- From: Sean Owen <srowen@google.com>
- Date: Tue, 6 Feb 2007 13:39:21 -0500
- To: "Dominique Hazael-Massieux" <dom@w3.org>, "Jo Rabin" <jrabin@mtld.mobi>, public-mobileok-checker@w3.org
On 2/6/07, Dominique Hazael-Massieux <dom@w3.org> wrote: > Le mardi 06 février 2007 à 08:39 -0500, Jo Rabin a écrit : > > > Probably so, indeed; but I don't know if it is really that much cheaper > > > or easier to define "sufficiently strictly the preceding steps" than to > > > define sufficiently strictly a mobileOK reference implementation. > > > > Well, most of what is suggested here as pre-processing needs to be done however you actually execute the tests, surely? > > Absolutely; I guess my point was that separating the pre-processing from > the tests execution may not be such a big gain in the end; but it's > totally worth a try. I agree that we should attempt to get a DOM out of the resource if at all possible, and write the rest of the tests in terms of the DOM where possible. A resource that isn't valid will still fail VALID_MARKUP but I think it's important to take this small step to be flexible and provide additional info. I'll look into some kind of Java HTML Tidy. Sean
Received on Tuesday, 6 February 2007 18:39:39 UTC