W3C home > Mailing lists > Public > www-style@w3.org > February 1998

Re: W3C Core Styles

From: Todd Fahrner <fahrner@pobox.com>
Date: Sun, 15 Feb 1998 11:28:54 -0800
Message-Id: <v03102806b10cdd2b2a77@[]>
To: Sue Jordan <sjacct@worldnet.att.net>, www-style@w3.org
Thus spake Sue Jordan:
> Given some of the css declarations, I wasn't surprised to see that
> Communicator 4.04 rendered poorly in all of the samples Gayle
> provided. Since, apparently, the decision has been made to exclude
> Netscape via @import, perhaps you should mention that the set of Core
> Style Sheets is presented for Win'95 IE 4.
> If no compromise(1) is possible, given the various buggy
> implementations extant, could you at least consider commenting the
> problematic declarations in the set, so that potential users would be
> aware of the implications of using those declarations?
> Sue Jordan
> (1) A possible compromise might be a set of Core Style Sheets with the
> 'problematic' declarations commented out or 'worked around', when
> possible.

Sorry I could not reply earlier - I've been off the e-grid on business this

The Core styles do indeed look pretty bad in all first-generation CSS
implementations, as well as in MacIE4. They are misinterpreted even in
WinIE4, albeit in subtle ways that tend not to look disastrous.

As a Mac user (bloody but unbroken), I wish as much as anybody that I
didn't need to run a PC emulator to see these sheets reasonably interpreted
on my hardware. If this project appears to favor WinIE4, then that is
strictly a reflection of the relative quality of its CSS implementation - I
needed at least *some* visual reference, and WinIE4 remains the only
reasonable choice at this point.

Make no mistake, I personally believe that portable hypertext markup,
separate from portable style, is key to the continued expansion and
enrichment of the Web, and to the creative and commercial opportunities of
all parties involved. This is why, finally, I would not consider it a favor
to Netscape or to other CSS implementors and authors to tailor these
stylesheets to any incomplete or buggy implementation, even within the
bounds of the CSS1 Recommendation. This would inevitably cloud the intent
of the project, perpetuating today's unfortunate behaviors as "legacy
support issues."

That said, Verso is working on ways to support the styled Web as a
comfortable container of the Windows desktop, rather than as a thing
contained. We hope to develop portable mechanisms to serve only as many
modules and declarations as are adequately supported by the requesting UA,
letting the browser default stylesheet fill in for the gaps. As has been
pointed out, most of the grave problems stem from the "composition" modules
- particularly the "vertical" submodule thereof. This can either simply be
withheld, or authors can modify as desired.


I'll be cleaning up and enhancing the modules over the next many days and
probably weeks, and will report any major developments here. My goal is to
achieve absolute line-by-line formal parallelism across modules of the same
kind, so that only the property *values* will vary.

Todd Fahrner

The trouble with knowing how to do things the wrong way is that your skills
will always be in demand.
Received on Sunday, 15 February 1998 14:24:12 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:26:46 UTC