6.2.1 #218
6.2.1 #218
Conversation
|
Tests failing :( some of them because of Error: ENOENT: no such file or directory, open '/home/travis/build/nodejs/readable-stream/test/fixtures/keys/ec-key.pem' |
|
we don't actually need that file, it's for the c api |
|
browser failures are probably caused by ngrok limits on free plan. Also there is bug in the ngrok module that prevents open tunnels from being closed, see defunctzombie/zuul#251 (comment). In light of #217 it might make sense to switch back to localtunnel (it worked fine for me in the last few days.) |
|
ok trying without ngrok |
|
@calvinmetcalf why don't we ask ngrok if they could sponsor an account for CI to the Foundation? Or maybe we can buy one. Another solution is to setup our own localtunnel server, I think we should try to avoid that as far as possible (but it's a solution). |
|
Another option is to use karma + karma-sauce-launcher instead of zuul but the ppc issue will not be solved. I don't think there is a ppc binary for sauce-connect. |
|
I think zuul can also run with sauce-connect and there is an addon for Travis, see https://docs.travis-ci.com/user/sauce-connect/#Setting-up-Sauce-Connect. |
|
@lpinca thanks will try that |
|
woot that works thanks @lpinca that being said there is a suble timing bug in one of the tests in 0.8 that just for fun, can be prevented by putting certain console.log statments in place |
|
ok from what I can tell, in node 0.8 setTimeout(fn, 0) will sometimes comeback before process.nextTick |
|
more specifically I think setTimeout(fn, 0) is the same as process.nextTick, making it setTimeout(fn, 1) fixes things |
|
well 1 doesn't but 4 does |
|
I think we should drop 0.8 asap and bump the major version. |
|
naw I got 0.8 working |
| }, | ||
| "scripts": { | ||
| "test": "tap test/parallel/*.js test/ours/*.js", | ||
| "browser": "npm run write-zuul && zuul -- test/browser.js", | ||
| "write-zuul": "printf \"ui: tape\ntunnel: ngrok\nbrowsers:\n - name: $BROWSER_NAME\n version: $BROWSER_VERSION\n\">.zuul.yml", | ||
| "write-zuul": "printf \"ui: tape\nsauce_connec: true\nbrowsers:\n - name: $BROWSER_NAME\n version: $BROWSER_VERSION\n\">.zuul.yml", |
There was a problem hiding this comment.
There is a typo here: sauce_connec → sauce_connect.
This reverts commit c213f1e.
|
hmm, there does seem to still be an issue :( |
|
sign yeah I think your right |
|
We should probably update this to 6.2.2 while we are at it #220. |
|
ok added #220 |
|
though in this case it was because I a new es6 feature was introduced |
calvinmetcalf
force-pushed
the
build-6.2.1
branch
from
ed4962a to
88b384c
Compare
|
bumped it, i don't think there is anything actually wrong with the pull, just the tests |
|
yeah I take this back a browser is failing legitimately |
calvinmetcalf
force-pushed
the
build-6.2.1
branch
from
88b384c to
fbbe09c
Compare
calvinmetcalf
force-pushed
the
build-6.2.1
branch
from
fbbe09c to
1345647
Compare
|
testing manually to see if I can reproduce the android error |
|
weird not able to reproduce the android error when running saucelabs locally |
|
@calvinmetcalf I didn't have issues with localtunnel lately, maybe try to switch again? 😄 |
calvinmetcalf
force-pushed
the
build-6.2.1
branch
from
27f7c9b to
5ec31ba
Compare
calvinmetcalf
force-pushed
the
build-6.2.1
branch
from
5ec31ba to
2e9f4ed
Compare
|
ok all browsers have had at least one green run so I'm calling this green enough |
what I like to see, only tests added