Version 0.7.11 (unstable)

[Isaac Schlueter]

This is the most stable 0.7 release yet. Please try it out.

Version 0.8 will be out very soon. You can see the remaining issues
on the github issue tracker.

<https://github.com/joyent/node/issues?milestone=10&state=open>

2012.06.15, Version 0.7.11 (unstable)

* V8: Upgrade to v3.11.10

* npm: Upgrade to 1.1.26

* doc: Improve cross-linking in API docs markdown (Ben Kelly)

* Fix #3425: removeAllListeners should delete array (Reid Burke)

* cluster: don't silently drop messages when the write queue gets big
(Bert Belder)

* Add Buffer.concat method (isaacs)

* windows: make symlinks tolerant to forward slashes (Bert Belder)

* build: Add node.d and node.1 to installer (isaacs)

* cluster: rename worker.unqiueID to worker.id (Andreas Madsen)

* Windows: Enable ETW events on Windows for existing DTrace probes.
(Igor Zinkovsky)

* test: bundle node-weak in test/gc so that it doesn't need to be
downloaded (Nathan Rajlich)

* Make many tests pass on Windows (Bert Belder)

* Fix #3388 Support listening on file descriptors (isaacs)

* Fix #3407 Add os.tmpDir() (isaacs)

* Unbreak the snapshotted build on Windows (Bert Belder)

* Clean up child_process.kill throws (Bert Belder)

* crypto: make cipher/decipher accept buffer args (Ben Noordhuis)


Source Code: http://nodejs.org/dist/v0.7.11/node-v0.7.11.tar.gz

Macintosh Installer (Universal): http://nodejs.org/dist/v0.7.11/node-v0.7.11.pkg

Windows Installer: http://nodejs.org/dist/v0.7.11/node-v0.7.11-x86.msi

Windows x64 Installer: http://nodejs.org/dist/v0.7.11/node-v0.7.11-x64.msi

Windows x64 Files: http://nodejs.org/dist/v0.7.11/x64/

Other release files: http://nodejs.org/dist/v0.7.11/

Website: http://nodejs.org/docs/v0.7.11/

Documentation: http://nodejs.org/docs/v0.7.11/api/

[Matt]

On Fri, Jun 15, 2012 at 3:48 PM, Isaac Schlueter <i...@izs.me> wrote:

* V8: Upgrade to v3.11.10

OK so we've had upgrades and downgrades and stability issues - do we actually know now why 3.11 was unstable in Node prior to this version?

[Marco Rogers]

Buffer.concat FTW! I just wrote this myself the other day.

[Tim Dickinson]

Just installed it going to see how well the code ports.

[Isaac Schlueter]

Matt,

The first install of 3.11 V8 was 3.11.1, which was way too early. (My
mistake to even try pulling it in, but hey, the tests passed, so it
must work, right?) That was in 0.7.9.

V8 3.11.1 didn't pass its correctness tests yet, and in some cases it
just did JavaScript wrong. That was bad. Also, it was crazy slow,
and whatever the performance regression was caused by, it was also in
V8 3.10. So, we rolled back to 3.9, which was stable, and got our
performance back. That was in node 0.7.9.

The V8 team saw the problem, corrected the bug, and informed us that
the 3.11 branch was worth trying again. So we did that, it was
faster, and that's where we are now.

It is in our best interest to be on a very recent V8, because it will
generally get more attention from the V8 team moving forward, which
benefits Node.js greatly. The only reason for the rollback was
because we weren't sure whether it'd be stable and performant in time,
but now it is.

[Robert Malko]

get a buttload of warnings when building 0.7.11.  mostly with ev.c, eio.c, and the openssl stuff up front.

not sure it matters because it builds but I thought I'd let you know.

on lion.

[Ben Noordhuis]

On Sun, Jun 17, 2012 at 1:54 PM, Robert Malko <robm...@gmail.com> wrote:
> get a buttload of warnings when building 0.7.11.  mostly with ev.c, eio.c,
> and the openssl stuff up front.
>
> not sure it matters because it builds but I thought I'd let you know.
>
> on lion.

It's mostly harmless. We compile with a lot more compiler warnings
turned on now.

ev.c and eio.c have lots of C constructs that gcc in C89 mode doesn't
like, just ignore it.

openssl has been deprecated by Apple. I kid you not[1].

[1] http://ludovicrousseau.blogspot.nl/2011/08/mac-os-x-lion-and-openssl.html

[Matt]

On Sat, Jun 16, 2012 at 5:22 PM, Isaac Schlueter <i...@izs.me> wrote:

Matt,

The first install of 3.11 V8 was 3.11.1, which was way too early.  (My
mistake to even try pulling it in, but hey, the tests passed, so it
must work, right?)  That was in 0.7.9.

V8 3.11.1 didn't pass its correctness tests yet, and in some cases it
just did JavaScript wrong.  That was bad.  Also, it was crazy slow,
and whatever the performance regression was caused by, it was also in
V8 3.10.  So, we rolled back to 3.9, which was stable, and got our
performance back.  That was in node 0.7.9.

The V8 team saw the problem, corrected the bug, and informed us that
the 3.11 branch was worth trying again.  So we did that, it was
faster, and that's where we are now.

Awesome - I just noticed a pattern and wondered if the team had info on it. Really appreciate the info.

 

It is in our best interest to be on a very recent V8, because it will
generally get more attention from the V8 team moving forward, which
benefits Node.js greatly.  The only reason for the rollback was
because we weren't sure whether it'd be stable and performant in time,
but now it is.

Great news. And I agree completely we should be on as most recent, but stable, version of v8 as possible.

Matt.