Web Standards

[stevemac]

I am looking to create an on-line business sometime before the end of this year.

I would describe my XHTML & CSS skills as amateur & have virtually no PHP skills. I have played around with OScommerce - copied & pasted mods, etc but that is as far as I've got. I'm not really too happy with the results.

During the last few months, I've been reading a few programming related books (XHTML/CSS). The main emphasis seems to be on producing 'standards compliant' code. With this in mind, I began looking for a possible candidate to design & code a website for me. Essentially, If I'm going to pay somebody to code a site for me - i would like it done properly.

My point is that I'm astonished at the amount of web designers who's own 'showcase' websites fail W3C validation. Very few display the W3C icons on their site to indicate that the code has passed validation.

.

[FireFleur]

W3C validation is all well and good, but W3C hasn't produced a browser in ages :)

It is something people tend be aware of, but the cut off for compliance is the target set of browsers consumers use.

With that said coding to standards nowadays does make a lot of sense, the browser wars have cooled down, and standard code tends to render better and be more extendable.

But there are bigger things to get right first rather than starting with strict W3C standards :)

[FireFleur]

Just as an example, and this was the first page I found on W3C

http://validator.w3.org/check?uri=http%3A%2F%2Fwww.w3.org%2FTR%2Fxhtml1%2F&charset=(detect+automatically)&doctype=Inline&group=0

That shows one warning.

Thankfully most of them appear to validate with no problems, but it is that simple to get out of full compliance.

[cycloneuk]

I have never bothered trying to get W3C validation, waste of time imo. Just look at all the big boys like Amazon and ebay, they get hundreads of errors and don't seem bothered.

[cmcp]

Good for you Stevemac I agree about the importance of web standards.

It's not about complying with some validator, it's about accessibility and consideration for all your users.

[FireFleur]

Well it is about complying with a validator in stevemac's instance.

As to accessibility well, people should be designing stuff that works for the blind, and colour blind, and degrades gracefully to text only, but hardly any do.

There are so many things, that I can assure you no one site has it all done correctly, there is always something to find wrong in a site. And a successful site can be successful for many reasons, but W3C compliance is hardly ever it.

It is useful to code to W3C and know when you are breaking the rules, primarily because XHTML is an XML derived language and the XSLT tools work better if the code is compliant, but that does mean supplying a charset, generally UTF-8, and a schema or dtd.

Firefox validates as it goes with XHTML, and the page will break if you are not compliant, with lovely error messages. And it could be interesting to see if that causes a problem if the W3C XHTML schema ever disappeared or changed slightly, you could DoS the web in a blink. For my own sites I use various doctypes, but for commercial sites making sure the majority of the users (not all) have access to the site is more important, and that is why you see transitional, and again the message is more important than being W3C compliant, so if the design warrants it some non compliant code can be used to make it work, often that is browser specific.

Reality and pragmatism trumps some notion of standards when it comes to commercial.

[DambTech]

W3c is the Basic standard that all websites should meet, if website design companies are not going to abide by W3C then when things like Standards are mentioned to customers they think they are being ripped off. (Even though it does not cost any extra whilst the site is being built.)

For website standards to be taken seriously then Web companies have to start sticking to them, otherwise there will be even more junk and clutter out there that looks like it has being put together by Muppets.

[stevemac]

I accept that websites need to be well designed, usability, etc - this has nothing to do with whether a site passes validation. Sites can & many do use different stylesheets for different devices. Indeed, many sites use seperate stylesheets for different browsers to get around some of the older IE issues - e.g.


<!--[if IE]>
<link rel="stylesheet" type="text/css" href="/stylesheets/ie.css" />
<![endif]-->
<!--[if IE 6]>
<link rel="stylesheet" type="text/css" href="/stylesheets/ie6.css" />
<![endif]-->



The most common problem that I've observed is the use of depreciated tags e.g. pages with a DTD of 'XHTML 1.0 strict' using depreciated tags that would have passed validation had the DTD been set to transitional instead of strict.

Second most common is closing tags that weren't open in the first place - possibly caused when updating existing pages.

These are simple errors that would take just a few minutes to correct!

[MartCactus]

Well it is about complying with a validator in stevemac's instance.

As to accessibility well, people should be designing stuff that works for the blind, and colour blind, and degrades gracefully to text only, but hardly any do.


Whilst of course local council sites, the health service, banking etc should be available to people who are blind, I do think some of the obsession with this is a bit over the top.

