[thelist] Tables vs Layers

aardvark roselli at earthlink.net
Fri Sep 15 00:32:12 CDT 2000

> From: Stevens, Sharon
> >
> 	no I think the main key point is
> accessibility and tables don't fit
> this mould

now this i heartily disagree with... for instance, how will
a screen reader interpret a table?  how will it interpret
layers?  in both cases, quite well if the code is good... if
i use DW, for instance, i can create my content bottom-up
and then lay out the divs to *display* top to bottom... so
firing it up in Lynx (just an example we can all test) would
result in a page whose content is all over the place... same
goes for tables if i don't consider the order of my cells...
accessibility has less to do with a specific markup style
and more to do with how it's coded...

bad code is bad code, regardless how you do it...

as for your other email:

> 	I hate designing with tables - I'm having a seriously
hard time on
> the contract I'm on at the moment because the company that
created our
> publishing package knew nothing about design and have
designed the site
> utilising tables (and poorly designed ones at that), I'm
not looking forward

using tables in and of itself is not bad design, or even bad
code... poorly-coded tables can be bad, but you only
indicate that their tables are bad *after* you've claimed
tables in general are bad... did they have a reason for
using tables?  did you ask them?  i know i wouldn't change a
handed-off CSS-P design to tables without a damn good
reason, so hopefully they had a reason or two instead of
just personal preference...

> to having to come back here next financial year (when
they've got money to
> pay me again) and converting the tables to useable layers
to fit with
> HTML4.01 strict.

why would you re-code it?  why would you bill the client to
make changes that to them, give them no extra features or
capability?  who's to say they didn't want them converted so
that they could maintain them?  or to address some browser
incompatibilities?  changing the code on their dime to
satisfy your own coding preferences is not good karma...
changing it because it offers tangible benefits, and the
client understands and approves them, is quite alright...

on top of that, why HTML 4.01 Strict?  what advantages does
it offer the client  (not you) or the client's site over
Transitional?  will some feature of the site not work with
transitional?  does the client want strict, or even know
what strict is?  who are the users, anyway?  are any of them
NN users?  or pre 4.x users?  are you aware precisely how
the site degrades through testing?  and if there are
problems, does the client know?

ultimately, it's about designing for your audience, as well
as your client... if your client has a good cross-section of
browsers in their user base, and their identity and page
design are important to them, why not use tables to support
the older browsers and guarantee more predictable rendering
for all?

i have no problems with CSS-P as a standard, but without
full or correct support, i have a problem implementing it...
given a captive audience (intranet, for example), i'm all
for it...

More information about the thelist mailing list