Slashdot Mirror

• Invalid document encoding
• CSS errors
• Accessibility Failure

But as long as their branding is good, because thats what matters, not that they are violating the UK's Disability Discrimination Act which applies to web sites. Overall I see no evidence these people are competent to mirror software, I do however look forward to enjoying their 'branding'.

• Invalid document encoding
• CSS errors
• Accessibility Failure

But as long as their branding is good, because thats what matters, not that they are violating the UK's Disability Discrimination Act which applies to web sites. Overall I see no evidence these people are competent to mirror software, I do however look forward to enjoying their 'branding'.

Flash sucks because it isn't standards compliant.
Websites that use flash navigation and provide no text alternative are totally unusable by my blind friend. Text -> Speech converters can't touch them.

Then again, you're probably right that most /.ers dislike it for other reasons. I mean, Slashdot does have a good 170 HTML errors. Even Microsoft beats that (although only by 4).

Flash sucks because it isn't standards compliant.
Websites that use flash navigation and provide no text alternative are totally unusable by my blind friend. Text -> Speech converters can't touch them.

Then again, you're probably right that most /.ers dislike it for other reasons. I mean, Slashdot does have a good 170 HTML errors. Even Microsoft beats that (although only by 4).

Think? What has parents thinking to do with w3c's names?

And no, they're Microsoft's invention.

Why can't Slashdot be validated??

Did you try to click your own link? It seems something changed in the /. web server configuration, as it is now allowed to validate the site directly withing Validator. Or maybe the IP address of validator changed and they were filtering it?

Maybe you should post this feature request to the Validator mailing list?

I've been on the list for a short while now, and the people there are very helpful and willing to take on suggestions. I'm not a html guru, but I was able to contribute in my own small way to the new validator, by pointing out some spelling/layout mistakes in the beta version.

Why can't Slashdot be validated??

Of course, we already know that /. is not up-to-date for good html, but meh.

After a quick look at the standards, one can tell that the browsers are not required to support blink

quote w3c.org:
Text blinks (alternates between visible and invisible). Conforming user agents are not required to support this value. http://www.w3.org/TR/CSS2/text.html#lining-strikin g-props

looks like its being maintained like this in css3 http://www.w3.org/TR/2003/CR-css3-text-20030514/#t ext-blink

After a quick look at the standards, one can tell that the browsers are not required to support blink

quote w3c.org:
Text blinks (alternates between visible and invisible). Conforming user agents are not required to support this value. http://www.w3.org/TR/CSS2/text.html#lining-strikin g-props

looks like its being maintained like this in css3 http://www.w3.org/TR/2003/CR-css3-text-20030514/#t ext-blink

From W3C CSS Rec.: UAs must recognize the keyword 'blink', but are not required to support the blink effect. I wish there was a way to get Firefox to ignore text-decoration: blink.

No need - there's already an organisation dedicated to making the web a more friendly place. The problem is, some, browsers don't follow the standards too well, choosing to implement proprietary extensions to published standards, and implementing standard extensions slowly at best.

<voice type="Blackadder-Wise-Woman">There are three solutions to this:

1. Persuade certain browser vendors to comply with modern standards;
2. Persuade the entire online community to switch browsers to modern, standards-compliant browsers;
3. Kill everybody!

The point is NOT to make a web site accessible to people who don't understand web sites! This is like the icon caption: "AOL 4.0: DOUBLE CLICK TO START". Don't do that. It's okay to say "Click here to view the page I have written about foo." instead of "I have written a page about foo.", but please don't take all that advice literally.

Yes, please do not take this advice literally, as it is wrong. The poster is correct that you are not to build a web site to people that don't understand web sites, but W3C actually recommends the exact opposite of his example of what is ok.

Also, making multiple links only seperated by spaces is bad. Not just because slashdot's filter (or whatever it is that adds the link domain) probably just garbled that, but those with poor sight cannot tell the difference when one link ends and the other begins (unless slashdot garbles it for them to point out a link is done and where it goes).

