The Net is Dead

Marco van Hylckama Vlieg on web technology, design, life and everything

Commentary

Join the discussion! Leave a comment through the comment form below!

• At 19 January '06 - 09:00 jazzle wrote:

  

  Great point – why does it matter if code is perfect in this situation.
  So, a space rocket blew up because of a single typo, but this is 2006 and computers are becoming more human. (By which I mean they can cope with, and more importantly ignore, small errors.)

  
  

• At 19 January '06 - 09:26 weefselkweekje wrote:

  

  On the other hand, there’s this movement right now that pushes writing clean code. It’s a momentum that will in the long run influence even people like microsoft’s IE team…
  
  Why give up being an evangelist? Why not change Pivot so it outputs the right characters? Why not let your site be one more reason to get these little things sorted out instead of yielding to them?
  
  The web really needs a presentation language to do it justice. We’ve been meddling with (non-x)html for way too long…

  
  

• At 19 January '06 - 11:12 Marco wrote:

  

  Let’s clarify things a bit here.
  
  I’m not ