[ale] CSS hell

[Christopher R. Curzio]

You said I was modifying HTML and adding "extra messy HTML tags" when in
reality I didn't change a single line of HTML. Plus, your unfamiliarity
with CSS has caused you refute your own argument. 

So what if bigcompany.com has lots of stuff in one stylesheet? So what if
they have multiple stylesheets? Contrary to what you believe, stylesheets
aren't downloaded over and over again. In most cases, they're downloaded
once and cached. 

In this example, there aren't any "extra css tags just to implement a list
of URLs." Your suggestion is to wrap that list of URLs in table code that
gets downloaded every single time the page is reloaded, versus a
stylesheet (if called externally, which in this case it is not) downloaded
once and called locally. Your solution ends up being bigger and slower.
And NONE of this "wastes the viewer's time" since it's all transparent to
the user. One large stylesheet can easily make up for the repeated
downloading of html-based layouts over and over. See ESPN and Wired for
examples. 

Also, a browser only calls the styles defined in the html document. It's
not "processing" the entire stylesheet repeatedly for every page that
calls it. If only two or three styles are ever needed by the end user,
only those two or three styles will be processed.

Finally, the CSS validator doesn't test any of this stuff; it only checks
the validity of the syntax and attributes used. Also, my site validates
(CSS and XHTML 1.1) just fine. 

-- 
Christopher R. Curzio     |  Quantum materiae materietur marmota monax
http://www.accipiter.org  |  si marmota monax materiam possit materiari?
:wq!



Thus Spake Jim Popovitch <jimpop at yahoo.com>:
Sat, 15 May 2004 12:13:15 -0400


> For starters, please don't use reply-all.  I am already subscribed to
> the list, I don't need two copies of the same thing.
> 
> All those extra css tags, just to implement an list of urls.  I suppose
> that it has it's merits, but the way most people implement it is a big
> waste the viewer's time.  Consider this: a company creates the website:
> www.bigcompany.com and decide to define all their neat-o wow styles in
> bigcompany.css.  Next month they create a targeted promotion at
> www.bigcompany.com/special-offer and they add a bunch more styles to
> bigcompany.css (yes, you and I know that they should have multiple .css
> files... but the truth is most don't).  Next month they create an
> intranet portal for their employees.... guess where they put the styles
> for the ultra cool fonts used on their intranet?  Now the real waste is
> not necessarily in all those extra bits, it may be in the extra
> processing it takes <insert_favorite_bloated_browser_name_here> to
> repeatedly process (and re-process) bigcompany.css when only 2 or 3
> styles might ever be needed by the end-user.
> 
> If you want to test this, use the css validator on w3.org, so far I've
> only seen Yahoo! pass. :)
> 
> -Jim P.
> 
> On Sat, 2004-05-15 at 11:43, Christopher R. Curzio wrote:
> > Err... What? Are you looking at the same thing I'm viewing? My
> > solution actually removed more data than it added, and I didn't modify
> > a single HTML tag.
> > 
> > What "extra messy" html are you talking about?
> 
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://www.ale.org/mailman/listinfo/ale