Or rather, they don’t do Web standards, either.

On a previous day, I summarized the standards compliance (usually noncompliance) of type sites. Long live Porchez for understanding his medium!

Now. What about design sites of the K10K ilk?

I don’t understand their purpose. I think they’re nice to look at, but really, that’s a positioning statement: “My site about design has to be well-designed. Q.E.D.” Nonetheless, they border on useless. Tiny links in iframes, usually given without enough annotation to clarify why you should care, barely constitute “content.” They’re designed around a kind of short attention span. The sites’ appearance, and much of the work they point to, are inexplicable even if you consider the tradition of personal work. They, and it, are simply insubstantial.

However, I’m not arguing that the opposite pole – turgid, overwritten design treatises – is any better. I’ve spent a good ten years trying to kill off that kind of design writing. Tarted-up illustration galleries are at least honest – though if these sites were thoroughly honest, they’d do everything in Flash or simply post GIFs (or, to continue their bias toward the pluperfect, PNGs).

In essence, design sites are about as informative as those failed old portal sites were. But since they look so very much nicer, well, you get to tell yourself you are hip, cool, and recherché when you look at them.

En tout cas, I guess the design is the content. But they’re using a structured document medium to deliver the design. And there are rules to follow. We know you’re all visual ’n’ shit, but we still expect valid code.

Is anybody giving it to us?

Not according to batch testing. My esteemed colleague Bryce Johnson ran a few dozen sites through the CSE validator, and none of them passed.

I then laboriously ran all the sites through the official W3C validator, and one of them did pass: NewsToday. Mazel tov!


  • Sites were chosen based on my own hip, cool, recherché knowledge and the listings at the misleadingly-named Stereotypography.
  • Individual iframes were not validated. This, by any measure, is a limitation. (Look at it as homework for somebody else.)
  • Many sites had no DOCTYPE (“ND” in the column below), and many of those needed a character encoding imposed upon them before they would validate (“+8859-1” or “+8859-2” below).
  • The two validators didn’t agree. Even though the W3C validator has occasional known faults, it is more trustworthy, a policy I’m sure the owners of NewsToday would support.
Site CSE errors W3C errors W3C faults
All Maple 7 100
Annuo 21 74
Archinect 5 43 ND
Attico 2 8 ND
Australian INfront 3 14
BD4D 8 188
Cobalt Revolver 2 163 ND
Core 77 21 161 ND
Coudal 17 87
Creative Behaviour 1 76
Deformat 21 40 ND
Design Made in Germany 404 20
Digital Ultras 2 24 ND+8859-1
Experimental 18 353 ND+8859-2
Half-Project 21 76
Infourm 4 15 ND+8859-1
Kiiroi 4 71
Linkdup 11 27 ND+8859-1
Moluv 21 314 ND+8859-1
NervousRoom 11 15
NewsToday 2 Nil!
Now Wash Your Hands 2 62 ND+8859-1
Pixelsurgeon 3 14
Scene360 21 55 ND
Stickernation 3 229 ND
Surfstation 3 18 ND
Swedezine 5 97 ND+8859-1
3D Realms 3 960 ND


  1. Most sites tested have trivial infractions that could likely be cleared up in minutes.
  2. Some sites could divide their cavalcades of errors in half if faults found early on in the source code were corrected.
  3. Design sites are viable candidates for standards compliance since they
    • tend to use oddball HTML structures like iframes, and
    • employ a lot of pictures of text.

    Getting those right is a challenge, which is exactly why they should do it.

  4. All the reasons articulated in my previous posting apply here: Sites dealing in visual artworks are not exempt from the need for standards compliance.
  5. No, Didier, ol’ pal, this does not constitute some kind of vendetta against or excoriation of design sites. I’m just running a few facts by everyone here.

Now, what about K10K? A mere three errors, fixable with a custom DTD. Still the king of the jungle. But look out, lads, for NewsToday – it packs quite the dagger when it snuggles up in bed.

Fact-check my arse

This study involved a lot of automation and also a lot of cutting and pasting (using CopyPaste, natch). I expect that errors will have been introduced. Do let me know of any. Among other things, I’m not sure I rendered each site’s name consistently, let alone correctly.


You can validate the sites yourself – if you have, for example, an urge to keep tabs on them, or what have you – by following these links:

  1. All Maple
  2. Annup
  3. Archinect
  4. Aticco
  5. Australian INfront
  6. BD4D
  7. Binah
  8. Cobalt Revolver
  9. Core 77
  10. Creative Behavior
  11. Coudal
  12. Deformat
  13. Design Made in Germany
  14. Digital Ultras
  15. Experimental
  16. Halfproject
  17. Infourm
  18. Kiiroi
  19. Linkdup
  20. Moluv
  21. Nervous Room
  22. NewsToday
  23. Now Wash Your Hands
  24. Pixelsurgeon
  25. Scene 360
  26. StickerNation
  27. Surfstation
  28. Swedezine
  29. 3D Realms

The foregoing posting appeared on Joe Clark’s personal Weblog on 2004.07.28 14:19. This presentation was designed for printing and omits components that make sense only onscreen. (If you are seeing this on a screen, then the page stylesheet was not loaded or not loaded properly.) The permanent link is:

(Values you enter are stored and may be published)



None. I quit.

Copyright © 2004–2024