|
|
Posted by Alan J. Flavell on 04/03/06 18:03
On Sun, 1 Apr 2006, Stan McCann wrote:
> Norman Swartz <swartz@sfu.ca> wrote in
> news:Xns97987A98B280Dswartzsfuca@64.59.144.76:
> > In IE, when one clicks on the font-size selector icon and chooses
> > a display size (e.g. "smallest" or "largest"), the on-screen text
> > remains exactly at 16 points.
Aren't you the lucky one? Hardly any Windows systems are calibrated
to display "exactly" 16 points.
> Due to a bug in IE.
It's hardly a "bug" that a browser conforms to the CSS specification.
In the sense that CSS is designed to be optional, it's certainly
permissible for a browser to ignore the author's sizing specification
(e.g when instructed by some user option) - indeed the web client
accessibility guidelines call for the provision of such an option; but
if the browser *does* implement the author's sizing specification,
then it cannot be a conforming browser unless it implements the
specification correctly. And an absolute size unit is an absolute size
unit: any other interpretation would be non-conforming.
So, in this sense the wretched MSIE is doing its best (modulo the lack
of dpi calibration) to be a conforming implementation; whereas
browsers such as Opera, and the gecko-based family, have decided to be
user-friendly *rather than* conforming.
If there's anything here that qualifies as a "bug", it's in the minds
of those crazy authors who specify absolute size units for a general
web display situation. Absolute size units are there for some good
reason - maybe printing, or some other specialised rendering
situation; but for making web pages for an unknown browsing situation,
they are totally unsuited. *As the CSS specification itself states
clearly to any author who cares to read it*.
> The user *should* be able to increase/decrease font size as needed.
Can't agree - to be a conforming CSS implementation, the user *should*
be able to readily[1] *override* the author's size specification. I
mean: to disable it. To put it another way, the author stylesheet
specification should either be implemented correctly, or not at all.
Re-scaling the font size, when the author has specified it in absolute
units, qualifies as neither "correctly" nor "not at all", and, as
such, is non-conformant to the CSS specification. Anyone is entitled
to have their own opinions about the relative merits of being
user-friendly versus complying with the specification, but calling
specification-conformance a "bug" is surely going too far.
regards
[1] MSIE fails this test by hiding away the override option in a place
where few readers manage to find it, namely on a rather obscure
Accessibility menu. But, other than that, this is one of the few
aspects of MSIE's behaviour that I find myself defending.
Navigation:
[Reply to this message]
|