Just slavishly following rules is also dangerous. I spoke with a client who'd built her site (related to volunteer sector) to comply with standards on accessibility. She then tested the site on a blind group she knew who were largely using screen readers. One thing they objected to was the practice (recommended in accesibility guidelines) or putting text filled into text field, that disappear when you click with the house on that field eg "firstname". They said that the screen reader doesn't require this if the field is probably marked up and labelled "first name". So on a screen reader annoyingly it reads it out twice.

Further investigation shows that the RNIB issue different guidelines to the "official" standards. Its best to aim for real usuability and accesibility rather than just following guidelines from a validator.

Also for some sites its just not really that necessary to have access for blind people. We've had clients selling adult videos query accessibility. Perhaps they think their clients might lose their eyesight over too much "use" of the videos?

[FireFleur]

Yes, the 'standards' compete against each other, because no one really has the standard.

In the case of the text to speech application I would argue that should be controlled at the application level, and the ability to disable pronunciation of default text in form fields is something that should be given.

The W3C uses the term guideline, and they know it is not that important and often will try and adapt to what is required, they are looking to build pragmatism into the guideline, not dogged acceptance. Eventually the W3C guidelines should work well, but IE is the constant sore point, Mozilla have had their moments as well.

Quirks mode coding is the opposite of a standard ideal, you are in the camp of pragmatism there, switching commented ifs is not part of any W3C guideline, it is an IE thing.

Ironically most who really believe in standards all the way through, would not be using PHP or ASP, again those langauges themselves are in a 'pragmatic' camp, they are not rigourous in their definition, and really designed as tools to allow simple tasks to be done.

Now the muppet comment, it is really hard when someone does that, and then you go and have a look at their site and it is not W3C valid, but I will resist from mentioning it. I will just generally say, that keeping to some idea of standards does cost and to get the validation will take an extra few hours or so, or possibly break a layout and take a lot longer than that.

[bluestagstudio]

Just to add my two cents..

Following standards doesn't guarantee success but disregarding standards altogether will result in almost certain failure.

Therefore I tend to follow the 'do everything by the book unless there's a reason not to' rule... designing a site specifically for the blind, as mentioned before, is a perfect example of when not to follow the rules.

It always amazes me at the number of website "developers" who have a website that doesn't work in IE6. Dread to think what their clients would do if they knew that their website might look a complete mess for 18% of it's viewers.

Choosing to ignore standards for the benefit of the end user is good, otherwise why do it?!

[/RANT]

[FireFleur]

Go check your portfolio page.

http://validator.w3.org/check?uri=http%3A%2F%2Fwww.bluestagstudio.co.uk%2F portfolio.php&charset=(detect+automatically)&doctype=Inline&group=0

Defining the type in a script tag is required even in XHTML transitional.

[sirearl]

Just to add my two cents..

Following standards doesn't guarantee success but disregarding standards altogether will result in almost certain failure.


[/RANT]

Utter rubbish as long as a site displays correctly on a browser and the selling procedure is right .

I never bother with web standards as they tend to get in the way of success.

Earl

[bluestagstudio]

Yea, none of the pages apart from the homepage validate yet, we wanted to get the live asap and validating or not validating I knew it looked good in all browsers and work better for us than the old one and in my mind that was a good enough reason to delay doing the sub pages.

The other pages have dropped further down the to-do list as we've been quite busy lately.

Utter rubbish as long as a site displays correctly on a browser and the selling procedure is right

On a browser? Probably better if it worked on any browser. And not just existing browsers, future browsers too. When firefox first came out loads of sites 'broke' because they didn't follow the rules and firefox did, but because they worked in IE at the time some people ignored the standards as they 'got in the way'.

[FireFleur]

It is what works.

So you generally code to what works in the majority of browsers, if the standards don't work in a browser with a big user base you ignore that 'standard'.

If your business relies on an XML toolchain with XSLT transformations then you have to keep to the guidelines, but again if an implementation is not done correct, you again ignore the standards and code to the implementation.

The whole thing is really nothing to do with consumers, it is all about the developers in the chain, and the consumer doesn't have to worry about it, but developers do. So, a guideline is produced by W3C the developer thinks hmm I will produce some tools, now they go off and implement, and if there is a problem with the guidelines they will make the implementation work and ignore the guideline. Normally this is the flagged back to the guideline creator, who then has to work out if the implementation is wrong or the guideline is.

If you are poducing for the consumer, then you have to produce something that works, not someting that meets the gudielines. It is bizarre to think that by coding to the guidelines that will be enough, you don't start there, you start with the browser and make it work there.

