- From: Tobie Langel <tobie@w3.org>
- Date: Wed, 27 Mar 2013 23:50:44 +0100
- To: Rebecca Hauck <rhauck@adobe.com>
- Cc: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
On Wednesday, March 27, 2013 at 11:46 PM, Rebecca Hauck wrote: > I'm somewhat of the mind that something is better than nothing. If a spec > doesn't have a person to make these assessments, we can only look at the > statistics that the spec analysis algorithm gives us. It's not ideal, but > better than completely guessing. Oh, absolutely. > BTW, I just closed my previous pull request and replaced it with a new one > [1] - json data is there for all of the specs in TV and Mobile profiles, > with the exception of CSS2.1. Awesome. That PR looks very similar to the previous one. Am I missing anything? > For some reason that spec isn't being > parsed. I'm debugging it right now, but didn't want it to hold up the rest. It's multi-page! > > -Rebecca > > [1] https://github.com/w3c/web-platform-tests/pull/43 > > On 3/27/13 3:25 PM, "Tobie Langel" <tobie@w3.org (mailto:tobie@w3.org)> wrote: > > > Right, I'm trying to validate the spec-analysis model. Hence trying to > > get out of this circular logic. > > > > --tobie > > > > > > On Wednesday, March 27, 2013 at 11:23 PM, Rebecca Hauck wrote: > > > > > If you mean the way they have Status: XXX% complete, no, not that I'm > > > aware of. I don't think I've ever seen any estimates for "how many tests > > > do we need to be complete?" that would inform that percentage number. > > > The > > > best we have to go on is the spec analysis + shepherd data for now. > > > > > > > > > > > > On 3/27/13 3:10 PM, "Tobie Langel" <tobie@w3.org (mailto:tobie@w3.org)> > > > wrote: > > > > > > > I was curious whether there had been any manual assessment of test > > > > coverage. Eg. how that had been done in WebApps[1]. > > > > > > > > --tobie > > > > > > > > --- > > > > [1]: http://www.w3.org/2008/webapps/wiki/PubStatus > > > > > > > > > > > > On Wednesday, March 27, 2013 at 10:57 PM, Rebecca Hauck wrote: > > > > > > > > > The landing page of Shepherd summarizes the test counts for each > > > spec: > > > > > > > > > > http://test.csswg.org/shepherd/ > > > > > > > > > > > > > > > You can drill down from there. > > > > > > > > > > With the shepherd API in the coverage scripts, we'll see the same > > > > > numbers > > > > > aligned with their respective spec sections with the word analysis. > > > > > > > > > > Note: If you don't see the spec you're looking for on the Shepherd > > > main > > > > > page, that means there are no tests. > > > > > > > > > > > > > > > On 3/27/13 1:22 PM, "Tobie Langel" <tobie@w3.org (mailto:tobie@w3.org) > > > (mailto:tobie@w3.org)> > > > > > wrote: > > > > > > > > > > > Hi, > > > > > > > > > > > > Do we have any info on the level of coverage of the different test > > > > > suites? > > > > > > > > > > > > Thanks, > > > > > > > > > > > > --tobie
Received on Wednesday, 27 March 2013 22:50:53 UTC