- From: Aubrey Jaffer <agj@alum.mit.edu>
- Date: Tue, 25 Sep 2007 11:26:42 -0400 (EDT)
- To: ot@w3.org
- CC: jkorpela@cs.tut.fi, www-validator@w3.org
| Date: Tue, 25 Sep 2007 16:06:53 +0900 | From: Olivier Thereaux <ot@w3.org> | | On Tue, Sep 25, 2007, Jukka K. Korpela wrote: | > As far as I remember, the problem was the idea of making the parsing | > itself something different from the kind of SGML parsing required by HTML | > 4, instead of just issuing warnings _in addition to_ doing the very job | > that a validator is supposed to do. | > Or maybe it was just the implementation, especially making the validator | > explicitly claim that a document is invalid when it is in fact valid. | | Indeed, after looking closer at the history, it seems that the "fussy" | parsing was unpopular because fussy parsing would report shorttag usage | as "invalid" when it is, strictly speaking, valid. It's just valid, yet | wrong most of the times. Giving warnings here should cause little fuss. | | I've hacked around this and made the shorttag-triggered messages | warnings instead of errors, and added some explanations to the obscure | warning messages issued by opensp. | | test at: | http://qa-dev.w3.org/wmvs/HEAD/check?uri=http%3A%2F%2Fqa-dev.w3.org%2Fwmvs%2FHEAD%2Fdev%2Ftests%2Fshorttags.html That looks good! One minor issue: in a small browser window there is nothing in the initially visible portion of the validation report indicating that there are warnings. It might be nice to give a parenthetical count of warnings in the "Result" field: Result: Passed validation (4 warnings)
Received on Tuesday, 25 September 2007 15:27:05 UTC