|AllTheWeb Validates to W3C Standards.|
Congrats to AllTheWeb [alltheweb.com]. They have recently become the first major crawler based search engine to validate [validator.w3.org] their home page to W3C HTML standards.
As far as I know, this is the first time ever a major search engine has taken such a stand in support of web standards.
There is some work to do on the serps, but given the environment that a search engine lives in, it is understandable to cut some corners to save bandwidth.
[edited by: Brett_Tabke at 1:57 am (utc) on Nov. 13, 2002]
Seeing as Altavista have just had a relaunch of their site, lets hope they can do the same.
Altavista validates? I checked. They don't.
Kudos to AllTheWeb.
One down. Loads to go. Why is this simple thing so difficult?
It isn't a particulary easy thing to do with a very large existing site. In fact, its a bit of a nightmare...
Does this piece of news deserve to be shown on the webmasterworld homepage? Ok, AllTheWeb validates, so what? Does this positively affect the search experience of AllTheWeb users? Do they even see a difference? I don't think so. AllTheWeb has been doing superficial things like a web design contest for its homepage with measly prizes and most recently this: HTML validation. They should better work on their search engine which is losing ground to Google every day.
>> Congrats to AllTheWeb <<
Ooopsies. The link points to [w3.org...] which I think isn't what you intended.
Was it meant to point to [alltheweb.com...] or to their Validation Results [validator.w3.org]?
I haven't got much to say but congratulations!. Although I am not a web designer the few times I have designed webpages I have made them validate. It is not that hard to read the guidelines. I don't think there is an excuse. But none the less, I would like to congratulate AllTheWeb for the great example (to others). (Although SE's have no excuse, if people don't want to have to learn HTML, they can use a standard-compliant editor - how many editors comply with HTML is another matter but IMO non-HTML-compliant HTML editors should be named and shamed as much as possible - having a non-copmliant page is bad but having a non-compliant editor is evil ;-) )
|Kudos to AllTheWeb. |
One down. Loads to go. Why is this simple thing so difficult?
<AOL>Yes</AOL> and <AOL>Yes</AOL>
|It isn't a particulary easy thing to do with a very large existing site. In fact, its a bit of a nightmare... |
More the reason for any webmaster to do it ASAP.
|Does this piece of news deserve to be shown on the webmasterworld homepage? |
Yes, it sets an example. It is a breakthrough because so few do even if the actual event itself may have little effect.
|They should better work on their search engine which is losing ground to Google every day. |
This may be true, but I would hope Google and co. follow their example. I doubt this has "wasted" much time they could have used doing SE algorithms etc. and surely the UI is a big part of a SE?
Validation is so overrated.
I mean what about some of the tags one uses to cater to the 1% who still uses NN? And tags that IE supports, but not other browsers? It's not like it breaks the other browsers, in most cases. As long as my site renders normally in IE, NN, Opera, on Macs, etc, I could care less if it validates.
MARGINHEIGHT & related BODY tags don't validate. I'll continue to use them anyway.
COLOR attribute in HR. An IE invention, but doesn't break common browsers, and hey, if 90% can see a colored bar without me making a table or a pixel graphic, I'm there.
ALT tags. Who really cares if my image doesn't have an ALT tag. Wasted bandwidth to put it in for clear pixels, etc. Aren't we past making pages for LYNX? Seriously.
Look at yahoo, they don't even put attribute values in quotes, and google... they don't have a doctype declaration. Even webmasterworld doesn't make it through the W3C validator becuase character encoding is not specified. Oh no! Those sites are sooo bad... right? whatever. As long as you follow basic HTML rules and don't do gross things, especially ones that affect rendering, 'violations' like I mentioned here aren't really violations, in my opinion.
Working through the W3C validation process made a big difference for my site. I cleaned up some sloppy code and accidentally fixed a rendering error with Mozilla that I had assumed was a browser-specific bug. I'm glad to see more support for html validation.
|Who really cares if my image doesn't have an ALT tag. |
Let's see... Individuals that are blind or have low vision
They rely on screen readers that can only read text. Images and image based navigation systems are completely inaccessible to those that are vision impaired if you don't use ALT tags. People that have turned off image loading
Many people, myself included, browse without images for speed when at home on a dial-up connection and while using wireless internet access. The Search Engines
Many search engines index the text of ALT tags. Why would you waste an opportunity for the legitimate placement of keywords?
