W3C home > Mailing lists > Public > www-style@w3.org > September 2006

Re: Sizing

From: Axel Dahmen <brille1@hotmail.com>
Date: Mon, 18 Sep 2006 11:30:38 +0200
To: www-style@w3.org
Message-ID: <eelp1k$d3l$1@sea.gmane.org>

> That's correct.  HTML is not a page description language.  Unfortunately
> designers often try to use it as a poor man's PDF, rather than using
> designs that tolerate this variability.

I see you point and I fully agree.. I have the impression, too, that we seem
to be those last "chosen" few implementors to always care for a fully
dynamic design ;)
However, there are occasions where .PDF is not possible (due to download
restrictions) and where HTML *is* to be used for strict layout. There is no
valid argument against using HTML with strict design to layout documents.
Fixed widths are present in the CSS specification and there is no such thing
like "good" and "bad". Anything else is attitude.

> Again, the need for this arises from bad design.  It does frustrate me
> that most pages will not print portrait without clipping and often
> won't print landscape, basically because the designer has forced an
> absolute width and hasn't even considered that anyone might want to print
> (vanishingly rare are pages where the navigation is automatically
> excluded when printed).

Agree with you again. We're talking about the cases of good design here,

> > b)  It's impossible to scale images down. In CSS you can't define that
> > (dynamically loaded) images should display in 50 % of their intrinsic
> > (which is pixels).
> However, you generally know the intrinsic size, so you can do the
> yourself.

Have you considered the amount of calculation to do so? I did this
implementatino already. If you need to get image sizes from database and
scale the page yourself, then you also need to calculate all margins,
paddings and font sizes before sending the page to the client. Not very

And did you consider that browsers are already implementing a quirks way of
scaling, changing font-size? So there is a need for scaling. But instead of
using user agent specific methods, what's left against scaling the whole
document in a standard manner, available from script?

> > > 72pt = 2.54cm by definition, so pt:px and cm:px cannot be defined
> > > independently.
> What I am saying is that once the mapping of pt to px is defined, the
> mapping of cm to px has only one possible value.

Agree. That's exactly what I'm saying. And what's your point again?
Received on Monday, 18 September 2006 09:32:10 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:26 UTC