HTML Validation?

Hey KAC

I just posted in that other thread on this very issue. I'm only an amateur coder, so you'll need to take mmy view with a large block of salt, but ...

The way I see it, validation is a way to get a good start on having your code display the way you want it to, both in current and, importantly, in future browser software. But I don't think there's a substitute for looking at as many different displays as possible to make sure you're getting what you want.

I suppose experienced professionals are fairly well able to anticipate problems with quirky setups.

Finally, I figure it's important to consider exactly what it is about your code that's not validating. Anything that comes to mind?

Validating is like spell-checking. It eliminates 90% of the problems right off the bat.

Then you need to do the grammar checking -- making sure it displays correctly in other browsing situations. Like real grammar checking, this is best done by a person, not a machine.

Additionally, it's good practice. Why? Because it helps you think more logically about how a page is put together. This will help you in the future when you advance your skills to include other web-based but not web-browser subsets of SGML, notably XML.

ADDITIONALLY, valid HTML helps significantly in using CSS, as it works best when applied to a valid HTML tree.

So you can work your *** off trying to get things working right without validating, or you can validate, learn, and spend less time working and more time making a profit. I'd rather work smarter than harder, eh?

Here's a list I've posted to an internal memo on the issue:

1. Things will (probably) be cleaner and more efficient. Less code means more speed (especially if you move to CSS layout!).
2. Better portability – when a course is exported it will be editable with any standards compliant tool or person; it's also becomes possible to use the code in different devices and for different purposes; you can apply a new design to your site without re-doing all the code from scratch!
3. Better functionality – code written to standard just works better; you don’t have to worry about things not displaying as intended, or how things will function in different browsers or even different devices (theoretically)
4. Forward thinking - standard code will work properly in future browsers as well as current ones
5. Accessibility – non-standard or otherwise poorly written code causes problems for alternative devices
6. Support for the w3c – if nobody supports them then their work is meaningless and web development goes back to the chaotic state it was in five years ago. We don’t want that, do we?

I guess an easier way of restating the first post is;

If a web page that is w3c compliant doesn't work properly on some browsers, then maybe the emphasis should be on testing more with every browser available with as many platforms than on the w3c validation, like some others mentioned.

My guess is the builder of the site listed in the first post of this thread thought that just because the site was w3c valid, that they were done. That the page would show properly cross platform - cross browsers. Which is not the case. You can blame it on the platform or browser, but in the end it pretty much negates the whole list of why your site should be w3c compliant.

sidenote:
It's kind of funny, but I feel like I'm debating religion or politics. I half expect someone to get angry just because I suggest that w3c may not be as important as some think.

MMI, usually the items that seem to invalidate my sites are background images, or not having the "type" on Javascripts. I live by the background image, and just end up having my site be invalid than remove them. how about you, do you validate?

Background images work perfectly fine in the CSS.

Anyway, it's not a matter of W3C to me, it's a matter of having it solid first, then tweaking for browser ERRORS.

You want to ignore validating and only tweak for visual performance, that's your business, but you're missing many of the points of validation (listed above) that have little, if anything, to do with the W3C as the impetus. Additionally, you're risking legal issues by not having accessibility in place.

I'll go a bit further and suggest that if you wanted a discussion, you should be willing to at least consider the answers to the question you posed.

It's kind of ironic - for years web designers were complaining so loudly about the browser differences and how hard it was to get things to work. Now that there's a system in place to help with that, these same people are resisting it!

Quote: Originally posted by Suzanne
I'll go a bit further and suggest that if you wanted a discussion, you should be willing to at least consider the answers to the question you posed.

I don't understand. Could you be more specific?

I don't give a whole lot of weight to validation, KAC. But like I always say, I'm an amateur coder, not a pro. If I were selling coding services, I figure writing to standards would be a good selling point for the reasons listed by Megan.

I like to use vaidators to diagnose problems with displays, along the lines of the spellchecker analogy Suzanne offered. If you have an open DIV or SPAN or whatever, the software will point to it.

Tbh, I haven't written much code in the past five years, and I didn't write much before that. Mmy work validates OK, but never completely it seems. I've used deprecated stuff like NOBR, LAYER and ILAYER, I still haven't converted formatting on some pages to CSS, stuff like that. Nothing I'm much concerned about.