Then you go make it standard compliant, or do you, because that is an extra cost. You have something that works, why add guideline compliancy?

Some do code from standards out, and that is a method, but it is slower than just putting in the basics to make it work.

There is nothing wrong with wanting standard validated code, but you will pay extra for it, and it will take longer, and at times you will have to compromise effect for standards, so there are negatives.

[FireFleur]

Yeah that is what you have done, you wanted to get the site out first, and working, nothing wrong with that, that is common business sense, speed to market.

Then later you may go and make it all compliant, but it will take time, and you will aim to do it in downtime.

So pragmatism generally wins.

[sirearl]

On a browser? Probably better if it worked on any browser. And not just existing browsers, future browsers too. When firefox first came out loads of sites 'broke' because they didn't follow the rules and firefox did, but because they worked in IE at the time some people ignored the standards as they 'got in the way'.

Well seems a bit of a waste of time trying to get compliant as the search engines don't give a hoot,and how else are you going to get an online business to be a success without playing footsie with them.:D

Putting a 200 page non compliant site up in a week will give you a much bigger head start than 10 immaculately conceived creations.:)

Earl

[stevemac]

Possibly I chose an unfortunate title for this thread, maybe I should have called it W3C validation instead. I can now see that 'web standards' means something slightly different to W3C validation.

My original observation was refering to sites belonging to web developers themselves & not necessarily to the sites that they construct for their customers. These are showcase sites that effectively say to the world 'Look at my site - this is the best that I can do - I could build one for you too'. It was also refering to what I would view as coding errors & not accessibilty issues, etc. I would have thought that web developers would have worn the 'W3C validated' icons on thier sites with pride & used this as a selling tool - maybe even to justify higher development charges. As I wrote in an earlier post, some do & I even managed to find 1 (UK based) web developer who now codes his site in valid HTML5 - the W3C don't even have any 'validated' icons available for that yet.

[quikshop]

Coding standards are shifting sands at best. Each iteration of the W3C guidelines and competing browser versions introduce new 'best practice'.

The truth is, W3C compliance is a red herring. As long as your website is useable for the audience you intend it for and can be rendered in the most popular browsers what more would you need?

Google pays no attention to coding standards and rightly so. Its own website fails on various 'standards'. Its easy to get hung up on a coding badge of honour which means very little in terms of SEO and conversion rates, and even less to your own customers.

[PoshCloche]

Also consider that if you get your web site built to a recognised standard it will be easier for another web agency to build or add on to it if you choose to change agency in the future.

[stevemac]

Also consider that if you get your web site built to a recognised standard it will be easier for another web agency to build or add on to it if you choose to change agency in the future.
Exactly!
.

[FireFleur]

W3C validation doesn't ensure a particular website will be easy to update.

It just means that the markup used validates against the DTD or Scheme supplied.

My advice would be to see if you can hire whoever wrote the tool you are using to work out if a site is valid or not, that to me would seem the most logical thing, but I bet you didn't think about that :)

[Esk247]

W3C Validation only checks to see if there are little bits of code missing e.g. /> not cloing properely or missing the initial part of a webpage stating what format it is in e.g. html or php

sometimes though it can kick out bits of coding and make additional /> appear and then you still have to go round and fix them when they appear or push divs out.

websites should really be made to please the visitor though..as said in some other post..they should be easily readable by partially sighted people and have a range of text sizes if they are a text based site..they should also be realllyyyy easy to navigate! and include bookmarks if they are information based and quite large in size. they also shouldn't scroll on too long..e.g. for 200 odd pages! nobody is going to read to the very end.

[stevemac]

My advice would be to see if you can hire whoever wrote the tool you are using to work out if a site is valid or not, that to me would seem the most logical thing, but I bet you didn't think about that :)
I do hope that hint of sarcasm wasn't aimed at me?

FYI - I am quite capable of writing my own (perfectly valid) XHTML 1.0 & CSS/CSS2. As I stated earlier, my search for a web designer began because I had a requirement to integrate PHP - a language that I have very little knowlege of. During my search, I did find several web designers who were able to produce very good looking work & also write perfectly valid XHTML code.
The purpose of this thread was to express my surprise at the amount that either didn't, couldn't or wouldn't. I certainly didn't expect to see some of the replies that have been posted.

[barginboyrob]

Try this

http://browsershots.org/


Haven't used it myself, but it might help finding out if your site shows properly in all browsers

[awebapart.com]

When considering web standards here are 10 things to think about:

1. There has been an old joke going around the IT industry for years, and many other industries where standards are discussed, and it goes something like this:

"The great thing about standards is that there are so many to choose from."

