- From: Ian Hickson <ian@hixie.ch>
- Date: Wed, 2 Sep 2009 22:27:20 +0000 (UTC)
- To: Mark Baker <distobj@acm.org>
- Cc: Julian Reschke <julian.reschke@gmx.de>, Jonas Sicking <jonas@sicking.cc>, "public-html@w3.org WG" <public-html@w3.org>
On Tue, 1 Sep 2009, Mark Baker wrote: > On Mon, Aug 31, 2009 at 6:24 AM, Ian Hickson<ian@hixie.ch> wrote: > > It is pointless to provide semantics of elements (or other features) that > > are obsolete > > Obsolescence is a forward-looking statement; don't use these features > in the future. That's fine and good, and I agree with the list of > obsolete features assembled in HTML 5. > > Obsolescence does not change the meaning of existing content which > happens to use those features, nor does it mandate that existing > content be updated to remove them. Sure, but that doesn't matter. Old documents will still be processed by software in a backwards-compatible manner, since the processing requirements on old features are still present in HTML5. > > other than the semantics that form the element's (or the > > feature's) normative user-agent conformance criteria, since the only > > effect of such semantics is in deciding whether the element (or feature) > > is being used correctly, and obsolete elements (and features) can never be > > used correctly, since they are obsolete and must never be used at all. > > It is a fact that obsolete features are in wide use today by content > served as text/html. Therefore, any HTML specification(s) normatively > referenced by the media type registration needs to define what that > content means. It needs to define how the content should be handled. What it means only matters from the point of view of authoring the content in the first place; after the content is authored, what matters is how it is processed. > FWIW, I had a look through the obsolete list last night and found some > which had properly defined semantics. For example, "plaintext" was > defined to mean the same thing as "pre". That's good, and what we > should strive for for the examples given by Julian. If you think that what is said for <plaintext> (i.e. a UA conformance requirement on the processing of the obselete feature) is acceptable, then what features are lacking these requirements? -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Wednesday, 2 September 2009 22:24:53 UTC