| 17:23 | <tobie> | TabAtkins, others: are there conventions as how to markup stuff that doesn't necessarily get auto-linked or that aren't vars (e.g. this, null, Function)? |
| 17:43 | <annevk> | tobie: varies, recommend following ES |
| 22:30 | <smaug____> | jgraham: does wpt has a place for adding tests for unstable or non-standard APIs? |
| 22:30 | <smaug____> | s/has/have/ |
| 22:31 | <gsnedders> | smaug____: like what? |
| 22:31 | <smaug____> | blink's webkitdirectory and the directory stuff in dnd |
| 22:32 | <gsnedders> | that should probs be in the compat spec given where we are now |
| 22:33 | <smaug____> | sure |
| 22:33 | <smaug____> | but while implementing the mess, could we put the tests to wpt |
| 22:33 | <smaug____> | and then finally document how it all works |
| 22:33 | <smaug____> | (Chrome's implementation seems to be super buggy. ) |
| 22:34 | <gsnedders> | idk if jgraham has stronger opinions, but compat/provision-webkitdirectory.html or something would keep me happy |
| 22:35 | <smaug____> | k |
| 22:54 | <gsnedders> | *provisional |
| 22:59 | <jgraham> | smaug____: There's a directory in m-c that doesn't get upstreamed if you want to use that for now (testing/web-platform/mozilla/tests) |
| 23:00 | <jgraham> | Or just make a directory for these tests under compat/ and don't worry too much. Maybe tell some people what you are doing so that Microsoft or whoever aren't surprised. |
| 23:01 | <gsnedders> | Well MS are/have implemented it |
| 23:02 | <gsnedders> | hence why there's basically no question about them becoming standard |
| 23:04 | <jgraham> | Right, more "surprised if they fail some tests" |
| 23:04 | <jgraham> | The badness I want to avoid is to write tests that don't match webkit and that accidentially cause someone else to change their implementation |
| 23:04 | <jgraham> | Which seems very unlikely but quite avoidable |
| 23:05 | <gsnedders> | s/webkit/blink/ here, but yes |
| 23:05 | <gsnedders> | (WebKit doesn't have it anymore. It was behind a compile-time flag that Apple-WebKit never shipped.) |
| 23:08 | <gsnedders> | jgraham: is it a crazy idea for me to come into office to hack on some wpt/csswg-test stuff in the next few days? I presume not tomorrow… maybe Wednesday so you're settled in a bit? |
| 23:08 | gsnedders | needs to look at the weather to work out what he's doing when |
| 23:10 | <jgraham> | gsnedders: Not crazy, although since we are moving next week, there might be chaos |
| 23:12 | <gsnedders> | well the crazy bit I meant was the whole moving thing |
| 23:13 | <jgraham> | I think it's still allowed. Ask Mandy? |
| 23:15 | <smaug____> | jgraham: I'm not going to write anything here. Outsourcing test and code writing to baku :) |
| 23:16 | <smaug____> | (which does however mean I will need to do the reviews) |
| 23:17 | <smaug____> | jgraham: webkit doesn't have these webkit prefixed APIs |
| 23:17 | <smaug____> | they are blink APIs |
| 23:17 | <gsnedders> | jgraham: I've always dealt with her by proxy :) |
| 23:17 | <jgraham> | OK s/webkit/blink/ in what I said :) |
| 23:17 | <jgraham> | gsnedders: #london |
| 23:17 | <jgraham> | Now: sleep |