Sure, in all of the above an ALT tag for a clear pixel won't be missed, but in many other cases it most certainly will be.
I use alt tags for all those reasons. All i'm saying is it should not be a compliance requirement. There are plenty graphics that do not require alt tags (graphical borders, clear pixels, column separators, combo icon/text links - like yahoo has, etc.)
Sure I could put in alt="" just to comply, but with all the other 'standards' combined together, why even bother. I don't think the masses will become totally compliant any time soon, as indicated by the chart. There are umpteen reasons why the sites in the chart aren't totally compliant, and [for the most part] not because they are poorly coded pages.
Are those missing quotes really hurting yahoo? Will future browsers become ultra-strict forcing them to change all their pages? Doubtful.
There is very little incentive to be compliant, which is probably why they -- and others -- aren't. Sure they validate and check their code, but not for the purpose of meeting W3C.
MARGINHEIGHT & related BODY tags don't validate. I'll continue to use them anyway.
COLOR attribute in HR. An IE invention, but doesn't break common browsers, and hey, if 90% can see a colored bar without me making a table or a pixel graphic, I'm there
can you not use css stylesheets to acheive this?
The reason for validating a website is the same as when writing copy we ensure it conforms to standards by inserting the proper punctuation and grammar in the right places, -to be professional and not misunderstood.
doesnt the fact that a site validates prove that the webmaster is competent?
<Generally rant - no offence intented>
At least standards evangelist has made a choice to support one thing or the other (wether it be Microsoft, w3c or some other specification).
Type-until-it-looks-pretty persons probably give a d*** about how the language arrived to the current state and how it will evolve, but please do not tell the people who does care about it to "shut up", unless you are prepared to suggest a better solution how to develop a language which is free for everybody to use.
Take a stand and by all means defend old coding specifications, browsers or company owned specifications with arguments if you like, but please do not tell people who cares about something that hey are overrated, unless you are ready to sum up some reansonable documented arguments.
</Generally rant - no offence intented>
dkoller> Just a small question out of curriosity. What are basic HTML rules? Who has defined them and where do I read about them? :)
Regarding your question about why graphic borders and other pictures, which only purpose is to be look pretty design elements, should have an alt attrib in the img tag you might consider that the image tag is actually designed for content specific images. All the whistles and bells are meant to be applied using CSS which today is quite possible, so that you can seperate content from style on a rather high level. ;)
Just my 2 cents
Remember the old school seo adage? <i>Creative HTML</i>?
We have no real way of knowing what spiders do and do not support.
I'm convinced 10-20% of the rankings problems people cry about are html validation based (code errors that spiders could see through).
|At least standards evangelist has made a choice to support one thing or the other (wether it be Microsoft, w3c or some other specification). |
What, Microsh*t have some standards? That question was only partly rhetorical - seriously - where are these standards?
|What are basic HTML rules? Who has defined them and where do I read about them? |
Was that question meant to be answered? Well, if so:
[w3.org ] ;-)
or [dmoz.org ] especially [url="http://dmoz.org/Computers/Data_Formats/Markup_Languages/HTML/References/Standards/"]the standards subcategory[/url].
P.S. I have no idea why the above link did not work.
[edited by: joeblakesley at 11:37 pm (utc) on Nov. 13, 2002]
|please do not tell people who cares about something that hey are overrated |
Ok, ok, maybe I should rephrase. I think 100% validation is overrated -- and I stand by that statement. (I never said an individual person was overrated, as you make it sound)
Don't get me wrong, I actually do appreciate W3C making a standard, but in the end it ends up being browser interpretation of the standard that really defines HTML law. IE, Opera, Googlebot (to address what Brett said), etc. dictate more or less what I consider acceptable code. (Whether that's bad or not you decide. I'd tend to say that's bad... but that's how things are. It can lead to bloat of the 'standard' surely...)
Does my site validate? No, it doesn't. Am I happy with my search engine rankings? Yes, I am. Do browsers display my site correctly? Yes, they do. Should I validate? *shrugs* What's in it for me?
|can you not use css stylesheets to acheive this? |
Yes, and I do use them. This is just one example, there are others too, analyzing the sites in the chart reveals some more of them.
|The reason for validating a website is the same as when writing copy we ensure it conforms to standards by inserting the proper punctuation and grammar in the right places, -to be professional and not misunderstood. |
Again, I think people certainly should check code for problems, but not necessarily make 100% compliance a goal. Yahoo, google, etc. have multiple 'errors' but are still professional, high quality sites.
|doesnt the fact that a site validates prove that the webmaster is competent? |
Yes, it does. But I think it would be unfair to say a webmaster is incompetent simply because a site does not validate 100%. Following all the rules can be limiting, in some cases.
