[whatwg] HTML syntax: shortcuts for 'id' and 'class' attributes

[Andrew Fedoniouk]

----- Original Message ----- 
From: "Robert Sayre" <sayrer at gmail.com>
To: "Andrew Fedoniouk" <news at terrainformatica.com>
Cc: "Boris Zbarsky" <bzbarsky at mit.edu>; <whatwg at lists.whatwg.org>
Sent: Thursday, November 30, 2006 10:15 PM
Subject: Re: [whatwg] HTML syntax: shortcuts for 'id' and 'class' attributes


> On 12/1/06, Andrew Fedoniouk <news at terrainformatica.com> wrote:
>>
>> Can this be considered as enough backward compatible ?
>>
>
> No. Try this rule of thumb: don't invent anything unless you
> absolutely have to, or you're giving a name to something that is
> already happening.
>
> The suggestions above are basically personal preference. I might pick
> your method if I were inventing the Web, but TBL beat us to the punch.
> Now we're stuck.
>
> There are billions and billions and billions of web pages out there,
> and most of them work in all common browsers. Trying to change them
> into something else is like serving tea in front of a tidal wave.
>

Robert, I pretty much agree with your arguments in general.

But!

Let's imagine that there are no such things as HTML5 and WHATWG yet.
Only HTML 4.1, CSS and JavaScript in the wild.

And here comes someone who will tell us: "Hey, something wrong in this 
triade -
it is not serving needs of Web Applications well. So let's start from HTML."

"No. Try this rule of thumb: don't invent anything unless you
absolutely have to, or you're giving a name to something that is
already happening."

Absolutely applicable! Isn't it?

In other words: what is so conceptually wrong with HTML 4.01 that
requires HTML5 to be designed?

Discussion on whatwg list is puddling about nuances of attribute use,
"/>" versus ">" and  the like "problems".  Why?

Really, with some minor exceptions, HTML 4.01 is good enough.

Real nightmare in modern Web technologies is that horrible way
of attaching scripting  and other methods of "activation" of Web pages.

JavaScript and its ugly prototype model is just... you know what.
And those "cool" onclick, onmousedown and co. attributes in HTML...
Archtecture of scripting is what need to be redesigned in the very first 
order.
That really should happen even yesterday.

CSS is also something....
In 10 or so years it failed to provide vertical alignment
so needed for Web Aplications. Instead people are adding
attributes in that flat table of CSS styles. There are 120 or so
attributes already in CSS. Who and when will stop this?
They say that CSS now is "modular" because specification
is splitted in 10 files instead of one. This is very strange meaning
of word "modular" to be honest.

Probably solution could be in creation of
"Open HTML/CSS/Script specification"
that will make conditions for competition of various
approaches/technologies. Who knows?

I do know that real evolution of  Web technologies
happend when IE4 was competing with NS4 - decisions made at that time
work now. I cannot see anything that changed technology
dramatically since then.

---------------
And yet... reading this:

<cite src="http://blog.whatwg.org/faq/">

  "Why improve HTML?
   Because that's what most authors are using and all browsers support."

</cite>

Something wrong with the logic of the answer. Robert, what do you think?

Andrew Fedoniouk.
http://terrainformatica.com