Also, if you come across sites referencing Bobby, please note that for some reason Bobby has decided that if an image is used anywhere in the page it will not pass their test. Previously, they would mark it off and warn testers that they must ensure that colors aren't used to differentiate context. Because of this, there are many sites out there that will claim Bobby compliance when they are not. That said, Cynthia Says provides a similar tester with a few more options. If you use Mozilla, Chris Pederick's Web Developer Extension contains an option under the Validation menu to validate against WAI Accessibility or Section 508 (as well as validate links, HTML and CSS).

Finally, it's worth noting that some you should check your pages in Lynx Viewer to see how the page would look in Lynx (or just run Lynx yourself). This is useful for when judging your content based on its textual equivalent (which in some instances is what is read off by screen readers). Also add a "skip to content" link whose CSS sets it to display none for graphical browsers (some people suggest leaving this on, but W3C's validator even uses this method so I go with them). If they are having the screen read to them after a while they will no how the navigation system works, and won't need to have all those links read to them and just want to get to the actual content in your page. If you go the full XHTML route, you'll also have accesskeys and tabindexes available so they can tab through your links corretly and can get back to the beginning of your page if you set a named anchor as the first thing they tab to (the second being the skip to content, thus they can simply hit the 'T' key on to take them to the top and then skip the content to get to the beginning of the content OR they can go through the navigation (of course, you could have given each entry in your navigation an accesskey, but that's not always helpful, and this is useful in case they forget what they key is for something in the navigation).

Wow, I can't believe I knew this much on the subject....

The point is NOT to make a web site accessible to people who don't understand web sites! This is like the icon caption: "AOL 4.0: DOUBLE CLICK TO START". Don't do that. It's okay to say "Click here to view the page I have written about foo." instead of "I have written a page about foo.", but please don't take all that advice literally.

Yes, please do not take this advice literally, as it is wrong. The poster is correct that you are not to build a web site to people that don't understand web sites, but W3C actually recommends the exact opposite of his example of what is ok.

Also, making multiple links only seperated by spaces is bad. Not just because slashdot's filter (or whatever it is that adds the link domain) probably just garbled that, but those with poor sight cannot tell the difference when one link ends and the other begins (unless slashdot garbles it for them to point out a link is done and where it goes).

Also, if you come across sites referencing Bobby, please note that for some reason Bobby has decided that if an image is used anywhere in the page it will not pass their test. Previously, they would mark it off and warn testers that they must ensure that colors aren't used to differentiate context. Because of this, there are many sites out there that will claim Bobby compliance when they are not. That said, Cynthia Says provides a similar tester with a few more options. If you use Mozilla, Chris Pederick's Web Developer Extension contains an option under the Validation menu to validate against WAI Accessibility or Section 508 (as well as validate links, HTML and CSS).

Finally, it's worth noting that some you should check your pages in Lynx Viewer to see how the page would look in Lynx (or just run Lynx yourself). This is useful for when judging your content based on its textual equivalent (which in some instances is what is read off by screen readers). Also add a "skip to content" link whose CSS sets it to display none for graphical browsers (some people suggest leaving this on, but W3C's validator even uses this method so I go with them). If they are having the screen read to them after a while they will no how the navigation system works, and won't need to have all those links read to them and just want to get to the actual content in your page. If you go the full XHTML route, you'll also have accesskeys and tabindexes available so they can tab through your links corretly and can get back to the beginning of your page if you set a named anchor as the first thing they tab to (the second being the skip to content, thus they can simply hit the 'T' key on to take them to the top and then skip the content to get to the beginning of the content OR they can go through the navigation (of course, you could have given each entry in your navigation an accesskey, but that's not always helpful, and this is useful in case they forget what they key is for something in the navigation).

Wow, I can't believe I knew this much on the subject....

