Forum Moderators: open
Here you will find easy, painless techniques and ideas to improve your Web site quality and make your Web site valid. This document is intended for HTML users, developers working on Web applications, and Web masters.
Thank you, thank you, thank you!
Take care,
Louis
With W3C standards, you'll be able to have very exciting Web sites.
The first thing I noticed was that the topmost heading had "slipped" a bit within its div, so the tails on the Q and y are invisible. It reads: "Conformance and Oualitv Assurance". (MSIE5.0, Win98, 800x600px)
"To validate and adhere to standards your site must look as dull as possible"
"To validate and adhere to standards your site must look as dull as possible"
Yours might...
there is no excuse now not to validate, well there is, lazy coding ;)
Don't get me wrong, I am 100% for web standards, and am constantly pushing my colleagues to make sure that their sites validate (it's an uphill battle, but we're getting there).
What get me really is that there is a mindset amongst some that to write good, valid presentational layer code, design must suffer. I disagree with this completely. As you said, it boils down to people being lazy about their code.
With the up and coming 'mobile device surfer' your site needs to downgrade nicely, which is a handy by product of strict coding
Please read this: [webmasterworld.com...]
I wrote there what I think about it.
However, I validate before (if I have to) kludging my code for NS4.x. This helps me to spot any errors that may crop up and gives me the peace of mind that the code I have written is good, clean code.
It's a necessary evil at the moment and we must still do this, in order to cater for the x% of NS4.x users your site will get - and don't forget that this percentage will vary dramatically based on your user demographics and target audience.
Here's a question that just struck me - if you have valid code and then insert (for example) the margin attributes into the <body> tag, with IE6 still go into Quirks mode, or will it just ignore those tags?
Here's a question that just struck me - if you have valid code and then insert (for example) the margin attributes into the <body> tag, with IE6 still go into Quirks mode, or will it just ignore those tags?
I'd have thought (assuming the document is set for Strict Mode) that the browser would ignore the tags. (Just like IE ignores the < abbr > tag, whatever the mode, grrrr.)
Here's a question that just struck me - if you have valid code and then insert (for example) the margin attributes into the <body> tag, with IE6 still go into Quirks mode, or will it just ignore those tags?
IE 6 will remain in Standards-compliant mode but will obey the attributes. Put a div on a page with width, padding, and borders set, and try with a combination of margin settings and !DOCTYPES.
I believe the modes strictly concern CSS rendering and do not address the issue of deprecated or proprietary elements and attributes. [Not that I think this is wise ...] Relevant information can be found in the MSDN Library [msdn.microsoft.com]. Mozilla has similar information [mozilla.org] about their own implementation of Quirks-mode on their website.
I can create good-looking, easy-to-use websites using standards-compliant code. It's not hard. It's just that it's not always cross-browser compatible.
I wouldn't mind if the W3C to make their pages look good. Somewhere on that site they have really gone to town with CSS just to prove that it's possible to create really daring envelope-pushing pages with standards-compliant code -- unfortunately, that page not only looks ugly, it's broken in IE6.
What W3C's "Oualitv Assurance" really proves is that it is just as possible to create pages that are ugly, badly designed and incompatible with a large number of browsers, while using perfectly standards-compliant code.
"Most of the Web sites on the Web are not valid. We may assume that this is the case for 99% of the Web pages
Why?"
Hmmm... I wonder.
Perhaps because 100% of the web's users don't use a fully standards-compliant browser? (as there isn't one)
Or perhaps because the w3c were *so* slow in publishing html standards in the 90s that all the browser makers went off and did their own thing?
Or perhaps because the w3c's new standards don't solve existing problems but do introduce a whole slew of new problems...?
Or perhaps because the w3c site is so hard to use? (And ironically, the WAI pages are actually pretty stupidly designed wrt blind users)
etc
The Emperor has no clothes! :)
I think the problem with the W3C attitude is that they insist you make every single byte validate perfectly.
