PROMISE AND PERFORMANCE
<<< By now, we all know that CSS (Cascading Style Sheets) is an
important standard. One that allows designers to specify the font
family, size, margins, and leading of type on the web; permits
designers to create advanced web layouts without abusing HTML;
and enables web designers, HTML practitioners, and programmers to
separate design elements (presentation) from content.
This ability to separate presentation from content further
empowers web designers to create attractive sites without
excluding visitors who cannot use graphical browsers, and paves
the way for the expansion of the web beyond the desktop computer
and onto a variety of hand-held and other Internet-enabled
devices. (More on that world beyond the desktop in a future issue
of ALA).
We also know that many times our best CSS efforts fail in one
browser or another.
CSS doesn't kill websites.
Browsers kill websites.
Even though the CSS Level 1 standard was finalized in 1996, the
first browser to meaningfully support it did not appear until the
year 2000 (Internet Explorer 5, Macintosh Edition). Netscape
Navigator 6, due around Christmas 2000, should also fully support
CSS Level 1. And Opera Software's Opera 4 browser was designed to
comply with CSS-1 and beyond. We have not run tests to verify
Opera's claims, but the browser was designed by the father of
CSS, and we have little reason to doubt that it gets even the
most academic nuances correct.
That's the good news. The bad? IE5/Mac and Opera 4 are used by a
minority, and Navigator 6 is still in beta.
Internet Explorer 5 for Windows, used by the majority, comes
close to perfection but misses the mark in a few critical areas,
and may not completely support Style Sheets before the release of
IE6. And Netscape Navigator 4, still used by tens of millions,
does such a poor job of handling Style Sheets that it has been
known to crash upon encountering them, as previously detailed in
these pages. (The Day the Browser Died)
Faced with these inconsistencies, many web designers have avoided
using CSS altogether. Others have gone ahead and fully exploited
the power of CSS in spite of the fact that it fails in IE3/AOL,
can crash Navigator 4, and delivers wildly inconsistent results
in newer browsers.
Other web designers and developers have followed the "No-Fault
CSS" plan previously outlined in ALA's "Fear of Style Sheets"
series, whether they picked it up here or figured it out on their
own.
Fear of Style Sheets
Fear of Style Sheets 2
Fear of Style Sheets 3
Still others - tricky devils - have created platform and browser
detection scripts to serve a variety of "appropriate" Style
Sheets to specific user agents. For instance, serving one Style
Sheet to an IE4/Mac user and another to a Navigator 4 user on
Windows NT. This approach was always unpleasantly complicated and
almost always added to the bandwidth requirements of each web
page, but at least it used to work. It no longer works. Throw it
away.
What works? Pixels - or no sizing at all. How can we make this
audacious claim? (Assuming The Master List did not
convince you.) Let's look at the alternatives: >>>
- Give Me Pixels or Give Me Death
- Promise vs. Performance
- Font Size Challenges: Points vs. Pixels
- Keywords to the Wise
- Relatives and Percentages
- Wrapup
|