The W3C (the people behind the HTML, HTTP and CSS specifications) have published the Web Content Accessibility Guidelines for exactly this situation.

The most sensible methodology is to write basic, meaningful HTML, and then use CSS, Javascript, images, etc where they will add to the value of the website where you can do so without constructing barriers for certain groups of people. For the most part, HTML is already accessible, and it takes screw-ups to make it inaccessible. Unfortunately, the types of screw-ups that make websites inaccessible are very common (things like leaving out alt attributes on images, trying to fix font sizes, etc).

As for it being aesthetically pleasing, accessible websites don't have to be boring, but if you have to choose one or the other (you rarely do), wouldn't it be better that somebody could actually use a website, rather than it being a pretty-looking, but ultimately useless toy?

The W3C (the people behind the HTML, HTTP and CSS specifications) have published the Web Content Accessibility Guidelines for exactly this situation.

The most sensible methodology is to write basic, meaningful HTML, and then use CSS, Javascript, images, etc where they will add to the value of the website where you can do so without constructing barriers for certain groups of people. For the most part, HTML is already accessible, and it takes screw-ups to make it inaccessible. Unfortunately, the types of screw-ups that make websites inaccessible are very common (things like leaving out alt attributes on images, trying to fix font sizes, etc).

As for it being aesthetically pleasing, accessible websites don't have to be boring, but if you have to choose one or the other (you rarely do), wouldn't it be better that somebody could actually use a website, rather than it being a pretty-looking, but ultimately useless toy?

NO. Do not use the font tags.

There are two lists of recommendations which would be helpful for you and your team:

• Web Accessibility Initiative (WAI)
• Section 508

(comment on HTML: needs a bloody fraction tag!)

That's what MathML is for. ;)

This examples site, you mean?
Not every page on the W3C's site is perfect. Whatever. This isn't CSS's fault; it could easily be set up so that this doesn't happen.

It's the result of absolute positioning gone horribly wrong. Of course, it doesn't occur for me until my browser's 430px wide. If people are browsing at that width, then they're going to have problems on just about any site.

Who's spec?

The Curl spec?

The RDF spec?

The XUL spec?

The
FLEX spec?

The web is growing up, and there are plenty of specs.

CSS does mundane and complex styling, but it is not, to my understanding, meant to replace tables exclusely. You had trouble getting a three column layout to work? Apparently you didn't take a look at CSS positioning. As for the "little space as possible" problem: not sure exactly what you're wanting to do but you can use width and min-width (minimum width), in addition to the CSS positioning I mentioned.

Want a header with centered text and left/right flanking images? Easily accomplished with several lines of structured code. I would provide an solution but I'm not sure how you're want it to be done. If you take a look at my website's CSS you can see it is very possible to use CSS to design an entire site.

CSS does mundane and complex styling, but it is not, to my understanding, meant to replace tables exclusely. You had trouble getting a three column layout to work? Apparently you didn't take a look at CSS positioning. As for the "little space as possible" problem: not sure exactly what you're wanting to do but you can use width and min-width (minimum width), in addition to the CSS positioning I mentioned.

Want a header with centered text and left/right flanking images? Easily accomplished with several lines of structured code. I would provide an solution but I'm not sure how you're want it to be done. If you take a look at my website's CSS you can see it is very possible to use CSS to design an entire site.

The best way to stay with (X)HTML/CSS compliance is to use the free (X)HTML validator and CSS validator provided by W3C. Also remember that using working drafted status technologies such as CSS3 is not advised since most browsers do not yet support it, whereas with XHTML 1.1 nearly all support it.

The best way to stay with (X)HTML/CSS compliance is to use the free (X)HTML validator and CSS validator provided by W3C. Also remember that using working drafted status technologies such as CSS3 is not advised since most browsers do not yet support it, whereas with XHTML 1.1 nearly all support it.

