search

web-platform-tests / wpt

Test suites for Web platform specs — including WHATWG, W3C, and others
https://web-platform-tests.org/
Other
5k stars 3.09k forks source link

Directory naming scheme for IETF specs? #10837

Open foolip opened 6 years ago

foolip commented 6 years ago

We have at least 3 directories for IETF specs: https://github.com/w3c/web-platform-tests/tree/master/cookies https://github.com/w3c/web-platform-tests/tree/master/http https://github.com/w3c/web-platform-tests/tree/master/x-frame-options

Perhaps more. @alvestrand is considering adding tests for https://tools.ietf.org/html/draft-ietf-rtcweb-jsep-24 but what should the directory be called?

foolip commented 6 years ago

Pinging owners of those existing directories: @inikulin @mikewest @annevk @mnot

alvestrand commented 6 years ago

Possibilities I've considered include webrtc/jsep/test (linking it to the w3c spec that uses it), rtcweb/jsep/test (making more room for other rtcweb WG-related specs) and jsep/test.

mikewest commented 6 years ago

shrug Call it whatever makes it most clear to developers what it's actually testing? It doesn't make much sense to me to lean too much on the place where the spec was written. We already have a webrtc directory. Throwing it under there seems fine? Having both webrtc and rtcweb would seem strange.

foolip commented 6 years ago

Sent https://github.com/w3c/web-platform-tests/pull/10838 (needs review) and https://github.com/w3c/web-platform-tests/pull/10839 to add a READMEs to at least point to the IETF specs under test, probably got it wrong :)

alvestrand commented 6 years ago

When actually writing a test to be put into the proposed directory, I thought that "webrtc/protocol" looked like a fine directory name - it's protocol specs rather than API specs, and it's strictly related to WebRTC. https://chromium-review.googlesource.com/c/chromium/src/+/1043886 - upstreamed as https://github.com/w3c/web-platform-tests/pull/10841

annevk commented 6 years ago

FWIW, I think that's reasonable.

jgraham commented 6 years ago

Yeah, I don't have a strong opinion here. I don't know of any generic tooling that particularly relies on the naming (CSS specific tools aren't generic), so we should focus on clarity.

annevk commented 6 years ago

(And it does seem useful to put URLs to the specification in a README, ideally the same URLs used to reference the standard from other standard, so you can grep for it.)