First, a confession: I'm an HTML Nazi. Second, another confession: I don't like DVB-HTML very much, and don't hold out much hope for it.

Why not? Well, it's linked to my first confession. Given the generally poor quality of the browsers that are currently available in terms of standards compliance, it's difficult to see how a complex specification that uses most of the latest Internet standards (which aren't even implemented in said browsers) will take off.

While there are some good browsers out there, such as my personal favourite Opera, there are too many browsers out there that don't implement the standards at all, or even worse, implement them in a buggy and inconsistent manner. Internet Explorer and Netscape are especially bad at this, but there are plenty of other bad browsers out there as well. Part of the reason for this is that the browser manufacturers have to deal with all kinds of crappy HTML that gets thrown at them, by both 'professional web designers' and amateurs who quite frankly shouldn't be allowed near a keyboard. OK, the professionals won't make mistakes like using a capital letter O instead of the number 0 when defining HTML colours (and yes, I have seen this in real live HTML that's used on a real web site), but there's plenty of abuses that do get committed in the name of design.

There seems to be a habit of designers forcing HTML to do things that it's not intended to to, such as consistent presentation and page layout, and it's easy to see why when the big two browsers have such half-hearted support for the standards like CSS2 that are supposed to make this easier. I've watched a senior Microsoft marketroid claim that they try hard to comply to W3C standards (at the WWW conference, of all places), but it's hard to believe this when their support for these standards sucks as badly as it does.

So please, browser manufacturers, fix your software so that those of us who want to write clean HTML can do. These pages all validate (and if you find one that doesn't please tell me so that I can fix it), but there are still a few gratuitous HTML abuses going on to work around broken or generally crappy CSS implementations.

At the same time, HTML authors don't get of lightly either. So it doesn't quite look how you want it to. Tough. That's the whole point of HTML. I don't care if you can't show your application in the tiny font that you want to use, with gratuitous use of formatting to produce a completely brain-melting layout for your L33T online ramblings (at least my L33T online ramblings try to conform to the standards).

I don't care how much you got paid for it, either: if I can't read it or navigate it, it still sucks. Please use HTML the way it's supposed to be used. And help the browser folks out by not writing crappy HTML. One of the great advantages of HTML is that it can be validated automatically. There's even validators that will do this. The W3C even provide one. So bloody well use it! If there's a tool that lets you find the most boneheaded errors that you've allowed to dribble into your document, use it to fix those errors.

The only way that DVB-HTML is going to take off is if the various browser manufacturers actually implement the standards in a competent way, and if the HTML authors actually write their HTML in a competent way. My past experience, and my cynicism from observing this kind of screw up many times before, lead me to believe that this will only happen if the next set of masters for Netscape get delivered by flying pig. HTML so far seems to be unique in the degree to which everyone who uses the standard can bugger it up so spectacularly. While I hope that doesn't happen with DVB-HTML, I'm not holding my breath.

I'll leave the last word to a post from James Cort in the newsgroup alt.sysadmin.recovery, because he says it better than I can:

From: James Cort
Newsgroups: alt.sysadmin.recovery

Date: Tue, 17 Apr 2001 19:31:41 +0100


Chris Hacking writes
>You know you need recovery when your sunday morning trip 
>to undernet to discuss gaming (#ufintruder), results in a 
>detour into #html, then a flame fest over why #htmls[1] 
>homepage doesnt validate through validator.w3.org
>
>"How come this channels homepage doesnt use w3c legal 
>HTML?
>Because it works fine like this too?"

IMHO a far greater crime than invalid HTML (provided that, 
if it's not valid, it's at least been well tested) is 
using browser-specific code (esp. javascript) for every 
sodding thing on the web site

I'm talking about links, I'm talking about rollovers, I'm 
talking about little bits of code which try to guess your 
browser then throw up errors when it's not 
IE5.5.4235.53254.rand(), then give you a dozen little
windows each saying "This site doesn't work very well 
in Netscape because it's been written for IE".

I'm talking about "designers" who hide behind 
statistics like "80% of people use IE so if we design it
to work in IE 5.5 and break everything else, we'll be OK",
without noticing that of those 80% a significant number 
aren't using 5.5 yet. People who state boldly"The reason
IE is ahead is because it's easier to design for".

And MS thrusting this product onto every single magazine 
cover CD, almost every single ISP CD and all releases of 
their "OS's" for the last four years apparently has 
nothing to do with it.

Dammit, I do *not* use IE, and nothing on God's green
Earth will persuade me to. The day I am forced to 
install IE on this machine an ad goes in the paper to 
sell it and I go into a new profession.

Yes, parts of my own web site are guilty of using
javascript, so perhaps I'm being slightly hypocritical.
But it's been tested in IE, Netscape, Opera, Lynx and 
any other browser I can lay my grubby little hands on.
And the visitor is informed of those parts that 
require javascript *before* they get to them. And for
me it's only a little hobbyist-type site. Is this so
hard for so-called "professional web designers"? 
Well, pardon me Mr. "professional web designer",
but I was under the impression that your client *wanted* 
business.

I needed that.