W3C home > Mailing lists > Public > www-style@w3.org > April 2002

Errata [was: css layout should be symmetrical]

From: Rijk van Geijtenbeek <rijk@iname.com>
Date: Mon, 8 Apr 2002 10:01:04 +0200
Message-ID: <9160593358.20020408100104@iname.com>
To: www-style@w3.org
Hello Ian,

On Monday, April 08, 2002 you wrote:

[..]

>  > Ian Hickson: "For the record, that section of the spec has been rewritten
>  > ..."
>  >
>  > Jesse McCarthy: "You're going to need to rewrite more of that spec to get it
>  > to conform to your revisionist version of history."

> Rewriting the spec (by adding to the errata) does not change history. That is
> why your allegation is misleading and offensive.

The history of changes is crystal clear. But if the rules can be
changed, is CSS2 ever going to be 'final'? You wrote that the WG
considers a CSS 2.1. That would be useful (if it comes before
parts of CSS3 are getting implemented).

Personnally, I have trouble with the Errata containing sometimes very
important changes. Like the infamous underscore in class selectors, or
the redefinition of the calculation of 'height' for underspecified
absolutely positioned elements.

I can't depend on reading the spec itself for interpreting CSS; I have
to look in the Errata if there might have been a change later on. And
not just once; as the Errata are continuously being added to. Wouldn't
it be possible to add a link to Errata entries in the spec itself?

When is a browser CSS2 conformant? When it does what it says, or are
the Errata normative? This means that the rendering engine might have
to be rewritten several times a year, in the worst case. Is Opera
'wrong' in not 'auto-shrinking' the height of absolutely positioned
elements that have a bottom but no top property set?


Greetings,
 Rijk                            mailto:rijk@iname.com

Mot du Jour:
If you think education is expensive, try ignorance.
Received on Monday, 8 April 2002 03:56:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:14 GMT