The point of the joke is that a standard suggests only one standard way of doing things, so having a choice of standards doesn't really make sense. This is true with the web standards out there too, do you go for a HTML standard or an XHTML standard, which of the many versions, etc.

2. However many standards there might be, they are really only two types of standards to be aware of, and to understand the differences between:

a. de Jure (http://en.wikipedia.org/wiki/De_jure) standards - these are standards made by either law or committee, and in this case of web standards, these are standards, or guidelines, provided by W3C

b. de facto (http://en.wikipedia.org/wiki/De_facto) standards - these are the actual standards that result from market forces and practical requirements, and in the case of web standards, these are an amalgamation of what does and doesn't work in the popular browsers out there

3. In the web world, due to different implementations of different browsers, and different interpretations of the de jure standards by different browser manufacturers, implementing a website to a simple de jure standard can turn into a very complex de facto standard implementation due to the various quirks of different web browsers. In some cases there might not be a de facto standard equivalent of some de jure standard feature, or a non standard "doing things differently for different browsers" de facto standard way of implementing a de jure standard feature (which then begs the question is it really standard at all)

4. In the web world, de facto standards will always be behind de jure standards, browser market share and browser implementations playing either the catch-up game or the backwards compatibility game dictates this situation

5. Tests like Acid test 2 (http://en.wikipedia.org/wiki/Acid2) and Acid test 3 (http://en.wikipedia.org/wiki/Acid3) show the difference between a de jure standard, and how standard compliant the browsers actually are. More importantly, sites like quirksmode.org (http://www.quirksmode.org) also show whether different browsers actually support different areas of standards

6. To emphasise the difference between de jure and de facto standards, it is possible to code a website to the very latest de jure W3C standards, e.g. latest version of HTML, XHTML or CSS, and for it not to work (or look broken) in the most popular browsers that are in common use today. So although a more standards compliant website sounds like it might support more browsers, it might actually mean that it supports less browsers. Sometimes making a website too standards compliant doesn't always make it more future proof either, as was noted with the IE7 release (http://www.theregister.co.uk/2006/10/31/ie7_causes_site_problems/).

7. It gets even more complicated when not just taking into account browser capabilities, but also taking into account email reader capabilities for HTML email, which are usually a lot further behind than the browsers.

8. Because of all these issues, there are many different approaches that web developers take to creating websites to work around these issues, all of these approaches might be standards compliant or not, but they may also be standard or non standard depending on what your definition of "standard" is. The only way to write one standard piece of code that presents consistently on many targets, is to use a lowest common denominator approach, using an old de jure standard which is old enough to have finally been adopted by most of the browser and HTML renderer implementations. As soon as you start doing things differently for different browsers you do not have one standard piece of code any more, so although it might be standards compliant it is isn't one standard code base. An alternative graceful degradation approach, which involves coding to a higher common denominator, can mean that the site in older browsers does not render in a consistent way, in which case you do not have a standard website presentation across different browsers.

9. With modern websites, the presentation layer, or what gets sent to the browser, the client-side code, the HTML, the XHTML, the CSS, is merely the tip of the iceberg of the total amount of code in the system, the server-side code, the PHP, ASP.net, SQL code, (even the client-side javascript AJAX code) etc, so having a W3C web standards compliant website is no indication of the overall quality of the code, or the quality of the system architecture, underneath it. In some cases it might be a good indication, in other cases it might be a warning sign that the developer has focused too much on one small area and not spent the time on other more important areas.

10. Create your website for your intended audience. If you create a site for people wanting to buy designer clothes, the people visiting your site wont be interested, and will even get confused if you have "blind them with science" icons like "W3C HTML v4.01" or "W3C CSS vX.x" on your home page. Placing links such as these on your home page can also introduce pagerank leakage (http://www.ukbusinessforums.co.uk/forums/showthread.php?p=861427#post861427), which goes against guidelines, best practices, or the "de facto standards", of implementing search engine optimisaton (SEO) on your site.

[J-Wholesale]

W3C validation won't get you any new customers. It won't increase revenue. A visually appealing site that is easy to navigate, and good content for search engines and visitors, WILL help win you new customers, and WILL increase revenue.

Which of the two should you invest almost all of your time in? I couldn't care less how many W3C validation errors our site generates as long as it keeps bringing in new business.

[FireFleur]

It was aimed at you and I am not being sarcastic.

