Re: [CSS21] Please endorse xml:id

----- Original Message ----- 
From: "J. King" <>
To: "Andrew Fedoniouk" <>
Cc: "www-style" <>
Sent: Sunday, June 26, 2005 6:41 AM
Subject: Re: [CSS21] Please endorse xml:id

| On Sun, 26 Jun 2005 01:22:51 -0400, Andrew Fedoniouk
| <> wrote:
| > CSS also asumes that this DOM represents "endless tape" coming
| > incrementally.
| > Consequence: famous vertical alignment.
| I hadn't considered that one.
| > CSS asumes that this DOM is always placed in some view having known
| > dimensions and capable to set flags :hover
| > :active :focus :link , etc. to its elements.
| Can you demonstrate a case where one would want to (or, for that matter
| could) render content onto a canvas of -unknown- dimensions?

To be precise in CSS 2.1 canvas defined as
 "The canvas is infinite for each dimension of the space,
but rendering generally occurs within a finite region of the canvas...

And real example:
In application
each note is HTML document. EverNote canvas (The Tape) has unlimited
vertical dimension. In some cases it has also
unlimited width. Therefore e.g. root { width:100%; height:100%  }
is not applicable.

| Interactive pseudo-classes aren't limited to HTML or XML, especially since
| they aren't even reflective of the document tree.  I suppose :link -was-
| specified with HTML in mind, but UAs for document types that don't have
| links can still be interoperable in every other respect---and
| interoperable with each other in every way.
| I think you're just splitting hairs here, Andrew.  The conformance

:) Yes, it seems so.

| requirements do state that lack of a feature in a UA due to technical
| limitations of a platform does not make it non-conformant.  Granted, they
| don't say anything about limitations of the document type, but that seems
| the implication.  Perhaps they should state this explicitly, though?

Andrew Fedoniouk.

Received on Monday, 27 June 2005 03:52:52 UTC