From: Chris Josephes <>
| The big thing about SPACE that was already pointed out was that it only 
| support measurements in PIXELS.  Thus, Emily Dickenson's greatest works 
| would only have been truely presentable on a seventy-two dot per inch 
| screen with allignment controled by SPACER and MULTICOL.  (Luckily, she 
| stuck to pen and paper).

Well, assuming a pixel is going to be a fixed size on a given device,
you can still assure some kinds of relationships between positions.
It gives the author a more convenient way to do the kind of quadding
you do now with empty images, without having to pre-generate all the
sizes of image you need.  I don't know why they didn't support units
other than pixels...

When you combine this with the new ability to specify attribute values
as JavaScript expressions, you *almost* have enough to do really powerful
layout control, but JavaScript doesn't seem to offer you the kind of
device-description capabilities you would need (for instance, a
rendered-width() method on strings, to determine how many pixels you
need to offset to match the length of a given string or a resolution()
methtod on windows).

| Stylesheets are more flexible because they were designed for true device 
| independence. Work is still in progress over such things as text flow 
| around images, scaling, output formatting, but the stylesheet work is far 
| superior than anything put out by NS lately.

I love stylesheets; I wish they had them in Netscape today.  I also
understand how it might be a major engineering job to re-work things so
they fit.  But even if they did faithfully implement CSS1 today, that
still wouldn't provide the functionality of MULTICOL and would only
partially address the goals of SPACER.


scott preece
motorola/mcg urbana design center	1101 e. university, urbana, il   61801
phone:	217-384-8589			  fax:	217-384-8550
internet mail:

Received on Tuesday, 2 July 1996 16:55:56 UTC