Changes to Sauce Labs

[coveralls]

Coverage remained the same at 97.89% when pulling a8a12634e46795511758011d0e990e522491f8f7 on sauce into 66c92136481a095f96a87a0de3f41587c9087045 on master.

[dchest]

@evilaliv3 what did you do to fix "403 Forbidden" error for badge (https://saucelabs.com/browser-matrix/evilaliv3.svg)? It giving the error when I was not logged in, but then it worked. I have now the same problem: https://saucelabs.com/browser-matrix/dchest.svg. Maybe it just needs time? :)

[evilaliv3]

IT WORKS!

[dchest]

Yes, your badge works, but mine gives 403 (when I'm not logged in - https://saucelabs.com/browser-matrix/dchest.svg)

I noticed the same with your badge first yesterday, but then it worked :)

https://github.com/dchest/scrypt-async-js/tree/sauce

[evilaliv3]

in relation to the svg, accessing directly now works without problems: https://saucelabs.com/browser-matrix/dchest.svg

the problems seems to lies in the emage cached by github CDN while showing badges. it seems that saucelabs does not use HTTP Caching headers correctly so that github reload the image online after a big refresh timeout. it happened to me also yesterday as you noticed.

[evilaliv3]

in addition if you want you can protect the saucelabs token this way: https://docs.saucelabs.com/ci-integrations/travis-ci/

i dont't find it so relevant as i think nobody will abuse your free saucelabs account but who knows :)

[evilaliv3]

you find how i did here: https://github.com/globaleaks/GlobaLeaks/commit/4613b8c94cc698dac55dd6072a77217025288e07

[coveralls]

Coverage remained the same at 97.89% when pulling ad389448e19ae6aa13f1603b33c5cfdae9ba0c8f on sauce into 66c92136481a095f96a87a0de3f41587c9087045 on master.

[coveralls]

Coverage remained the same at 97.89% when pulling 39c5e29eac686d26626087e1f7c0a3d3f13d8848 on sauce into 66c92136481a095f96a87a0de3f41587c9087045 on master.

[evilaliv3]

@dchest probablly this pull can be closed?

[dchest]

Waiting for CI to finish and merging. Thanks for your help!

[evilaliv3]

glad to help

i'm really interested in a patch that in case the interrupt step is set to -1 will skip totally to use the SetTimeout as i found that using a value like 1000000 is totally different from skipping it all due to the overhead introduced and the context change. are you still working on it? if now i will do today.

[dchest]

Yes, I wanted to do it today.

[coveralls]

Coverage remained the same at 97.89% when pulling ff448f173e574d1e716bdc601f6eb20dcf7048c3 on sauce into 66c92136481a095f96a87a0de3f41587c9087045 on master.

[coveralls]

Coverage remained the same at 97.89% when pulling 69db1d42da8b971d3a211fa486a0a009b8ae2108 on sauce into 66c92136481a095f96a87a0de3f41587c9087045 on master.

[coveralls]

Coverage remained the same at 97.89% when pulling 9f91c38b7658f32a187efdbcd73f555c9f77d534 on sauce into 66c92136481a095f96a87a0de3f41587c9087045 on master.

[coveralls]

Coverage remained the same at 97.89% when pulling 15d106ca1d4463d5c7222e28b79cd5923a6dbb42 on sauce into 66c92136481a095f96a87a0de3f41587c9087045 on master.

[evilaliv3]

@dchest i would suggest a little edit: https://github.com/globaleaks/GlobaLeaks/blob/devel/scripts/travis_saucelabs.sh here you find the configurations that are interesting for current saucelabs support. i've added: opera, android, ipad

here you have a demo: https://saucelabs.com/u/globaleaks

[evilaliv3]

here you can read about the browsers/platforms available: https://docs.saucelabs.com/reference/platforms-configurator/

give it a look an let me know that if there is something more that you would find useful

[dchest]

Having Opera is pretty useless, as they are on Chromium now, but having Android and iPad would be good. Thanks, I'll add them.

[evilaliv3]

i've updated the list: https://github.com/globaleaks/GlobaLeaks/blob/536435c60ad4ccb47d60094a556852ed9c9a499f/scripts/travis_saucelabs.sh

as starting point fo GlobaLeaks each browser on it's main OS we are running a test of any version. e.g. firefox is tested one for the stable version on linux, but all version on Windows e.g. chrome is tested one for the stable version on linux, but all version on Windows

we will run this for some days until we will recognize some of the differences/errors and when we will have a clear idea of the differences we will make a reduction to the relevant versions.

i will put an update here.

\cc @fpietrosanti