inline CSS - score so far

The score so far.

Reasons to keep style=

+1 Convenience.  People use style= in rapid development, even when their
   final goal is to use CSS rules.  (I do things the same way.)
+2 Populism.  People like style=.  It's easy to grasp, and it gets
   results.  XHTML will be ignored if the WG does not listen to the
   people's demands.
+3 Occasional near-legitimate uses.  For example, when writing *about*
   HTML or CSS, I might use style= in examples.  Another example:  when
   I'm quoting a source that isn't semantically marked, I may want to
   quote its layout rather than interpret that layout, and recast the
   quote with my own semantic tags.
+4 General conservatism.  It's widespread; many pages use it.  (But this
   shouldn't stop the WG from deprecating something, imho.)
+5 Unfairness.  <b>, <i>, and <tt> are in a Text Presentation module.
   (I'd really like to hear an answer to this objection.  I agree with
   it.)


Reasons to drop style=

-1 Redundancy.  It's a redundant feature with ID selectors.  A page that
   uses style= can be converted automatically to use ID selectors
   instead.
-2 Abuse.  The most common uses of style= are abusive; when you see
   style= it is almost always a mistake, if not an appalling
   technological snafu.
-3 Inflexibility.  style= attributes are treated as text/css by default;
   the only way to override this is by using an even more embarrassing
   hack (http-equiv).
-4 Accessibility.  It has been claimed style= has accessibility issues.
-5 Modularization.  It'll be available in the Legacy module.
-6 Forwardthink.  Determination to make XHTML2 a strong semantic
   language, and the need to set the stage for that now.

One more argument that could go either way

=1 Usability.  Dropping style= would make it more difficult to write
   pages with CSS that applies to single individual elements of a page.
   Most web page designs need this because they're using HTML
   as a page description language.  It looks like the HTML WG is trying
   to cut off that usage and steer HTML back into semantic territory.

All these arguments are boring; none is convincing.  Usability combined
with Populism is pretty strong.  But the arguments on the other side
make it seem a real nonissue.


Things that will make Jason scream

- Just one more copy of that Excel 2000 barf appearing in my
  mailbox.  Ever.  Just one.

(*pause*  *scream*  Thanks, Arjun.)

-- 
Jason Orendorff
not affiliated with Microsoft, Opera, the W3C, or the Bavarian Illuminati
considering adding an "MS-XML" spreadsheet to his signature

Received on Wednesday, 23 February 2000 06:33:34 UTC