>> Look at yahoo, they don't even put attribute values in quotes, and google... they don't have a doctype declaration. Even webmasterworld doesn't make it through the W3C validator becuase character encoding is not specified. <<
.. then they stand a good chance that the site breaks in IE 8 or 9, or so, and will have to be reworked yet again. I'm fed up with the 'recode your site to cater for latest .01 version number increment of the current browser' mentality.
>> Don't get me wrong, I actually do appreciate W3C making a standard, but in the end it ends up being browser interpretation of the standard that really defines HTML law. <<
However, these browser-specific things are not all stated and defined in a formal specification and interact in different ways on different browsers.
>> Does my site validate? No, it doesn't. <<
I hope that you still use a validator on your site, at least to check for code typos, nesting errors, and so on. I'm not bothered if a site uses a few proprietory NN or IE extensions.
[Ah, you did confirm that you check the code. I was still writing when you posted].
[See also: [webmasterworld.com...] ]
Lots of people think "well, if IE and Netscape display my page, then why should I validate? Isn't it a lot of bother for no gain?" Fair enough, if you're sure that when the next major versions of the popular browsers come out your site will still be viewable, then don't lift a finger. Only if you're sure, though.
Now that AllTheWeb validates and uses CSS layout, they can be sure that even people with handhelds and any devices that come out in the future will be able to use the site, as well as new browsers.
As for me, well, my site validates too, so I can be sure I'll have a full head of hair when the next browser versions are released, and my hair won't be turning prematurely grey.
Two quick points:
It is a shame that the title of this thread is not "All The Web Validates to W3C Standards" - now that really would be something to shout from the rooftops about but something that will never happen. I suppose the web is supposed to be a free (as in freedom-of-speech) medium and validation could never be made compulsory anyway, but it is rather like ignoring the or making up your own rules of grammar/spelling/punctuation in RL publications (except this might be excused more as English does vary and is a real language as opposed to a mathematical or prgramming language where correct "grammar" is crucial). Will companies wake up to the fact though? (It should make life easier for them in the long run.)
|Lots of people think "well, if IE and Netscape display my page, then why should I validate? Isn't it a lot of bother for no gain?" Fair enough, if you're sure that when the next major versions of the popular browsers come out your site will still be viewable, then don't lift a finger. |
This seems pretty closed-minded especially as I think very few people at all use MSIE or NN out of choice. IMO, they are both very hoggy et relatively featureless browsers that crash easily and I rarely use either.
Can't use it. Character encoding a site where you have interactive dynamic user generated content from people all over the world using multiple platforms and no less than a dozen browsers isn't possible in the real world. You try to set it one way, and UserX will scream about funny characters. You try it another way, and your text inputs come back in another char set and you have to guess what they are sending. I worked on it for two weeks here and it was nothing but trouble at every corner. I figure to get it even close to right would take two to three months solid work.
Try it some time and see how far you get.
btw: 99% of webmasterworld DOES validate [validator.w3.org].
|they can be sure that even people with handhelds and any devices that come out in the future will be able to use the site, as well as new browsers. |
Don't believe the hype. Check for yourself. This doesn't hold for any of my devices, and doesn't seem to apply to new devices either.
The w3c defines standards that say how things *should* be, not how they actually are. IMHO a lot of folks don't get this.
I've been working on a large, publicly funded site using transitional html and as much styling in css as poss. The aim is to make the site accessible (in every sense of the word) to as many people as possible. Unsurprisingly this means breaking w3c-valid html.
The main thing I've learnt on this project is: support for w3c valid html and css among current user agents is buggy as hell.
You absolutely /cannot/ just w3c-validate and assume your page will work in all browsers. Many of the issues we've had are bizarre rendering quirks from Opera, arguably the most standards compliant browser.
Having everybody mark up html perfectly and every browser agent support w3c standards completely is not enough. You also need every single user agent to be completely bugfree - a somewhat unrealistic expectation ;)