I think you want someone into strict validation to build a site for you. And if someone is going to create a browser plugin in to do that then that is the person for you.

Or as it turns out yourself :)

Well you use a template system like pear and phplib, probably the former, then you write all your lovely strictly validated code, and the php guy just makes reference to the templates you have created, voila.

That is the way to do it by the way, there is no sarcasm that is the solution most places use - though normally the server side guys gets to comment on the unstrict validation of the templating code, but in this instance it won't be a problem.

Do tell us the site, I will set a cron up to visit it a year after release, and see if it is still strict :)

[stevemac]

Andy Budd & the rest of the team at Clearleft Ltd even manage to code their 'showcase' website in perfectly valid HTML5 - a standard so new that the W3C haven't even produced any icons for it yet!!

Several web designers (registered here on this forum) are also able to & do produce code which passes W3C validation. Some of which, I know are watching this thread with great amusement!

[stevemac]

While I'm on the subject - have any of you ever heard of dabs.com?

[FireFleur]

Well prepared to be shocked again, nay horrified, stevemac.

Run your validator over their sites and their client sites and watch the errors and warnings tumble in.

It is anarchy out there stevemac I for one salute you and your quest for W3C validation, Godpseed stevemac.

May your opening brackets be less than and your closing braces be greater, let no rogue browser specific tag pass your frenetic whizz kid markup skillz, much lolz stevemac, referer my man, keep it valid :)

[edmondscommerce]

it all reminds me of that zen proverb about the finger pointing at the moon...

standards are great, but a great web site that doesnt meet the standards is infinitely better then a rubbish website that is perfectly standards compliant.

search (http://www.google.co.uk/search?hl=en&rlz=1B5_____enGB325GB326&q=zen+finger+not+the+moon&btnG=Search&meta=)

[awebapart.com]

While I'm on the subject - have any of you ever heard of dabs.com?
The dabs.com home page is an excellent example of a page which is W3C compliant, de jure (http://www.ukbusinessforums.co.uk/forums/showthread.php?p=868557#post868557) standard compliant, but does not render correctly in different browsers, so it is not de facto standard compliant. At a quick glance, the rendering problems include the "Products that you may be interested in" section at the bottom going outside the box container at 1024 width on Firefox 2 and Chrome, and the right column containing the basket and other ads going half-way off-screen in IE6 at 1024 width. Some of these issues might also be related to their choice of a fluid web design width (http://www.awebapart.com/home/faq/Web_Design_Width_and_Web_Screen_Size) as well as lack of thorough browser testing.

Does any of this detract from dabs being a good site and good company overall? a site to buy from? not for me. I don't care what the underlying code for dabs looks like or whether it is compliant or not, or even whether there are some rendering issues in some browsers, because I buy from dabs because of the prices, availability and the service.

Andy Budd & the rest of the team at Clearleft Ltd even manage to code their 'showcase' website in perfectly valid HTML5 - a standard so new that the W3C haven't even produced any icons for it yet!!
Clearleft should be applauded in doing a good job of their website, not so much because it is HTML5 compliant, de jure standard compliant, but because first and foremost they have taken the time to test it and get it working in different browsers, including IE6, to make it de facto standard compliant. Whilst the site is HTML5 compliant, it is not coded purely for HTML5, but with common browsers in mind too, which makes it a HTML4 and HTML5 website.

The dabs.com and clearleft sites are also good examples of "the great thing about standards is that there are so many to choose from", because dabs.com has gone the XHTML route whereas clearleft have gone the HTML5 route. XHTML verses HTML5 (http://mezzoblue.com/archives/2009/04/20/switched/) is a hotly debated topic at the moment.

Whilst I don't care whether dabs.com is compliant or not, I am happy to see and would expect to see a compliant website from a web design company like clearleft, especially since clearleft are at the level where they are giving paid for talks, seminars, presentations to other web designers. Last year I was less than impressed to see that a good percentage of the speakers at the "Future of Web Design" seminar had websites that were not standards compliant (http://www.ukbusinessforums.co.uk/forums/showthread.php?p=441306).

The clearleft guys are clever guys, but there is also a risk that you can be too clever sometimes, and similar approaches that follow the latest standards too closely and introduce hacks or workarounds for other non-compliant browsers, have in the past fallen over when new browsers have entered the market, as was the case with the IE7 release (http://www.theregister.co.uk/2006/10/31/ie7_causes_site_problems/).