The best way to stay with (X)HTML/CSS compliance is to use the free (X)HTML validator and CSS validator provided by W3C. Also remember that using working drafted status technologies such as CSS3 is not advised since most browsers do not yet support it, whereas with XHTML 1.1 nearly all support it.

The best way to stay with (X)HTML/CSS compliance is to use the free (X)HTML validator and CSS validator provided by W3C. Also remember that using working drafted status technologies such as CSS3 is not advised since most browsers do not yet support it, whereas with XHTML 1.1 nearly all support it.

The best way to stay with (X)HTML/CSS compliance is to use the free (X)HTML validator and CSS validator provided by W3C. Also remember that using working drafted status technologies such as CSS3 is not advised since most browsers do not yet support it, whereas with XHTML 1.1 nearly all support it.

The best way to stay with (X)HTML/CSS compliance is to use the free (X)HTML validator and CSS validator provided by W3C. Also remember that using working drafted status technologies such as CSS3 is not advised since most browsers do not yet support it, whereas with XHTML 1.1 nearly all support it.

The best way to stay with (X)HTML/CSS compliance is to use the free (X)HTML validator and CSS validator provided by W3C. Also remember that using working drafted status technologies such as CSS3 is not advised since most browsers do not yet support it, whereas with XHTML 1.1 nearly all support it.

a word in favor of tables: they're NOT deprecated for tabular data, and DESPITE the "zealots'" insistence that they are inherently EVIL for layout, they ARE still "legal" there.

The HTML 4.01 specification directly contradicts you:

Tables should not be used purely as a means to layout document content as this may present problems when rendering to non-visual media. Additionally, when used with graphics, these tables may force users to scroll horizontally to view a table designed on a system with a larger display. To minimize these problems, authors should use style sheets to control layout rather than tables.

It seems pretty clear that <table> elements aren't legal for layout purposes, with or without quotes around the word "legal".

There are still issues: fonts are (render?) slightly larger in Moz 1.4 than in ie5.5 or 6.

This is the web; font sizes will vary depending on a whole range of factors and if you are expecting any particular size, then you are doing something wrong.

CSS positioning is fundamentally broken. It's simply impossible to acurately recreate some of the stuff tables do

Do you mean CSS positioning is fundamentally broken, or do you mean Internet Explorer's CSS positioning is fundamentally broken? Before you answer that, take a look at the CSS specification, particularly the section all about display: table. All major CSS supporting browsers apart from Internet Explorer support that part of the CSS specification.

Even something as simple as "center this element on it's parent" is impossible without absolute positioning.

Completely false; use margin: auto.

If you're going for pixel-perfect positioning, you'll want CSS though.

If you're going for pixel-perfect positioning, you'll want to stay as far away from the web as possible. Variations in rendering are both possible and desirable in terms of both HTML and CSS.

The cross-browser issues are a real pain with CSS - it's worse than the hacks we all used back in 93 to make stuff render correctly.

No, you're just so used to the workarounds associated with table layouts that you forget they exist, or are so unaware of recent browser developments that you are allowing your sites to break without knowing it.

For example, what about the fact that tr, th and td elements are all defined with optional ending elements? I don't see anyone whining that they need to explicitly close them even though the specification says otherwise. But describe a similar limitation of user-agents that happens to include CSS, and suddenly CSS is broken? Give me a break.

Similarly, I don't see anybody whining about the inconsistency of table layouts, despite the fact that plenty of the whiners use them to lay out sliced up images that break in recent Gecko versions (hint: images are inline; table cells need to include space for descenders). But start talking about inconsistencies in CSS rendering, and suddenly CSS is broken again.

IE may be broken and non-compilant but it's still got the 80-90% market share and if you're doing any large-scale site it MUST render correctly on IE. That means giving up on a whole ton of CSS functionality

That's true. What you fail to mention, however, is none of that functionality is needed to replace table layouts, and so is irrelevent to your argument.

Maybe CSS 3 will adress some of the issues.

