2c will be released soon. https://mta.openssl.org/pipermail/openssl-announce/2015-June/000033.html We'd better wait it for a moment.

shigeki commented 9 years ago

1958 is opened for upgrading to openssl-1.0.2c

shigeki commented 9 years ago

Upgrading to 1.0.2c was finished. Please go ahead to release.

Fishrock123 commented 9 years ago

CI: https://jenkins-iojs.nodesource.com/view/iojs/job/iojs+any-pr+multi/823/

Fishrock123 commented 9 years ago

test-cluster-worker-wait-server-close.js appears to fail on master as of recently.. No error log in TAP.

Fishrock123 commented 9 years ago

https://github.com/nodejs/io.js/pull/1961 should land beforehand. The readline: allow tabs in input is better done in a major.

EDIT: landed!

rvagg commented 9 years ago

test-cluster-worker-wait-server-close.js is failing because of a timeout - note how it's at 60s for each failure, we can't release without this one resolved because it's across so many build slaves, so I'm investigating but other help would be appreciated too if anyone else has time.

rvagg commented 9 years ago

ah, looks like it landed @ https://github.com/nodejs/io.js/pull/1953, thanks @sam-github

https://jenkins-iojs.nodesource.com/job/iojs+any-pr+multi/825/ running again to see whazzup

trevnorris commented 9 years ago

My alternative test shows that it was timing out because the master client couldn't connect to the worker server. Id it's run directly on the same box it should pass. Something to do with setting the LOCALHOST environment variable. @jbergstroem was helping me test this out and has a gist with more information.

jbergstroem commented 9 years ago

@trevnorris LOCALHOST is not really relevant -- It's used to verify ipv4 testing in FreeBSD jails where routing 127.0.0.1 will give you the jail ip, making tests like localhost == 127.0.0.1 fail.

@rvagg looks green to me! (I also verified this being fixed by #1953 but took at bit longer to land since the merge to iojs/master was slightly delayed)

rvagg commented 9 years ago

working on release notes, this is a big deal cause of so many commits ..

rvagg commented 9 years ago

How does this look, anyone?

2015-06-13, Version 2.3.0, @rvagg

Notable changes

rvagg commented 9 years ago

ping @saghul, @zkat, @shigeki, @indutny, @cjihrig, @vkurchatkin for review before I push the release button ^

jbergstroem commented 9 years ago

@rvagg perhaps also mention that uv_os_homedir is exposed in iojs as os.homedir().

Edit: Scheduling a new appointment at the eye doctor.

cjihrig commented 9 years ago

I'm not commenting on the OpenSSL bit, but the rest of it LGTM.

rvagg commented 9 years ago

tagged & building @ https://jenkins-iojs.nodesource.com/job/iojs+release/73/

amendments to the changelog can be done via PR if anyone has any, sorry for the short time on call for review!

shigeki commented 9 years ago

LGTM

rvagg commented 9 years ago

running smoke test and seeing some failures on coffeescript, grunt and request but I don't think they are related to this release .. quickly investigating

alexlamsl commented 9 years ago

Just downloaded the MSI (x64, Windows Server 2012 R2) and perform an install over 2.2.1.

npm package manager now changed to Feature will be installed when required - is that intentional?

rvagg commented 9 years ago

nope, not intentional, nothing has changed with msi building for a long time, can you screenshot that for us?

alexlamsl commented 9 years ago

capture

alexlamsl commented 9 years ago

Completely uninstalling and deleting io.js files from the system works around this issue, i.e. on fresh attempt all components are marked to be installed.

saghul commented 9 years ago

The libuv part LGTM. On Jun 13, 2015 5:44 AM, "Rod Vagg" notifications@github.com wrote:

ping @saghul https://github.com/saghul, @zkat https://github.com/zkat, @shigeki https://github.com/shigeki, @indutny https://github.com/indutny, @cjihrig https://github.com/cjihrig, @vkurchatkin https://github.com/vkurchatkin for review before I push the release button ^

— Reply to this email directly or view it on GitHub https://github.com/nodejs/io.js/issues/1939#issuecomment-111666531.

silverwind commented 9 years ago

All binaries including armv6 are up, so this release is done.