Although I have pointed out some rendering problems with the dabs site, it is important to stress that the dabs site is far more complex than the clearleft site, as most ecommerce sites are. It is easier to do a good job on a simple site with a simple layout, and perhaps one which isn't so content driven and dynamic.

I wouldn't say the clearleft site is perfect, no site is perfect. Aesthetics and design aside (which are subjective), both the clearleft site and the w3 site have one bug bear of mine, which is to include easily spam farm extractable email addresses (http://www.ukbusinessforums.co.uk/forums/showthread.php?p=298096), something that wouldn't pass my own personal QA, standards, good practice guidelines.

Several web designers (registered here on this forum) are also able to & do produce code which passes W3C validation. Some of which, I know are watching this thread with great amusement!
Don't mistake someone saying on this thread that W3C validation isn't so important as other things, as a statement that they do not do W3C compliancy at all. Whilst I do not see it as so important as some others do, W3C compliancy is something we strive for with our sitebuilder service, and we have also put some effort into even making more W3C compliant our custom version of osCommerce (which is embedded as a shop section within the sitebuilder).

[creospace]

I don't see the need for the 'medals of honour ' to be displayed on designers' site. Despite what the site owners may think they are quite meaningless to average Joe, they mearly serve to notify other passing web designers that they are working to standards.

[FireFleur]

It is quite amusing takes a bit of time to find one on dabs, as the person who did the markup thought 'stuff the semantic it is all divs' :)

But, here is one.

go to email registration and your off that domain still part of the site though, and wallop 29 errors.

http://validator.w3.org/check?uri=http%3A%2F%2Fwww.prospectemail.com%2Fdab s%2Fmain.asp%3Fpage%3Dregistration%26competition%3 DDBCOMP0001%26source%3DDBW0002&charset=(detect+automatically)&doctype=Inline&group=0&user-agent=W3C_Validator%2F1.654

It is the templating engine and the fear of using anything more exotic than a div that tends to give them compliance to the doctype.

Oh dear, found one, just search for a nic.

http://validator.w3.org/check?uri=http%3A%2F%2Fwww.dabs.com%2Fsearch%3Fq%3 Dnic&charset=(detect+automatically)&doctype=Inline&group=0&user-agent=W3C_Validator%2F1.654

Now for the css validation:

http://jigsaw.w3.org/css-validator/validator?profile=css21&warning=no&uri=http://www.dabs.com

No challenge there.

Anarchy :)

[driansmith]

My point is that I'm astonished at the amount of web designers who's own 'showcase' websites fail W3C validation. Very few display the W3C icons on their site to indicate that the code has passed validation.
.

Most major sites do fail. My view therefore is - mostly irrelevant.

Focus on the aspects that will make your site successful - namely well thought-through SEO.

[stevemac]

Most major sites do fail. My view therefore is - mostly irrelevant.

Focus on the aspects that will make your site successful - namely well thought-through SEO.
Many of the articles that I've read had led me to believe that validation would carry some form of SEO advantage. From reading the replies in this thread, opinion appears to be that may not be the case. Unfortunately, due to forum rules, I can't post any external weblinks to the articles & websites that I'd used as a source of information.

This appears to be a classic case of a 'real world environment' differing from that which is taught by various training resources & establishments. Kind of like leaving university, thinking that you know everything, only to be swiftly brought down to earth!

[FireFleur]

If the site is using valid code, it will tend to validate better against the parsers search engine companies use.

Don't turn your back on validation now steviemac you have stumbled onto an advantage of validated code, it works better for most parsers.

Often you will see people claim you write websites for humans only, well they are not right, a lot of websites are written for computer programs as well, and a DTD or Schema is used as part of the parser ruleset. Spiders and browsers need to understand the markup to display the information appropriately.

So, if a page doesn't validate it may be moved down the line, as it is harder to understand the page, for search engine ranking.

And if you view source in Firefox you will notice red marks where the markup is not valid, and internally a browser has to decide what to do with non valid code, in XHTML1.1 Firefox will error, so you cannot get it wrong.

But again, using a div where a p is more appropriate, is more important than say missing a closing html tag.

Google breaks the rules on their own site, probably because they want to reduce the amount of information transmitted overall; a small reduction in page size could improve performance across the system, and responsiveness is important to them.

[driansmith]

If the site is using valid code, it will tend to validate better against the parsers search engine companies use.

Don't turn your back on validation now steviemac you have stumbled onto an advantage of validated code, it works better for most parsers.

