Well, having been involved with The TV Room here's my two-penn'orth.
1. Yes, I agree - the code is rather weighty and font and other style tags being opened and closed repeatedly is a waste of code space, and is very difficult to edit
HOWEVER - and this is a big HOWEVER
2. Websites with huge amounts of content to organise can very often produce code which is not as efficient as it could be. It's called "content management" and sometimes is a lot easier to perform if copy/paste/find/replace methods are used, unless of course you're a multinational publishing company who can spend a million quid or two getting a turnkey system produced.
I think Mike does an excellent job given that it's NOT his life's work, despite what many people obviously seem to think it is.
The problem is much code-generating software is abysmal. I think we're still in the infancy of the HTML world, as many page editors will continually generate something like <TAG VALUE="X"></TAG>Blah<P><TAG VALUE="X">Blah<P> rather than <TAG VALUE="X">Blah<P>Blah<P></TAG> because HTML is, by its very design, a blocky, serial horizontal/vertical wrap and table system. Everything has to be defined as lego blocks that fit together - unlike Flash you can't just say - "put this object in the middle of the screen" - you have to have padding of some sort to get it there.
You could say that it's more code-efficient to remove tags that set the same attribute to the same value, suspend further calls and leave the closing tags to appear only where those attributes then need to take on another value.
That's code efficient. But it's not 100% "compliant" in exactly the manner that the original HyperText Standards intend things to work, i.e. each element of a page is supposed to be coded-up as a discrete object which would appear the same way even if surrounding code chunks were removed. Code-efficient source makes this difficult and many page editors wouldn't, if tags were stripped-down, be able to resolve the necessary retention of attributes either side of a new object, if that new object was to have different values for those attributes. Because if each table cell or paragraph either side hasn't got it's font attributes or style attribute set, the whole thing will fall over because the closing tag to everything BEFORE the new object won't be found, and the original closing tag will be seen as a closing tag to the tags opened when creating the new object. So the necessary formatting will have to be re-applied to everything AFTER the new object too.
Remember too that HTML itself is forever changing - after all, where did CSS come from? I don't think that existed when HTML first came out - it has HAD to evolve because users constantly find themselves coming up to natural barriers in HTML that stop them from doing what they want to do in an easy to code way.
I get sick of people ranting on about HTML standards when, if it wasn't for the focus groups sitting down to talk about the latest round of changes that are going to be made to the standards, the standards now would still be the standards of 10 years ago and we'd all be viewing Ceefax-style web pages in XTerminal, using the cursor keys to move a square cursor over to the hyperlinks which would HAVE to be (because it used to be the standard) surrounded by square brackets and underlined with no choice of font.
