Google

[skinme!]

Has anyone here ever looked at Google's HTML? Urgh. Yuck. Eek. How could such a premiere site on the net have code like that behind it. It's got no CSS except to define the font each element should use. It's got horrid attribute definitions without quotes. It's got a tabular layout. And to find all this out, you have to run HTML Tidy on it, which generates 29 warnings. Just what are the folks at Google on? Did they go for the 'retro' approach or do web standards mean nothing to them?

Furthermore: there isn't even a doctype specified to validate it against. Perhaps the doctype should be "HTML 4 RubbishMarkup". Overriding it with HTML4 Transitional gives this mess of results.

[newmand2]

Well ill probly get flamed for saying this, but do they really need to? Its not like they dont get enough visitors or anything, and they only really have 2 pages so if i were them i wouldnt care. The only thing they offer is a search engine.

[skinme!]

Originally posted by newmand2
and they only really have 2 pages so if i were them i wouldnt care. The only thing they offer is a search engine.

But they don't only have 2 pages. They have hundreds, if you take into account all the other type of search they offer, their information pages, etc. etc.

The idea of Google is to make the Internet easier to use. The idea of web standards is to make the Internet easier to use, especially for those with disability (if standards are used responsibly). They promote the same cause, yet Google completely rejects any standards.

Moreover, it's not a few niggles over their code. Not a small part which is invalid. The entirity of it is invalid. It doesn't validate to HTML4, HTML3.2 or HTML2 and definitely not to XHTML. Their site is totally inaccessible.

Also, it's one of the most heavily accessed sites, yet it could be much more lean in its code if it employed proper HTML and CSS properly. Just think how much bandwidth is wasted sending that mess of code over the net.

[firepages]

Their site is totally inaccessible

totally the opposite.

Whilst your site may get 10 or 20 visits a month with browsers that do not understand say even CSS1 , google gets millions.

despite the banter and fluster about web standards ALL major browsers and most of the minor ones understand straightforward HTML , many differ in interpretation or do not support many of the 'standards' you speak of.

Google wants I assume to be useful to as many users as is possible which is exactly what it does.

Web Standards will be fully supported and eventually be the norm and nothing less acceptable, until then using what works seems perfectly acceptable.

[liorean]

There is another reason for Google to not embrace web standards: Web standards require larger documents. A Google engineer some time back added a 26 character long greeting to their index page - they removed it not long after, after seeing a noticable performance loss in their service.

[me']

Originally posted by liorean
Web standards require larger documents.

But using semantics could make this difference up.

[liorean]

Not if you have Lynx, Mosaic and Netscape 0.9 as your lowest common denominators to which you code for.

[ionsurge]

This is like a few other threads we have had recently. Don't critisize a website if the coder has not requested it.

[whackaxe]

i suggest we all boycott google and ask skinme! to search 4.2 billion web pages for us instead. standards 0wnz0rz.

[MysteryMan]

Re: Google

Originally posted by skinme!
Has anyone here ever looked at Google's HTML? Urgh. Yuck. Eek. How could such a premiere site on the net have code like that behind it. It's got no CSS except to define the font each element should use. It's got horrid attribute definitions without quotes. It's got a tabular layout. And to find all this out, you have to run HTML Tidy on it, which generates 29 warnings. Just what are the folks at Google on? Did they go for the 'retro' approach or do web standards mean nothing to them?

Furthermore: there isn't even a doctype specified to validate it against. Perhaps the doctype should be "HTML 4 RubbishMarkup". Overriding it with HTML4 Transitional gives this mess of results.

Seeing as how google is perhaps the most universally used website in the world - it is THEIR use of code which defines what is good and what isn't.

So to say google's code is bad is to say good code is bad. They define what good code is, because their code leads directly to their success.

If you aren't using the code they use, perhaps you are the bad coder, not them.......

Google doesn't define what is good, they ARE the definition of good.

[brothercake]

It is possible to make an accessible page in the old-fashioned tag-soup stylee. I pride myself that my pages were reasonably accessible and worked in text-browsers and netscape 3 long before I started using CSS or abandoned tables for layout.

But look at Google - look how much is on their front page. Almost nothing. They don't really have that many coding challenges to overcome, because there isn't very much to the site. Even the inner pages like the press center and Zeitgeist have very minimal styling, and very little layout or presentation.

That isn't the norm. Most sites are far more complicated and involved. Making a complicated site throws up loads of new issues that Google will never have encountered - things like how to linearise a 3 or 4 column layout; how to incorporate DHTML, plugins or multi-media content in a semantic and accessible way; how to reconcile an image-heavy design into something that is still useable in screenreaders and text-only browsers.

Most sites benefit from being coded semantically. Very simple sites do not. There is little need to divide and seperate content with semantics, when your page only has one thing on it.


And then there's the flipside - Google does not want the semantic web. Google's success is due, in part, to their ability to extract meaningful information from tag-soup coding. If everyone made semantic accessible pages, they would lose that edge, because writing efficient search algorithms would be significantly easier.

[skinme!]

Why should we not criticise a site if it hasn't asked for it. Should we not criticise government because they want us to be nice to them? Furthermore, I was offering constructive criticism and I think it started an interesting topic.

I still maintain that even if Google doesn't go down the semantic markup avenue, it should at least validate.

[rmedek]

Moreover, it's not a few niggles over their code. Not a small part which is invalid. The entirity of it is invalid. It doesn't validate to HTML4, HTML3.2 or HTML2 and definitely not to XHTML. Their site is totally inaccessible.

Skinme, just what do you mean by inaccessible? All I know is that Google is the one page on the web I can count on to view properly in anything anywhere. Also, just because a page is standard, doesn't mean it's totally accessible. If you don't believe me, just try to view a non-hacked valid CSS page in IE...

-Rich

[Nightfire]

Just because someone uses invalid code doesn't mean their site is inaccesable. If no one could use it, I'm sure google wouldn't still be using the coding they're using now it'll cost them thousands or millions of dollars in loss. But what is the point in this thread though, if the coder(s) aren't even here to provide their side of the story or atleast be here to see what you have to say. WEe could have millions of these threads showing bad code of ppl's sites, but what purpose do they show? It's not like you're going to get the developers to suddenly change their ways by people they have no idea who they are and for 'advise' they never asked for

[oracleguy]

Originally posted by Nightfire
But what is the point in this thread though, if the coder(s) aren't even here to provide their side of the story or atleast be here to see what you have to say. WEe could have millions of these threads showing bad code of ppl's sites, but what purpose do they show?

Exactly.