W3C home > Mailing lists > Public > public-css-testsuite@w3.org > December 2011

Computed value of 'border-width: thick': 6px for Opera 11.60, 5px for other browsers

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Sat, 10 Dec 2011 20:03:40 -0800
Message-ID: <75a25231d8e1b9e5da49709b5ad5aba4.squirrel@gtalbot.org>
To: "Public CSS test suite mailing list" <public-css-testsuite@w3.org>

I stumbled on an issue which prevents me from making reliable reftests
in a few tests.

When 'border-width: thick' is declared, Opera 11.60 uses a border-width
of 6px.

OTOH, Firefox 8.0, Chrome 15.0.874.121 and Konqueror 4.7.3 uses a
border-width of 5px. (I'll check later IE8, IE9, Hv3 TKHTML and Amaya

And there is nothing in the spec which prevents or forces or favors
either value.

    A thin border.
    A medium border.
    A thick border.


The interpretation of the first three values depends on the user agent.
The following relationships must hold, however:

'thin' <='medium' <= 'thick'.
8.5.1 Border width

The same situation could happen with a new browser or a not-so-popular
browser and for 'thin', 'medium' or 'thick' keyword values.

The only solution I see is that all Ian Hickson's tests (starting with
) which uses 'thick' as a border-width value would be replaced with

The long-term solution is that CSS3 specify what are the pixel
equivalents to keywords 'thin', 'medium' and 'thick'.

regards, Gérard
Contributions to the CSS 2.1 test suite:

CSS 2.1 Test suite RC6, March 23rd 2011:

CSS 2.1 test suite harness:

Contributing to to CSS 2.1 test suite:
Received on Sunday, 11 December 2011 04:04:21 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:13:23 UTC