- From: John Cowan <cowan@ccil.org>
- Date: Tue, 23 Jun 2009 12:45:38 -0400
- To: "Grosso, Paul" <pgrosso@ptc.com>
- Cc: public-xml-core-wg@w3.org
Grosso, Paul scripsit:
> If you are going to discuss an issue, please use the numbers > to
refer to issues and potential resolutions.
My takes:
1: agreed
2: agreed
3: I prefer g and reject the others, because AssocSS defers to HTML4,
and HTML4 says in 12.3.2:
When the LINK element links an external style sheet to a document,
the type attribute specifies the style sheet language and the
media attribute specifies the intended rendering medium or
media. User agents may save time by retrieving from the network
only those style sheets that apply to the current device.
The plural, style sheet*s*, in the second sentence clearly licenses
browsers to accept more than one style sheet.
4: agreed
5: I prefer g and reject the others, by analogy with XML attribute
processing (it is a fatal error to have duplicate attributes).
6: I prefer b and reject a. Bad charrefs are garbage, and bad
(non-predefined) entity refs are an attempt to invoke a feature that's
not supported.
7: I'd like to see the requirements tightened here. Specifically:
href MUST be a LEIRI
type MUST have the syntax of a RFC 2045 media type
media MUST be a Name or comma-separated list of Names
charset MUST be a Name
alternate MUST be "yes" or "no"
or the PI is ignored. Alternately I can live with ignoring just the
bad pseudo-attribute.
8: prefer b.
9: prefer a.
10: reject a, can live with b, prefer c iii or c iv if my version of 7
gets accepted.
11: prefer b iii 2
12: prefer a.
13: prefer a.
14: prefer a.
15: prefer a, can live with c.
--
Go, and never darken my towels again! John Cowan
--Rufus T. Firefly http://ccil.org/~cowan
Received on Tuesday, 23 June 2009 16:46:17 UTC