Posted by Alan J. Flavell on 02/01/18 11:47

On Tue, 2 May 2006, Jonathan N. Little wrote:

> CSS works quite well on the whole, most times it just improperly
> applied.

I think the worst examples are from those kind of folk who take some
existing bag of parts that has been scrabbled-together with HTML/3.2
+/- proprietary extensions, and then insist on trying to translate
that mess, piece by piece, into CSS.

Typically, the results show all of the disadvantages of presentational
quasi-HTML - combined with the disbenefits of the "if only the web
could be DTP" deezyner mindset; and none of the benefits of aiming for
a relatively presentation-neutral HTML, with one or more stylesheets
aiming for flexible presentations on various classes of device.

I won't for a moment pretend that HTML and CSS are absolute ideals for
their purpose; but they're what we've got, and they /are/ far more
serviceable than the above exercises tend to indicate.

IMNSHO, things go much better if one starts from first principles with
a CSS-based design; explores what can be done flexibly and resiliently
in CSS, as against what tends to break at the slightest sight of a
variation in browsing situation, and then designs accordingly. And
puts out of one's mind these eternal calls for translating old (bad)
constructs by rote into new.

IMO and YMMV, natch.

[Back to original message]