Release proposal: v1.6.1

[rvagg]

Primarily to get #1208 sorted.

@iojs/collaborators please don't merge any semver-minor changes to v1.x until we get this out, it'll just make it more complicated.

/cc @mscdex

• [3b9eab9779] - build: make check aliases test (Johan Bergström) #1211
• [4c731042d4] - configure: use cc and c++ as defaults on os x (Ben Noordhuis) #1210
• [8de78e470d] - path: reduce type checking on some methods (cjihrig) #1216
• [c9aec2b716] - querystring: fix broken stringifyPrimitive (Jeremiah Senkpiel) #1213
• [a89f5c2156] - querystring: parse numbers correctly (Jeremiah Senkpiel) #1213
• [2034137385] - smalloc: don't mix malloc() and new char[] (Ben Noordhuis) #1205

[rvagg]

• path: iojs@1.6.0 breaks compatibility with previous versions #1215 with a fix for path.resolve(undefined) -> '.' is another candidate here

[rvagg]

2015-03-20, Version 1.6.1, @rvagg

Notable changes

• path: New type-checking on path.resolve() #1153 uncovered some edge-cases being relied upon in the wild, most notably path.dirname(undefined). Type-checking has been loosened for path.dirname(), path.basename(), and path.extname(), #1216.
• querystring: Internal optimizations in querystring.parse() and querystring.stringify() #847 prevented Number literals from being properly converted via querystring.escape() #1208, exposing a blind-spot in the test suite. The bug and the tests have now been fixed (Jeremiah Senkpiel) #1213.

Known issues

• Possible remaining TLS-related memory leak(s), details at #1075.
• Surrogate pair in REPL can freeze terminal #690
• Not possible to build io.js as a static library #686
• process.send() is not synchronous as the docs suggest, a regression introduced in 1.0.2, see #760 and fix in #774
• Calling dns.setServers() while a DNS query is in progress can cause the process to crash on a failed assertion #894

Commits

[rvagg]

ptal at the above changelog entry @Fishrock123, @mscdex, @cjihrig.

[rvagg]

CI @ c9aec2b: https://jenkins-iojs.nodesource.com/job/iojs+any-pr+multi/349/

[mscdex]

LGTM

[cjihrig]

For the path issue, the reported problem was path.dirname(undefined), not path.resolve(undefined). Other than that, LGTM.

[Fishrock123]

Internal optimizations in querystring.parse() and querystring.stringify() #847 prevented Number literals from being properly converted via querystring.stringify() and querystring.escape() #1208, exposing a blind-spot in the test suite. The bug and the tests have now been fixed (Jeremiah Senkpiel) #1213.

My edit:

Internal optimizations in querystring.parse() and querystring.stringify() #847 prevented Number literals from being properly converted via querystring.escape() #1208, exposing a blind-spot in the test suite. The bug and the tests have now been fixed (Jeremiah Senkpiel) #1213.

Might as well not mention the stringifyPrimitive silliness in there, or we'll have to add a few more sentences. Besides, it was definitely subtly broken all along.

[rvagg]

thanks @Fishrock123 & @cjihrig, fixed

[rvagg]

CI happy, going straight to release

[rvagg]

release build @ https://jenkins-iojs.nodesource.com/job/iojs+release/47/

[rvagg]

done and released @ https://iojs.org/dist/latest/

ping @iojs/website

[Fishrock123]

Website updated: nodejs/iojs.org@0837801

[jbergstroem]

Changelog.md updates are missing for 1.6.0 and 1.6.1.

[Fishrock123]

Huh? They seem to be there: https://github.com/iojs/io.js/blob/v1.x/CHANGELOG.md

[jbergstroem]

..they at least were missing from my local repo, which wasn't properly updated 😞. Sorry for the noise.

[gopat]

Sorry to bother anyone with a ridiculous question but... will there be armv6l builds for this release?
They seem to be missing from https://iojs.org/dist/v1.6.1/ ...also all arm ones from https://iojs.org/dist/v1.6.0/ , but i don't think they'll be missed and this is not the correct gh issue anyway.

Also, wonderful work, everyone. Thanks so much for it.

[Fishrock123]

@gopat I suspect the armv6 1.6.1 builds should be up today sometime. They take considerably longer to build.

[silverwind]

1.6.0 arm builds should be ready according to #1157 (comment), but it appears @chrisdickinson didn't release them yet :)

[gopat]

@Fishrock123 You're right, from your comment ("considerably longer") i just remembered they build on rasppis themselves (don't know how i could've forgotten).
Please don't take my previous comment a sign that i'm in a hurry. I just got a bit concerned that they could be discontinued for some reason.

@silverwind Oops, didn't see that comment (when i knew of 1.6.1, my focus switched here), that would have taken the worries out of my head.

Take all the time you need and more, people, and again, thanks!

[chrisdickinson]

There may be an issue – while promoting the linux arm builds for v1.6.0 I accidentally continued on to v1.6.1, resulting in a "partial" promotion, where the SHASHUMS256.txt and the linux arm tarball were uploaded, but because I couldn't sign the release the rest of the files (including SHASHUMS256.txt.{asc,gpg}) were not updated. At the very least I'll probably need @rvagg's help to get the appropriate SHASHUMS256.txt{,.gpg,.asc} uploaded. Sorry all.

[rvagg]

all sorted, arm builds and signed shasums should be in place for both 1.6.0 and 1.6.1, sorry for the delay and confusion folks

[rvagg]

I should reinforce though--armv6 builds are slower, tho using ccach makes them faster than if you were to do them from scratch yourself, so you just need to expect delay in promoting them, some times there's a personnel delay, as in this case, but usually they should be there between 1 and 24 hours after a release.