[thelist] Browser Detect with JavaScript

[aardvark]

> From: "Aylard JA (James)" <jaylard at equilon.com>
> 
> 	Perhaps I was a *bit* overstated, especially in light of the fact
> that not even IE 5.5 has full support for HTML 4. However, aside from a
> relative handful of elements (<span>, <frameset>, <frame>, and <noframes>
> are about the only ones that readily come to mind), Netscape 4.x supports
> very, very little of the new features introduced in the HTML 4 spec (and
> since frames had been a de facto cross-browser standard long before the W3C
> issued the HTML 4 spec, that's more or less a gimme anyway.)
> 	Netscape 4 *doesn't* support new table elements such as <col>,
> <colgroup>, <thead>, and <tfoot>; accessibility attributes such as title,
> accesskey, and tabindex; form elements such as <button>, <label>, and
> <fieldset>; and it doesn't support <iframe> -- and those are ones more or
> less OTTOMH. There are also a lot of new internationalization features that
> it doesn't support. For all practical purposes, it doesn't support HTML 4.
> Granted, an absolutist argument would rule out all versions of IE except IE
> 5 Mac, but I was intending to make a pragmatic argument, not an absolutist
> argument.

i completely see where you are coming from, but i code many a 
page to the HTML 4.01 transitional spec that works fine in *all* 
browsers, so i don't see an inability to support HTML4...

i guess it's because i don't rely on new tags or attributes that 
control whether or not the page can be viewed.... for instance, i 
used <label> on all my forms (plus many other attributes) because 
it is a great feature for all browsers that support it, but for those 
that don't, there is no loss of functionality... in short, it degrades 
nicely...

so i wouldn't use <button> or <iframe> because it cuts out too 
many browsers... yeah, 70% of my audience uses IE5/wintel, but 
that means i'm still willingly cutting out the users of all other 
browsers... that's too much for me to use those tags...

> 	Agreed. But again, the CSS 1 properties unsupported by IE 4+ are a
> small handful compared to those unsupported by Netscape 4.x, especially when
> bugginess is thrown into the mix. Although here I would say that it would be
> wrong to suggest that either IE 4 or IE 5 is CSS 1 compliant, because there
> are some sizable holes. IE 5.5 closes some of these, but we all know that
> some remain.

agreed, none of them are perfect, but i don't use the elements that 
are unsupported or insanely buggy...

> 	Bottom line: if someone says to me that he or she wants to develop a
> site based on HTML 4 and CSS 1, I would still say that with IE 4+ it can be
> done, with limitations; but with Netscape 4.x, it really can't be done. The
> limitations far exceed the abilities, IMO.

i guess that's where we diverge, because i have many pages that 
validate to HTML 4.01 Transitional and CSS1, but that work in all 
browsers on all platforms... i could provide you with a list to prove 
it, but i think you get my point... i don't use every attribute and 
element just because i can, i only use the ones that degrade 
gracefully...