Where have you been? CSS 2 addressed those issues. Microsoft had already won the browser war by then though, and never bothered with a decent CSS implementation.

It's certainly a step up on font tags

Look, if you want to be taken seriously when talking about web development, stop harping on about "tags". When you say "tags", what you are actually talking about are "elements", and you just look stupid when you get it wrong.

Up to a few weeks ago, I would have laughed at the idea of CSS being used in serious web design for anything other than simple color and theme coordination. On commerical sites, CSS seemed the exception to the rule, and I scoffed at sites that used and tags to layout their page.

But that was before I actually saw the power and wisdom of incorporating CSS.
I recently updated my personal test site to use full CSS for the structure and design, and was very pleased to learn about the two key benefits of CSS website design: structure and format.

Structure:
For ages my own coding methods involved nested table within nested table, until the complexity of my pages got so complex that a simple updates became a gamble of helping or shredding the resulting page. Nested tables are also nearly impossible to coordinate for pages that must be scalable for accessibility, or simply stretching the viewable resolution for more modern video display sizes. Tables were originally meant for one thing: formatting text data, not carrying the workload of page structure. The truth is, it is much easier and precise to define a site's structure using CSS positioning. I am a minimist at heart when it comes to my source code, and CSS has not only helped to reduce clutter in my source code, but in most cases has reduced the source size by about 20%.

Format:
Using relative font sizes and design templates for formatting text not only makes universal page design easy, but it also makes browser loading faster, since CSS can be cached by local browsers while hard-coded or code includes must be reloaded every time the user clicks a link or refreshes the browser. That means reduced server load and increased load speeds for the user, too.

Now I realize, CSS is not the end-all and be-all of web design. There are some weaknesses, and the typical cross-browser support that needs to be worked out. But for the serious web designer, you can't ignore the elegance and the design concepts that make CSS a very powerful (and in some cases, superior) design tool.

If you are interested, the W3C site has some great CSS howto's and examples on replacing table-based structure on your site.

I suggest adding a new error code number, 5000, to the list of valid HTML error codes for this purpose.

Or how about 507 or 508 or 520? A four-digit code would probably confuse some browsers.

Or, even better, just use 402! The RFC defines it as "402 Payment Required."

The book's site doesn't validate, either. That isn't a huge sin in my book, but I see that the site uses named colors (black, gold, navy) for attributes. This is a no no.

yup, and they even got the wrong order of the psuedo-classes

a:link
a:visited
a:active
a:hover

should be

a:link
a:visited
a:hover
a:active

and they aslo use properitary css, speaking of the scrollbar "css". propitary css which is not properly marked with vendorname Vendor-specific extensions
and to qoute w3.org: Vendor/organization specific extensions should be avoided

yup, and they even got the wrong order of the psuedo-classes

a:link
a:visited
a:active
a:hover

should be

a:link
a:visited
a:hover
a:active

and they aslo use properitary css, speaking of the scrollbar "css". propitary css which is not properly marked with vendorname Vendor-specific extensions
and to qoute w3.org: Vendor/organization specific extensions should be avoided

I may be a bit strange in this, but I learned most of the CSS I know by reading theW3C Recommendation. I started reading W3C recommendations when I wanted to learn how to code SVG. Tutorials and examples were relatively rare on the web, so I just found the specification and went from there. I find W3C recommendations to be very readable, and I've since read the recommendations for CSS and every HTML since 4.0. My web design has changed dramatically (and my dislike for IE has deepened).

Has anyone else learned a web technology strictly from the specs?

True, although it's deprecated. It is, however, still part of the CSS standard. There was some discussion about allowing "centre" as a synonym, but it didn't make it into the final standard IIRC.

What kind of /.er are you?

Anyway, I think you give them too much credit.

Not to mention the farther reaching XHTML2!

This has been on the table for a while. Mozilla has "reasonable" (as they describe it at mozilla.org) support for it.

