W3C home > Mailing lists > Public > public-test-infra@w3.org > July to September 2012

Re: Metadata flag for automated tests

From: Robin Berjon <robin@berjon.com>
Date: Thu, 19 Jul 2012 15:01:25 +0200
Cc: "<public-test-infra@w3.org>" <public-test-infra@w3.org>
Message-Id: <00EC1713-0EF1-47ED-8E88-32B1582841EE@berjon.com>
To: Ms2ger <ms2ger@gmail.com>
On Jul 18, 2012, at 19:02 , Ms2ger wrote:
> On 07/18/2012 02:37 PM, Robin Berjon wrote:
>> I've looked at the list of known flags and have been trawling through a number of suites, but I can't seem to find any flag that is used to indicate that a test can run fully automated (by which I mean not just run as script and provide an immediate result, but also use testharness.js so that the results can be automatically captured and submitted).
>> 
>> Have I missed something?
>> 
>> If not I'd like to propose that we add the ***drumroll*** "automated" flag to tests so that these can be prioritised and get to run first.
>> 
>> Thoughts? Screams? Meh?
> 
> Fully automated should be the default; everything that isn't flagged as a reftest or a manual test is automated. I would object to requiring more boilerplate for the kind of automated tests we want to encourage.

Yeah, that's what I initially thought, but there seem to be quite a few tests that aren't automated (or, in a few cases, that are but that aren't using testharness.js; e.g. the Element Traversal tests). I'm fine with sticking to that plan, and I'm certainly all for an end to all boilerplate, but in that case we'll need people to update their metadata in a number of cases (I plan to do an all-encompassing review of existing suites in the system next month anyway, so I can add that on my checklist  in fact I might be able to automate it).

-- 
Robin Berjon - http://berjon.com/ - @robinberjon
Received on Thursday, 19 July 2012 13:02:00 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:20:33 UTC