What, you would prefer that they provide validation tools that don't validate correctly?
It's not like they will impose sanctions against you or send out their super-secret team of Catholic nun assassins if you don't follow their standards in your own website creations.
What W3C's "Oualitv Assurance" really proves is that it is just as possible to create pages that are ugly, badly designed and incompatible with a large number of browsers, while using perfectly standards-compliant code.
Irony aside, bashing the W3C because the site is not a award-winning-design art site that displays perfectly on every legacy browser (even the mighty and ubiquitous MSIE) is really only picking nits.
The W3C is in the business of creating standards, so one should expect them to ask people to conform to those standards, and give reasons why they are a Good Thing(tm).
I ask everyone who has argued in this vein: Do you really disagree with the actual written points made in the essay?
Standards are all about cross-platform, cross-browser compatability. That is the whole point, so that we web designers can create without having to worry about the volatility of the market, or which browser happens to have the greatest user base at the moment.
However, standards are not about completely supporting every legacy browser that has ever been created. True, if a particular browser doesn't support standard correctly, then the W3C QA [w3.org] site is not going to display perfectly. Also true, the default style for the W3C CSS page [w3.org] is ugly as sin, and none of the alternate styles are stunningly attractive. But it's not the W3C mission to author beautiful, legacy-browser supporting websites.
Again, I ask: Does anyone really disagree with the actual written points made in the essay? Or with the general argument that vendor-independent standards are not a Good Thing(tm)?
Of course sites should work on as many browsers as possible, be free of markup mistakes, be accessible, have no broken links etc. But what have any of these to do with w3c standards?
I've been doing web sites a long time. I've always done sites that are accessible, and work on a wide range of devices. And all that time I've completely ignored the w3c. Why? Because w3c standards are just another set of rules with associated quirks, like the nn4 set, the ie4 set the ie5 set, etc. But the difference with the w3c set is that nobody at all browses using that set of rules. Nobody. At all. So why even pay it any attention?
w3c validation <> accessible <> usable <> cross-device etc
BUT the danger is that novices will think that just by passing a page through a validator with no errors means that it "works"
As soon as someone gives me a reason to think that the w3c has a clue (any clue) I'll sign up to be a full card carrying member.
Following a vendor-independent standard that is not actually supported by any vendor (and which actually has enough grey areas to ensure rendering quirks for all time) is just insane to me. When did the world go mad? :)
with the site validated I can be fairly sure that it won't fail catastrophically...there are a number of well known "quirks" that have to be dealt with and otherwise I don't have to panic
I can generally relax about new browsers being introduced...it's getting rare for a new browser to fail badly on valid mark up or css
above all it tells me about all my typos
Those standards define what "markup mistakes" are, for example. The same standards also define the difference between a working and a broken link, and they make it signficantly easier to write pages that work on as many browsers as possible.
Because w3c standards are just another set of rules with associated quirks, like the nn4 set, the ie4 set the ie5 set, etc. But the difference with the w3c set is that nobody at all browses using that set of rules.
This is simply wrong. The various "quirks modes" of most modern browsers all have a little brother called "standards mode". If you use one of those browsers to view a site that includes a doctype header and is encoded according to that doctype, then that means that you are browsing using "that set of rules". You may not have noticed any difference (probably because it just worked without making a big fuss about it), but that doesn't make it any less true.
Don't get confused by the fact that all existing implementations still have bugs. Unfortunately, bugs are inevitable in software, but that is really no argument against the attempt to implement the standards that all browser vendors have officially(!) agreed on, by participating in the relevant w3c working groups.
w3c validation <> accessible <> usable <> cross-device etc
Validation is the lowest common denominator in determining technical correctness. Accessability and useability involve a lot more than just valid HTML, but in the strict sense, they are impossible without. The same is true for cross-device compatibility.