It is always nice to see a site change when I hit my 0){var cs=!document.styleSheets[0].disabled;for(i=0;iTogg le CSS bookmarklet. Too bad Slashdot uses shitty fat HTML. I guess Slashdot doesn allow links that start with the javascript: pseudo protocol

var i=0;if(document.styleSheets.length>0){var cs=!document.styleSheets[0].disabled;for(i=0;i

PostgreSQL has always been OS, but it didn't appear until 1996. The original Postgres engine goes back much further, but didn't support SQL. Also Stonebraker and other Postgres people saw it as the basis of commercial products not free software.

You're probably right, anyway -- even if Interbase and PosgreSQL had been available for free in 1995, their greater complexity would have kept most web developers from using them. But they weren't so we're dealing in theory.

1. Ability to parse is not the same as ability to use. I'm able to parse French. I know all the letters. But I'm unable to use it. I could use it if I get a dictionary. However there is no dictionary for Microsoft XML available.
2. XML is not Microsoft format. It was defined by W3C. Microsoft may have publish the same information or create extensions, document them and publish those. But I prefer to stick with W3C.
3. MS documents are sometimes (e.g. format of .doc is not) accessible but they are not open - you can't contribute.

...it says that Microsoft is involved with developing the format...

Not to long ago there was a push for Microsoft to adopt open file formats for their office suite. They naturally didn't follow through. Their reason is they have a virtual monopoly in office suites - despite very viable alternatives. If they adopted an open file format then that would, in their mind, strengthen the competitors and weaken customer lock-in.

Their motive for advocating an opne 3D graphic format is that they have no stake in the 3D imaging market. If an open format is adopted then that gives them a leg up on taking over the 3D image market.

The interesting thing is how Microsoft "embraced and extended" the SVG format - only to make their own incompatible format wvg. This is inspite of the fact that Microsoft was involved in the specification. I would suspect they will use the same strategy with U3D.

Plenty good enough to get started with, WYSIWYG and all. Produces much nicer HTML if started in HTML mode rather than writer mode, but even so in writer mode, it's chalk-and-cheese better than the abominations MS-Word spits out.

My own website, while hardly a paragon of usability or graphic design, is mostly built on OOW-edited HTML that's been fed to a gawk script which rips off the head and tail, replacing them with PHP calls to generic top-and-tail scripts which do the preamble, headings, menu, links-here, translation form (thanks Google) etc.

This makes consistency much easier, it's quick to edit stuff up (I use Linux, but that remains true even on MS Windows) and massaging MS-Word docs and the like to suit (precious few of those on my site but I do this elsewhere too) is fairly straightforward, although I usually have sed discuss some of the resulting HTML's shortcomings up close and personal before feeding it to the top-and-tailer. You'll notice that all of the W3C buttons work.

If I've just got to add an item to a menu or whatever, simple little tasks, it's vim all the way, and of course for a larger, more complex site I'd take a completely different approach.

I have read several posts attributing HTML as an invention of TBL.
This in my opinion is incorrect, the WWW dates back to 1980.
HTML is a derivative of SGML which dates back to 1960's and is
a descendant of IBM's GML.
Check this for some more history.
--
Han Tacoma

But you CAN link to characters 314-395 of document URL://whatever just use an XPointer ;-) , see http://www.w3.org/TR/xptr-xpointer/#b2b1b1b3b6b7 for example.

And to see how this could be used for annotation to documents that are kept separate to the document have a look at W3Cs Annotea page.

Well your current browser might not support that yet but at least theoretically its all there!

But you CAN link to characters 314-395 of document URL://whatever just use an XPointer ;-) , see http://www.w3.org/TR/xptr-xpointer/#b2b1b1b3b6b7 for example.

And to see how this could be used for annotation to documents that are kept separate to the document have a look at W3Cs Annotea page.

Well your current browser might not support that yet but at least theoretically its all there!