I suppose I'd just laugh at someone suggesting I don't sell site building because I don't write to standard. I already try to sell a high-quality and very affordable professional service to site owners and web developers and almost all of them are too stupid t' buy it.

Quote: Originally posted by mmi
I don't understand. Could you be more specific?

No, I don't believe I can be. It's extremely clear that despite asking for reasons why validation is important, KAC is only concerned with the idea that it's not.

There are people more concerned with profit than good practice -- eventually they are found to be negligent. This industry is still very young, but like building codes, ethical practices, et cetera, I think it's inevitable that using bad coding practices will eventually become actionable in a court of law.

Non-valid coding is bad because it causes an unnecessary expense for the client when it needs to be revamped and "brought up to code". It makes it harder to edit, it can trap the content so that the cost of maintaining information outweighs the benefits, leading to a loss of information for society.

Personally I think any professional in this field needs to be aware of these issues and saying the W3C isn't relevant is just posturing.

Quote: Originally posted by KAC
then maybe the emphasis should be on testing more with every browser available with as many platforms than on the w3c validation. (emphasis added)

Quote: My guess is the builder of the site listed in the first post of this thread thought that just because the site was w3c valid, that they were done. That the page would show properly cross platform - cross browsers. Which is not the case. You can blame it on the platform or browser, but in the end it pretty much negates the whole list of why your site should be w3c compliant.

I sure don't see how he's "saying the W3C isn't relevant" or that he's "only concerned with the idea that [validation is] not [important]."

Quote: Please don't jump at me, it is an honest question, about something that I've wondered about for a while. ... maybe I should do what I do now, check my sites with all the browsers I can and worry less about validating at w3c.

Quote: It's kind of funny, but I feel like I'm debating religion or politics. I half expect someone to get angry just because I suggest that w3c may not be as important as some think.

It looks to mme like his concern was well-founded.

I agree that unless you're building for a gov't or some other organization that insists on strict and complete validation, you need to keep the occasional warning in a proper context.

Imo, the idea that a lot of lawsuits are going to be brought over this issue is nonsense. Accessibility is important; in some contexts it's vital (gov't, etc). But I can't see courts spending half their time deciding how much to award a plaintiff because Mom and Pop web builder used a deprecated tag.

Piffle. Let's let the gangs police the streets. Let's let society develop without a plan.

That's what you're saying.

The W3C is a PLANNING committee. It's involved in helping shape what the web will become. That's why it's relevant, not because of the validator.

Clearly no one can express in a satisfactory way to you why standards (which is really the point, not the validator which only checks them) are important, so it's not a discussion on why they are/are not important, it's posturing.

Posture away.

I think that's an absurd misrepresentation of what I'm saying.

I agree the standard is important and, imo, "clearly" never suggested it wasn't.

I don't appreciate your characterization of mmy view as "posturing," although I don't expect that is of any concern to you.

Here're two more views:

Quote: My position with regards to the W3C validation tool is, and remains, that satisfying it is a distant second to making sure that our code works on as many browsers as possible. Someone pointed out to me that it is possible to write code that is entirely W3C compliant and will work on all browsers. The problem is, however, that these two characteristics are completely orthogonal: writing compliant code does not guarantee that it will work on all browsers.

Thus, while I will worry about someone telling me that, say, my site doesn't work well with Opera 6 on Linux, and try to correct that deficiency right away, I will not worry about someone telling me that the site doesn't validate with the W3C. Why? Because validating with the W3C gives me no advantage from a technical or business perspective.

I'm sure that others will disagree with me, and I respect their opinions. They all run their websites, and I'm sure that they take care to ensure that they are compliant. I"m not sure whether they work with all browsers, but if they weren't I would still respect their choices—it's a matter of setting one's priorities straight.

I don't want to say that we have better things to do than worrying about W3C compliance—but the truth of the matter is that while it takes significant amounts of time to write compliant code (and the resulting code is also much larger), there is no direct, tangible benefit in doing so.

