- From: Robin Berjon <robin@w3.org>
- Date: Wed, 19 Sep 2012 20:53:28 +0200
- To: James Graham <jgraham@opera.com>
- CC: Rebecca Hauck <rhauck@adobe.com>, Tobie Langel <tobie@fb.com>, Kris Krueger <krisk@microsoft.com>, Odin Hørthe Omdal <odinho@opera.com>, "public-test-infra@w3.org" <public-test-infra@w3.org>
On 19/09/2012 19:17 , James Graham wrote: > On Wed, 19 Sep 2012, Rebecca Hauck wrote: >> BTW, I nominated this topic as a session idea on the Plenary day at TPAC >> next month. >> >> http://www.w3.org/wiki/TPAC2012/SessionIdeas#Moving_W3C_repositories_to_git > > I see that's listed as a session for someone else to organise/lead. If > no one else wants I am quite happy to run such a session. +1 >> On 9/19/12 9:33 AM, "Tobie Langel" <tobie@fb.com> wrote: >>> On 9/19/12 6:27 PM, "Kris Krueger" <krisk@microsoft.com> wrote: >>> I guess that depends whether we're planning to rely on the developer >>> community to significantly help with test authoring... or not. > > Indeed this is a factor; github gets lots of exposure so we are much > more likely to get test contributions from authors encountering bugs in > their day-to-day work if we have a presence there. It is pretty annoying > to get to the stage of making a pull request only to be told "do some > midly compelex magic to format the patch for hg, then apply it to a > different repository". Indeed I would expect most people to give up at > that point. Completely agreed. Having our test repositories in our own corner is a great way to ensure that no one finds them outside of our (small) community, and that no one ever contributes. And we don't need to make this complicated. I'd rather forgo hg entirely, and come up with a way of having some GH repos push straight to w3c-test.org. It's technically possible (though I have to look at the PHP-reviewing step). Additionally, GH has an API (if an ugly one), which means that we can build tools on top of it. > FWIW I strongly doubt that authors will ever be providing the majority > of testsby number or anything. But their tests are valuable because they > come from actual problems they have experienced. So I really don't think > it is wise to operate in our own ghetto. I think that depends entirely on how much momentum things like TTWF manage to create around testing. It's an uphill battle, but I wouldn't give up just yet :) -- Robin Berjon - http://berjon.com/ - @robinberjon
Received on Wednesday, 19 September 2012 18:53:37 UTC