Often you will see people claim you write websites for humans only, well they are not right, a lot of websites are written for computer programs as well, and a DTD or Schema is used as part of the parser ruleset. Spiders and browsers need to understand the markup to display the information appropriately.

So, if a page doesn't validate it may be moved down the line, as it is harder to understand the page, for search engine ranking.

And if you view source in Firefox you will notice red marks where the markup is not valid, and internally a browser has to decide what to do with non valid code, in XHTML1.1 Firefox will error, so you cannot get it wrong.

But again, using a div where a p is more appropriate, is more important than say missing a closing html tag.

Google breaks the rules on their own site, probably because they want to reduce the amount of information transmitted overall; a small reduction in page size could improve performance across the system, and responsiveness is important to them.

I am sorry but I do not agree. The engines are not interested in 'purity' of coding. They want information for their customers - searchers.They also recognise that 'small errors' in coding can occur. Provided it is not a complete dogs breakfast, they have no problem.
Do not waste your time on validation, concentrate on good content and well thought-through SEO.

[FireFleur]

Well I didn't say purity, that's not a term used in parsing.

But the parser is effected by non standard code.

Google and the big search engines have to juggle processing, they have huge data centres but an awful lot of information to get through.

Parsing is done in multiple stages, anything that errors on a first parse could well be moved to the bottom of the heap and anything clean goes through first.

That is the logical way to do it. The errors cause more processing power to be dedicated to sorting out the interpretation of the markup.

[FireFleur]

You have to understand weighting with the search there is a whole load of criteria that determines the ranking position.

> The engines are not interested in 'purity' of coding.

That statement is false, they are interested because it costs them more to process a page that is badly marked up, in fact some doctypes will do better than others, depending on their complexity.

> concentrate on good content

that is good advice, the weighting is highest on that more than anything else. I would use the word information rather than content though, data + context = information.

Markup should be used as context.

Now unless Google wants to release their parser engine open source :) which would be nice, but unlikely, no one can say for sure.

But, they have a computer science and mathematics background and they are hiring most of the known computer science guys, and parser optimisation is computer science 101, back in my days we had to build mini compilers with lex and yacc, badly formed data is rejected, too annoying to handle.

Google won't be that harsh, but it will weight against the page, and part of seo is encouraging google to visit to increase the number of iterations of the page ranking algorithm, and for it to update its page information as soon as possible after a change.

[wood1e2]

In a way Validation may be a waste of time, as it is not directly effecting people buying products from a website. Design, ease of use, layout, flow and price of prodcut all effect that.

But if it doesn't effect your website in a bad way and it is possible to obtain validation then why not go for it.

If it can help with cleaner code which in simple terms makes it easier for search engines bots to scan your websites then even if it only helps you website by 1% then that could put you ahead of someone else.

As for coding for various browsers, well it is just lazyness not to, as not everyone uses IE7.

Plus why not have a website that is viewed correctly (as possible) by 100% of the market.

The same thought process goes for blind/colour blind.

Finally, I am not keen on any badge of honour...W3C etc why should I advertise someone elses products on my website. I even go as far to say I hate seeing 'This website uses Paypal' or 'Google bucket' whooppee...not interested is the website providing what I need?

That if you like is the long and the short of it.... :)

[ukbizbook]

my guys at banilla are compliant and I would argue that all code should conform to the basic standards. Many of the big sites such as Amazon etc can't as they have code that is embedded by users so can't always be 100% compliant but it's non compliance that gives us such marvels as IE6 and all of the hassle of floating div tags not working and so many other things that work in one way with IE and another with other browsers.

my advice, learn it - love it - let yourself get away with a few minors

Ukbizbook

[quikshop]

> The engines are not interested in 'purity' of coding.

That statement is false, they are interested because it costs them more to process a page that is badly marked up, in fact some doctypes will do better than others, depending on their complexity.

There is no evidence that Google pays any attention whatsoever to the 'purity' of mark-up or the coding standards used on any particular website. Google's own website is not standards-compliant and it has not affected any of the websites I have been involved with since 1998.

I'm not sure why you are trying to imply that Google will punish poorly formed websites for cost reasons! In terms of time / energy used, a small badly formed page will take less time to process and therefore be cheaper than a large standards-compliant page.

Its all very good thinking that it would make sense for Google to start taking notice of HTML / CSS / Accessibility standards, but it doesn't make it so ;)

[DotNetWebs]

Well seems a bit of a waste of time trying to get compliant as the search engines don't give a hoot...