It's definitely an arbitrary choice, but I'm not the only one who has made it. The PHP Site's main page doesn't validate. Microsoft's doesn't, either. These are two sites from two well-respected groups that represent very different development philosophies—and they both share this common trait. - http://blogs.phparch.com/mt/archives/2003_07.html

Quote: In yesterday’s quite lengthy Report (prompted by the release of the Wired redesign and the many issues it raised) we touched on purity of validation and the reasons it is rarely achieved on large-scale, commercial sites, even when the initial templates validate and the client and builders are fully committed to supporting W3C specs. Admittedly, clients and builders who care are still too rare a breed. But even those who do often find themselves flummoxed by outdated middleware, compromised databases, and other such roadblocks.

In a 10 October editorial, Joe Clark points out that independent designers and developers have no problem authoring to W3C specs (and many even enjoy fixing commercial sites with which they are unaffiliated, bringing such sites up to spec with a few hours’ work).

Our point is not that XHTML 1, CSS, and Priority 1 accessibility are hard. They’re easy for any halfway sophisticated designer or developer to achieve. The problem lies in the large-scale systems and third-party content that must be integrated into most big sites. We need CMS systems that encourage compliant, accessible authoring practices instead of breaking them. (Translation: We often build our own if the budget permits.) We need backend folks to recognize that compromising the front end “to force a display issue” is no longer an acceptable option. We need clients who are willing to invest in fixing systems that “work” at the expense of forward compatibility. It will take persuasion and it will take money.

There is no shortage of persuasion. Money is another story.

In a healthy economy, companies invest in R&D, training, and long-range planning. In a sick economy, companies focus on cutting costs, eliminating personnel and processes, and keeping their doors open for the next 24 hours. Encouraging designers and developers to learn compliant authoring methods is easy: show them the benefits. Coaxing struggling companies to invest in the long-term health of their web presence is the real challenge.

The Vignette-generated validation errors that afflicted Wired.com have now been fixed and the site validates. Not every company has the technological chops to fix errors introduced by Vignette and its brethren. And as we’ve just said, few companies have the dough. The systems themselves must become standards-compliant for the web to progress. Site managers must tell CMS vendors compliance matters, just as designers and developers once told browser makers. If enough customers do this, the vendors may be willing to upgrade their tools.

This screenshot of zeldman.com in NCSA Mosaic, submitted by Usman Farman, shows how backward compatible transitional web layouts can be. Mosaic was the first graphical browser, and was last updated in January 1997, a month after the W3C introduced CSS1 (which Mosaic naturally did not support). Zeldman.com is built to current W3C specs yet looks acceptable and is legible and usable in the long outdated Mosaic browser. - http://www.zeldman.com/daily/1002a.html

If the vendors (i.e. not the W3C) cared enough to "invest in the long-term health" of the web presences of their clients, and by vendors I mean Mom and Pop Web-builder as much as Vignette, then what would happen?

Valid code, as I said before, is the FIRST step. It's not the LAST step.

We're still working on getting software developers (browsers and agents used to view content on the web) and application developers (CMS, eCommerce, regular web pages) to fully embrace standards. The more they do, the better it is.

A few years ago, writing valid code would have WILDLY different results in different browsers. And I mean WILDLY. Insanely, make no sense, tear your hair out, it's a freaking table, why won't it display wild.

Now the issues are primarily small glitches (minor display imperfections), and programming differences (JavaScript -- these can be major if people choose to use complex DHTML for functional aspects of the page). Those are now being standardized as well (to the DOM, another W3C recommendation).

Microsoft is *not* an example of a good web citizen. PHP.net is built by programmers who deal predominantly with server-side technologies -- they need, but do not realize they need, a front-end specialist to fix their generated code, much like those folks producing CMS and advertising applications that generate crap coding.

You want to be a slave to whichever browser maker, rss reader, refrigerator viewer comes on the market?

I don't suppose you remember a time when people had to create multiple versions of websites because of the disparity between browsers? When changing a design meant updating every single page, often in the 100s, of a website EVEN when the pages were dynamically produced?

Having to add a few CSS hacks to the one stylesheet seems somewhat anti-climatic in comparison. And this is possible today because of the clamouring for standards and the work of the W3C to create and recommend those standards in the first place.

