On Monday 2013-02-04 18:38 -0500, Kyle Barnhart wrote: > To summarize, some maintain that our parser library needs to be compliant > with the sytnax rules of the specification and so the parser library tests > should ensure that. The parser section is not required but useful. I > maintain that the parser library must conform the parser rules in the > specification, and that checking syntax rules is not required but useful. > We both maintain that the test for the parser library should reflect the > part of the specification that is required.
To help understand how the specification is written, it may be helpful to read http://ln.hixie.ch/?start=1140242962&count=1 , in particular, on the distinctions between different sorts of conformance requirements. Well-written specifications generally distinguish between conformance requirements that apply to content producers and those that apply to processors of that content. (For a slightly longer read on a similar topic, it may also be useful to read http://www.w3.org/TR/qaframe-spec/ .) And as roc said, it's proven very important on the Web that implementations follow the same rules for rejecting invalid content, and that we test that the rules are implemented as specified. If this isn't done, then authors will end up unintentionally writing and publishing invalid content because it works in the implementation in which they tested it, only to later discover that it doesn't work in another implementation. -David -- 𝄞 L. David Baron http://dbaron.org/ 𝄂 𝄢 Mozilla http://www.mozilla.org/ 𝄂 _______________________________________________ dev-media mailing list [email protected] https://lists.mozilla.org/listinfo/dev-media