Your way of thinking seems to be firmly anchored in the past. The w3c standards define the future. You will be assimilated, resistance is futile... ;)
The fact that their site is ugly and partially broken is not nit-picking, it's seriously detrimental to the W3C's own case. "My Web site is standard!" it proclaims, to which I reply, "My website isn't quite standard, but at least I don't get about 200px horizontal scroll for no apparent reason".
with the site validated I can be *fairly* sure that it won't fail catastrophically...there are a *number* of well known "quirks" that have to be dealt with and otherwise I don't have to panic
Now that's *exactly* what I mean! ;)
The fact that their site is ugly
I guess I'm the only one who actually *likes* the design of the W3C and Jakob Nielsen sites. Personally, I would like the whole web to look like W3C, Nielsen, Yahoo and all the other totally simple, basic, easy to navigate sites.
I guess it's the fact that I'm a very textual person *and* color blind, all of which biases me against "design-oriented" layouts.
Tom
Hmmm... so you really prefer a full collection of quirks to a standard with some quirks?
The most compelling argument against validation that I have heard so far went: "We have always done it that way!" ;)
I always say that design is how you present content. Consistent layout and navigation, good-sized fonts, careful use of colour, high contrast where it matters. One thing seriously lacking at W3C is consistency -- at least, that's my experience.
A small amount of eye-candy is a bonus, but a very useful one if it doesn't interfere with content, navigation, accessibility etc., and doesn't impinge too much on (perceived) download times.
Of course, there is a lot of personal taste involved here. But... I just can't bring myself to warm to the W3C site.
Don't get confused by the fact that all existing implementations still have bugs
Now thats exactly what I mean as well! It's madness! ;)
It's not the existence of standards that bothers me, it's the way they are constantly pushed as if they were some sort of panacaea, and anyone who lets slip a proprietory attribute in their code is some kind of felon. Articles like the subject of this discussion seem to be less about tips on making your sites look attractive while making sure they validate (which is a good thing), and more about making sure your sites validate.
Hmm. I guess a lot of that is perception, and that's a valid argument. ;)
In my view, this is exactly the stance that the W3C should take (well, I don't know about the felon part). As a standards organization they should create standards, provide tools to validate against those standards, and should argue for the use of said standards with some conviction. Consider if the W3C took this stance:
"Authors shouldn't really worry about having your pages validate, and browser manufacturers should feel free to ignore us whenever it suits their fancy. We really hope that you listen to us, but you don't have to if you don't want to."
I would argue that this would not be a very effective approach.
It would help (even uninformed) clients identify when their web designers have secretly given them a "this site works only in ie5.5 sp2 q123456 beta 2" or even when an Overwrought Flash Extravanganza means the site has no text for search engines (which, of course, may or may not be a problem). You could stick a bobby-type thing on there as well... A score of 100% means no obvious problems but still needs to be checked by an expert, 50% = (?)
Now. Who's got 5.75K usd? :)
As a standards organization they should create standards, provide tools to validate against those standards, and should argue for the use of said standards with some conviction.
True, but not on this page, which is supposed to be about making your pages both valid and attractive.
A more honest and helpful approach on a page which is (according to the introduction) aimed at designers, not browser manufacturers, would be:
"There are many advantages to validation, such as:
[insert long list here]
Not all browsers interpret code correctly, however; most browsers have bugs, some simply don't implement certain features. We recommend therefore that you test your pages thoroughly. In some rare circumstances, you may find that a workaround involves using invalid code or proprietory markup. Using non-standard code may have unforseen drawbacks, however, so use it only as a last resort and with great caution. It will also mean, of course, that your pages fail to validate.
On these pages, however, we show you how you can create some attractive-looking websites with perfectly valid code, which will display correctly in most current browsers and which degrade gracefully in older browsers..."
(See? The text admits to some problems, yet manages to argue that this is due to sloppy work by programmers who need to get their acts together and that there are ways of working around them...)