Pop, the problem with that site you mentioned, it's just another webmaster advertising for work.
Check out the 'proper' way to do meta tags (keywords related to each page), then check out that sites meta tags on any page ,same tags throughout the site and they also spam keywords, ie: flower shop which is no way related to the site. They are proud of their rankings, done by spamming keywords. right or wrong?
I use the same keywords throughout my sites but dont spam words, if I use ads on my site most of the time they wont validate and makes it look like the site wont.

What's wrong with a webmaster looking for work? Besides, I thought this discussion was about validation.

Your comments about the Meta tags are like the pot calling the kettle black. Look at the error in your own Meta tags. The closing tag, "/> is not correct on any of them. Not to mention all of the other HTML errors on all of your pages. Not a single page passed the validatiion test.

Sorry Pop I didn't realise it was your site, but I didn't say it was wrong that you are looking for work, I just stated you were.

Which one of my sites did you look at for the meta tags? my sites are done in XHTML not HTML and XHTML requires all single tags to be closed. If you mean my HTML site, most pages on there validate to W3C's specs (link at bottom right of each page to check yourself), pages that dont validate either mention it doesnt validate or doesn't validate because of the banner advert which I have no control over. My mschat site wont validate now because of the banners as well, but there is also a link at the bottom of all pages on that site to check yourself
and most of my CSS also validates (some may have warnings) but the links are there to validate them also.

I checked munchtech.com. I'm not very familar with xml. I think the closing tag of the Meta tags do not require />. Drop the / and try validating it.

munchtech.com is valid XHTML 1.0 transitional, I can't even validate my site via w3c.org, the htaccess thinks w3c is a bot and kicking it else where.

but via Opera (right click, choose frame/validate) it comes up with "This Page Is Valid XHTML 1.0 Transitional!"

If you don't know what XHTML is you're in big trouble, briefly XHTML is the latest version of HTML, can read more about it here http://www.w3.org/MarkUp/ a quote from their site:

Quote: XHTML 1.0 is the first major change to HTML since HTML 4.0 was released in 1997. It brings the rigor of XML to Web pages and is the keystone in W3C's work to create standards that provide richer Web pages on an ever increasing range of browser platforms including cell phones, televisions, cars, wallet sized wireless communicators, kiosks, and desktops.

XHTML 1.0 is the first step and the HTML Working Group is busy on the next. XHTML 1.0 reformulates HTML as an XML application. This makes it easier to process and easier to maintain. XHTML 1.0 borrows elements and attributes from W3C's earlier work on HTML 4, and can be interpreted by existing browsers, by following a few simple guidelines. This allows you to start using XHTML now!

Your arguement is it should be valid code, true, but shouldn't it also be up to date as well? XHTML has been around for 3 years.
This isn't to cause an arguement, just to open your eyes a little.
But you are right, my sites don't validate to HTML standards because they are all done to XHTML (and topic of the thread is HTML, not XHTML)

I'm somewhat familar with xml and have followed it since it started a few years back. I bought the domain name xmlwebmaster.com in anticipation of getting into writing xml code. It just hasn't happened yet and at 69-years-old I may never get to it. I also bought a few books about xml and browsed thru them.

XML is not XHTML is not HTML. But they all have the same parent, SGML. Validation (and indeed some measure of precision) are ideal for them all. XML will not work at all if you have errors. It's extremely strict. XHTML will work if it's not valid, but not as well as when it's valid, unless you are (properly) serving the document as xml -- but since this only works in advanced browsers like Netscape and Mozilla variants, it's not something most people will bump into. HTML is the most forgiving.

In XHMTL like in XML, all tags must be closed. You can do it like this:

for a tag that has content within it

But for elements that have no *displayed* content, or where the content displayed is in the attribute values (like for forms), you can "self-close" a tag like this:

The space isn't needed, but it causes no harm. It's left in for readability and for browsers that would crash if it were there (older browsers).

How about Meta tags in an xml document? Will they validate if closed with "> or must they be closed with "/ >.

They must be closed with a " />" -- no space between the forward slash and closing angle bracket, but you can have a space before the forward slash.

That holds true for XML or XHTML documents.

