[lug] Web page test

[Dhruva B. Reddy]

Getting OT, but I would like to respond to this by sharing some lessons
learned from my last web development project...

I have unfortunately found that merely making your site standards
compliant is not enough.  This is because the standards are still not
quite consistently implemented, especially the newer ones (CSS, HTML 4+,
etc).

As an example, I took a (rather complex) template designed for use with
IIS and adapted it.  It had some graphic hyperlinks at the top, which
were designed to fit flushly with a horizontal bar.  We had to put these
into a form and turn them into graphic submit buttons.  It displayed fine
in Mozilla 1.0, but M$ IE raised them.  I ended up hacking the style
sheet and arrived at an unhappy medium between the two browsers.

The least painful way to have an attractive, user-friendly site viewable
on any browser (and they do exist!), is to keep it simple.  This should
be kept in mind when designing the site (this, of course, requires the
designer's buy-in).

-d

On Thu, 19 Dec 2002 at 09:22 -0700, Peter Hutnick soliloquized thusly:
> J. Wayde Allen said:
> >
> > An acquaintance of mine is experimenting with a web tool and has created
> > the following test page:
> >
> >    http://www.nobledead.com/tab-test.asp
> >
> > He is curious how compatible this is with various browsers under
> > different OS's.  If any of you get the chance to check it out, please
> > let me know what OS, distribution, and browser you are using, how it
> > displays, and if clicking on the tabs changes the displayed window pane.
> 
> A nice place to start is with actual HTML.  The page that that script
> generates is pretty far from.  Standards are the only rational way to
> approach cross-product compatibility.  You can't test with every browser,
> but if your page is valid and it doesn't work with a given browser that's
> /their/ bug.
> 
> It is hard to even get the W3C validator (http://validator.w3.org/) to
> even run against your friends page, since it doesn't declare an HTML
> version (which kind of makes sense in light of the fact that it isn't
> HTML) or a character encoding.
> 
> At a guess I tried UTF-8 and XHTML 1.0 transitional, 4.01 transitional and
> 3.2.  It doesn't qualify as any of the three.
> 
> Now, it may be that what he is trying to do is not possible in pure HTML,
> but there is no reason why he can't declare his doctype, encoding, and use
> proper script tags.
> 
> Having said all that, I'll give it a go on Konq and Phoenix when I get home.
> 
> -Peter
> 

-- 
"Advice is what we ask for when we already know the answer but wish
we didn't." Erica Jong (b. 1942); US author