- From: Mike Schinkel <mikeschinkel@gmail.com>
- Date: Wed, 6 Dec 2006 18:09:44 -0500
Ian Hickson: >> Validators are allowed to give any warnings or notes >> they like. (The spec only specifies that a validator >> must give no errors if there are no errors and must >> give at least one error if there are any, IIRC.) Is it possible for the spec to suggest/recommend that validators present warnings in certain circumstances, i.e. in the case of meaningless markup that is nonetheless allowed? >> > "Implementors SHOULD NOT use string concatonation and >> > SHOULD use one of the HTML5 components listed here: >> > http://wiki.whatwg.org/wiki/Implementations" >> It is not the job of specifications to limit implementation strategies. Which part of my suggestion did you find limiting: a.) "SHOULD NOT use string concatonation", b.) "SHOULD use one of the HTML5 components listed here:", or c.) both? My *intention* for the wording was to give guidance to implementors such that they avoid all the Bad Things(tm) that you and others rant about, and that they are both "made aware of" and "given permission to use" HTML components instead of string concatonation. I'm sure if I can understand your objections with the effect of my wording I can come up with acceptable wording to achieve the same ends. -Mike Schinkel http://www.mikeschinkel.com/blogs/ http://www.welldesignedurls.org/
Received on Wednesday, 6 December 2006 15:09:44 UTC