Busy and Suzanne
Well, well. Since reading your posts, it appears I need to head to Books-A-Million to buy a book about XHTML.

I wonder where XHTML will leave thousands of todays HTML webmasters. It is pretty evident the vast majority aren't getting the simplest form of HTML to validate. How will they ever cross over to something that requires the use of more brain power?

I agree with what Busy wrote about being in deep trouble by not knowing XHTML. Thanks to both of you I'm going to accept the challenge.

There are many primers available on how to make the transition.

HTML tables were confusing as anything else when they first came out, not to mention frames (which still confuse!) - I'm sure with the support of software vendors that are creating "no code" solutions, it will become easier and easier.

http://www.nypl.org/styleguide/ is a good start. http://www.alistapart.com/stories/betterliving/ is the "unauthorized companion" article that complements it.

For what it's worth, xhtml has been around for a few years now, and is supported by all the modern browsers. It's also in use in the larger (and a large number of the smaller) web development organizations as the main level for coding.

hi, i think if your site looks fine to yourself and its a personal website, dont stress if it validates or not.

But if you are selling templates or offering work to build sites, then yes of course validation does matter.

Its sort of like getting building work carried out on your home, you would want a builder with a good reputation and building standards, so you know its proffesional.

No need to buy a book, the change isn't that great and all the info is on the web.
Main things are:
all tags to be lower case
all values to be quoted
all single tags to be closed etc
and everything laid out properly, no crossed nesting stuff.
and XHTML will drop a few tags like font, center etc one day and use CSS in it's place

even do a search in these forums, should find everything here, if not ask

P.s. Age only matters if your a cheese

Just by conforming to HTML standards I have found myself coding (almost) in XHTML by accident...I'm already quite a pedantic person who loathes bad code, all you need to be able to code in XHTML is a healthy desire to make things valid!

umm... microsoft is a big lumbering beast. it thinks it's larger enough it should be able to arbitrarily set things and how they should be done. any display issues in ie with validated code is an ie issue. most non-windows people don't use mie because of the fact it's known to ignore w3c among other things. if people stop with the microsoft proprietary **** (asp/.net) then microsoft will have to work to be compliant to the same places as everyone else. if people don't stop doing that, then eventually everything else will become microsoft compliant just for compatability, which will mean they have other security issues that currently are inherint in how ie handles things that the others don't have as a result of w3c compliance.

insummary: whenever you can, ignore the ie display issues and do what you can to reverse the assimilation!

Having spent all this week on the subject of XHTML, reading, searching and in general pondering the subject, my conclusion is XHTML and XML are a long way from acceptance by the majority and may never be accepted. In my heart of hearts I feel both will come to a slow death.

If a vast majority of the site Owners or Webmasters now on the WWW have not taken enough care to write good code what makes anyone think they'll get the urge to do so now, especially the big name sites.

I may be wrong, but at this point I see no real strong incentive that would cause me to want to convert a 700 plus page site of mine to XHTML nor to XML. Perhaps we may see new sites that are starting up use the code, but I just don't see millions of site owners going back and converting their code to XHTML or XML. That will never happpen!

I wonder where we'll go from here?

espn.com
wired.com

Many larger corporations in other countries are embracing XHTML as an advantage. Don't write it off yet.

RSS uses XML, so does a number of other industry specific internal applications that then generate XHTML for the web, using XML for their print documents.

Heart of hearts is a fine thing, but when it comes to business, it's the head of heads that counts.

Pop, to me, XHTML is HTML Strict (which many don't realise there are 3 flavors of HTML - transitional, strict and frames) and the normal HTML is HTML transitional.
HTML is a very sloppy code and IE supports this very well, but Netscape and others don't, and in some cases wont even display anything.
XHTML (transitional) is how HTML should of been set out all those years ago, all lower case, values quoted, proper nesting of tags and all tags to be closed (single tags are a bit much) and thats about all there is to XHTML.
But XHTML Strict is more CSS flavored and does away with a few tags which opens the way for new stuff (CSS).
If you want to get really keen you can use XML to create your own tag sets, 90% of people would never use the XML side of it but even some WYSIWYG editors are using XHTML tags and/or option to convert and validate to them.
Since XHTML is a lot cleaner (code wise) it has more change of displaying on more browsers, it still wont display the same in all as not all the browsers support enough of eachothers stuff for that to happen but there is more chance it will.
Also using XHTML with CSS you can get your load speeds down as well.
You'll probably find 80% of webmasters out there use an editor of some sort, of which doesn't support XHTML and the operator proably only knows the basics of HTML so knows no better.

