- From: Ian Hickson <ian@hixie.ch>
- Date: Wed, 25 Feb 2004 10:46:03 +0000 (UTC)
- To: Etan Wexler <ewexler@stickdog.com>
- Cc: www-style@w3.org
On Tue, 24 Feb 2004, Etan Wexler wrote:
>
> Interoperable practice at this point in time seems to be to accept the
> mixed <number> and <percentage> terms. Safari 1.0, iCab 2.9.7, and
> Internet Explorer Macintosh Edition 5.2 each accept the mixture.
> Mozilla 1.6b rejects the mixture.
Opera also rejects it, so there are two interoperable implementations.
Generally, the WG didn't want to unnecessarily change things from CSS1.
When we do we tend to be chastised for continuously changing the rules.
> Additionally, a Hicksonian realism predicts that authors will make this
> mistake and concludes that the intent is clear, so there is no sense in
> punishing end users with a rejection of the value.
The "Hicksonian realism" (or is that "Hickson Ian realism"?) is about
making sure that specs specify interoperable error handling behaviour, and
making sure that the handling is not draconian ("entire document must be
perfect or you lose"), since if it is draconian you are likely to end up
with a faulty implementation gaining market share and forcing other
implementations to follow suit in not being draconian.
At the level of individual constructs, there's no big deal. I agree it
would have been easy to define the mixed case here, but for whatever
reason CSS1 didn't, and changing it now would just penalise the UAs who
got it right, making them unhappy with the spec "continuously changing
the rules" and making them less likely to want to write compliant code in
future because "what would be the point, the rules will just change in a
few months anyway".
--
Ian Hickson )\._.,--....,'``. fL
U+1047E /, _.. \ _\ ;`._ ,.
http://index.hixie.ch/ `._.-(,_..'--(,_..'`-.;.'
Received on Wednesday, 25 February 2004 05:46:08 UTC