Standards or not ?

[Mat Colton]

Am Samstag, 25. Januar 2003 05:42 schrieb Andreas Pour:
> Mat Colton wrote:
> > Am Freitag, 24. Januar 2003 22:31 schrieb Andreas Pour:
> > > Hi,
> > >
> > > Just a short point about standards:  you use the term a lot but do not
> > > define it.  What you seem to mean by "standard" is "the draft
> > > specification developed by w3c.org", then you are using:
> >
> > It's a recommendation, not a standard. Since they don't own the web they
> > can't call it a standard.
>
> Exactly my point - except it was being bandied about as a standard and
> something which should be strictly followed simply b/c some big companies
> published it. Don't get me wrong, it may be a good idea to follow certain
> parts of it strictly - but that is a different question from what is a
> standard.  In fact if anything is a standards its IE

Well, I would rather call it a de facto standard. It's not a standard, and 
besides that, it's not even a very good browser. But that's a different 
story. :P

> so if your goal were
> to comply to standards you should just do whatever IE does.  Right?  

No, I do what reaches the most customers. That happens to be sticking to the 
W3C. As I pointed out earlier, it's rather future safe. And it's not a hack. 
Everything else is more or less a hack, since it works around bugs.

>(And
> arguing IE is not a standard is being political, 

We have different views on that. I used to use the IE as dev browser. Back in 
1998 IE 4 was by far the best browser around. This has changed. So for me as 
a developer it is easier to use Mozilla, stick to the recs of the W3C, and 
see what quirks other browsers have when I'm done. It's not political, it's 
about time and money. Of course every web site should run on the IE, but it 
should run on every damn browser. Otherwise the coder did a bad job. Easy as 
that. Don't misunderstand me. I don't expect a modern site to be graphically 
full blown on IE/NS4 as on IE 6, but blocking visitors is bad for the 
business.

>when 90% of people do it
> one way, that *is* a standard, whether you like it or not.)

A de facto standard, not a standard. That's a BIG difference.

> > >   * a completely undemocratic organization
> >
> > So is KDE. So what?
>
> The point is one should not "annoint" it under the banner "standard" and
> slavishly do whatever they say.

Yaaaa, sure, just trying to tease you. :)

> While "slavish" is extreme, this all started out with my comment that
> kde.org should work with Konqueror 3.0 and the response was, in effect, all
> we worry about is complying with this standard. 

Umm, as I stated above, it should run with every browser. Of course this 
includes Kongi. Wierd, who would be so insane and recommend code for the 
kde.org site which doesn't run on Kongi. But making the site XHTML 1.1 and 
run on every browser works. But I must admit that at least NS4 will get a 
text version nowadays. This is were the discussion starts IMHO. 

> If compatability with what
> exists in the wild is also a top priority then I have nothing to say :-), I
> am reacting to the possible view of treating some bureaucracy controlled by
> some huge corporations as being more important than reality.
>
> > >   * an organization which seriously considered making standards subject
> > > to an individual's monopoly (i.e., requiring royalty payments),
> >
> > This is true and sad. But it hasn't happened so far and I hope it won't
> > happen in the future.
> > Oh, BTW, Appsy could be seen as an individual's monopoly. Your Appsy has
> > a great impact on the community. Not meant as a critic, just trying to
> > point it out.
>
> The following essential elements are missing from the analogy :-) :
>
>   * everyone using the Internet having to pay MieTerra by force of law

Hey, cool, can I join MieTerra? :)
But I don't get it, nobody forces you to code to the W3C recs.

>   * anyone saying "do it this way and only this way b/c MieTerra says so"

They not only say so, they also tell you why. It's not totalitarian, nobody 
forces you. Code as you wish. Same for KDE apps. 

> > > If that is your def'n of "standard", then I pass on standards, thank
> > > you very much.
> >
> > Yes, I know, I had a look at Appsy. ;)
> > Ok, I know, it's old. When I look at my code that's as old as Appsy it's
> > poor as well. But you get the idea? It doesn't scale, it's not
> > accessible, it's not printable (in the standard version at least). And,
> > it is quite obviously optimized for 2 browsers. IE and NS.  Thanks to the
> > browsers error tolerance your site works. So go on and ignore the W3C's
> > recommendations...
>
> I don't care about them one bit as a "goal".  What I care about is that it
> works.  

As I first visited Appsy NS 4.5x would crash. every time. So I used IE to get 
the stuff and then reboot to Linux... When visiting with lynx it's "no aalt 
tag hell" and more. So don't act as if it runs on all browsers. I can name a 
dozen browsers which cough on the code. But, hey, it works with IE, isn't 
that great for a GNU/Linux site. :P
No, honestly, the code it old and evil.

>It may be that following the recommendations is what makes it work;
> and that case I will.  It may be otherwise, as it in fact was when the site
> was written; and that I won't.
>
> Still today you can make a site in a way that Konqueor or Mozilla or IE
> won't display it but it complies with the "standards".

Sure, no prob. Never argued on that. I remember the first time I visited the 
CSS test suite and nothing worked... That sucks.

> > But again, it's very old, so what.
>
> Exactly.  Maybe if my house were rebuilt a few nails would not be as
> crooked. But I am lazy, or better I have other things to occupy my time, so
> I leave it as it stands.

Well, I guess it works on most popular browsers. If that's a good site for 
you, then YOU WIN! :)
I really didn't want to start a flame war. But I did want to point out that 
ignoring the W3C recs is as out of touch with reality (and business) as it is 
to blindly follow the recs.
-- 
Bye,
Mat