Quote: Originally posted by Pop

If a vast majority of the site Owners or Webmasters now on the WWW have not taken enough care to write good code what makes anyone think they'll get the urge to do so now, especially the big name sites.

I think it's an "invalid" exaggeration to describe code that doesn't validate as "bad."

Pages that don't validate at http://validator.w3.org/:

http://www.att.com/
http://www.citigroup.com/
http://www.ge.com/
http://www.gm.com/
http://www.hp.com/
http://www.honeywell.com/
http://www.ibm.com/
http://www.intel.com/
http://www.jpmorgan.com/
http://www.3m.com/
http://www.yale.edu/
http://www.harvard.edu/
http://www.berkeley.edu/
http://www.umich.edu/
http://www.berkeley.edu/

Is their code "bad"?

And code that does validate may not display as desired. E.g., on http://www.webmasterdisaster.com/ , & #8221; (without the space) doesn't come up as a quotation mark in mmy Mozilla browser. It does here in the forum ( ” ) -- dunno why.

There are other flaws in the front page of this site I would point to as a copyeditor. One is that the author uses 8221 on both sides of the quote in

Quote: Go ahead, do a search for ”webmaster disaster” and see for yourself.

Moreover, there are weaknesses in the use of punctuation. Could this page consequently be described as "bad"? Would a copywriter or editor who received payment for working on it be subject to civil litigation?

I don't really buy the argument about validation being such a determinative boost in SEO. "Disaster master" isn't a string that is found very often anyway (121 in Google). With invalid code, mmy site lists third out of 6080 for "multi media information." I don't expect forcing validation would bump mme to #1.

And perhaps more importantly, why would a surfer be looking for either term? I'm second for "editing websites," but not listed for "editing web sites." I tried putting that in mmy meta tag; didn't seem to help. Maybe I should try again. Or perhaps I should toss editing standards in the trash and use it both ways in the text on that page. I suppose I'd take the issue more seriously if the hundreds of people who know ALL ABOUT my high-quality, affordable service weren't so determined to keep their fifty bucks and their sloppy copy.

Btw, no one has ever contacted mme saying they found mmy site on Google. I suppose the content needs improvement, but I have to wonder about the audience.

Here's a discussion of the code on php.net:

http://blogs.phparch.com/mt/mt-comments.cgi?entry_id=65

I have just one question,(probably been asked a hundred times) but why do you type mme?

I get about 50% of my hits from Google. About 35% are blogs, personal sites, bookmarks, et cetera. The rest are other search engines and directories.

That's another thing about valid code -- it makes it MUCH easier for search engines to find your content and therefore to index your content and get you listed.

Mmi, you may have issues with "website" v. "web site" because Google throws out duplicate results, but I'm not sure -- check Google for "web site" and "website" alone, then choose which one is in common use. Of course, the more focused your content is, the better.

On why validation is so important -- http://www.w3.org/People/Bos/DesignGuide/introduction.html

Fwiw, I agree (and I expect just about everyone does) that validation is important. I think the issue I (and perhaps others) have is with the idea that a page is "bad" if it doesn't completely pass the W3C validator. As I said earlier, I think you need to look at why it doesn't pass and what the consequences are.

Quote: why do you type mme?

I would think the answer is rather obvious: so people will ask mme why I do it?

Such an interesting thread. And to think that only one year ago there were people dissing XHTML. I wonder what a thread like this would be like in 2004??

Validation is a tool and should be used from the very beginning when building a site. You will be glad that you did in the long run. It's not that hard if you put your mind to it.......and just do it!

I get a lot of hits (80-90%) from search engines to my HTML site, and the most popular page, with nothing even coming close to second is the 'bad html' page.
People are looking to tidy up their code

I'm using "HTML Candy" - http://www.anetto.com.ru
It is more useful than HTML Validator.