I agree with this but still think, provided you don't have to make to many compromises, it's still worth making the effort as there are SO many browsers out there and making your site standards compliant will give you a fighting chance of your site rendering [at something close to] correctly on them. I am not just talking about desktop browsers here but also browsers embedded in devices like mobiles phones and games consoles etc.

As for the icons, I have included small and subtle ones in the footer of Visit Horsham mainly as an aid to me checking the site. Most people probably don't notice they do occasionally help 'sell' the site to new subscribers.

Regards

Dotty

[sirearl]

I agree with this but still think, provided you don't have to make to many compromises, it's still worth making the effort as there are SO many browsers out there and making your site standards compliant will give you a fighting chance of your site rendering [at something close to] correctly on them. I am not just talking about desktop browsers here but also browsers embedded in devices like mobiles phones and games consoles etc.

As for the icons, I have included small and subtle ones in the footer of Visit Horsham mainly as an aid to me checking the site. Most people probably don't notice they do occasionally help 'sell' the site to new subscribers.

Regards

Dotty

well I agree that sites should view properly in all browsers,but don't need to be compliant for that.

There is also the point from an SEO view that having to stick to a set of rules can inhibit ones flexibility.

Not sure about mobile phones and games consoles on the grounds that I think I would not want a sub species as a customer.:|

Earl

[FireFleur]

Not accessibility standards. It is down to the parser.

So for example, and this is one I found recently.

<html>
<head>
<meta
name = "description"
content = "foo bar buzz"
<link
...
</head>
<body>
...

Ok that is non compliant markup it is missing a closing bracket on the meta.

When the parser first runs over this markup, it looks to separate out the elements and put them in DOM data structure so it gets


html
head
error
body
it now has to check from the meta to the end of the link tag to determine what it could be, but if the tag was there it would just parse in one go for the DOM and get:


html
head
meta
link
body


Handling side cases takes more processing power, it has to add a layer of logic that tries to determine best fit, rather than just parsing and moving on to the next phase of parsing.

Don't get me wrong, it is minor in the grand scheme of things, but can give or take an edge on a SERP.

[DotNetWebs]

well I agree that sites should view properly in all browsers,but don't need to be compliant for that..

I agree again but my point is that where you have unknown new browsers W3C code will probably render better than something created in FrontPage 2000.

i.e it is a basic standard the should work across the board.

Regards

Dotty

[FireFleur]

W3C sets the core for the language. Browser developers will use those specs to build their parsers and render engines.

So, it will be more likely to work across unknowns because they will generally be taking from the W3C specs.

But, there are CSS shortcuts such as padding : 0; which is compliant to W3C but won't work in some mobile browsers.

W3C has a big juggling thing to try and pull off, there are people who want smaller tighter code, there is the semantic web:

http://news.bbc.co.uk/1/hi/technology/8047076.stm

which generally requires good use of tags, high relevancy proper use of parameters.

There are spiders that want to process data as fast as possible, they want simple rules nothing complicated, and there are designers and users looking for a fresh and seamless experience from a site.

Admins want to keep the load low.

All of these are debated and then we have browser makers fighting to create lockin and sort of steal the standard, so the W3C has to balance all of this generally just on contribution from both working groups and people donating.

So ignore W3C at your own peril, but become dogmatic to it and you may end up with something that doesn't work.

[stevemac]

With regard to different browsers, devices, etc. Would it not be standard practice to use the same XHTML web pages for all & just link them to a set of browser/device specific style sheets?

[DotNetWebs]

With regard to different browsers, devices, etc. Would it not be standard practice to use the same XHTML web pages for all & just link them to a set of browser/device specific style sheets?

My point was nowadays you DON'T KNOW what all the browsers are out there, or if you do a new one will come along soon.

OK so many are derived from common ones like Opera etc. but even then they have their quirks. e.g I use the Wii browser to view the Internet on my TV in the lounge.

Regards

Dotty

[DotNetWebs]

...
Not sure about mobile phones and games consoles on the grounds that I think I would not want a sub species as a customer.:|...

IMO mobile phone browsers will have more and more relevance. Especially in the 'local' Internet field that I develop in.

Regards

Dotty

[FireFleur]

Mobile accounts for most of the CPUs in the world.

[sirearl]

A little bit from Jill on validation.

http://videos.webpronews.com/2007/04/19/ses-high-rankings-ceo-and-founder-jill-whalen-on-duplicate-content-and-tags/

Earl

[stevemac]

One problem with living miles from nowhere - my broadband connection is little faster than dial-up..:mad:

[edmondscommerce]

check out satellite broadband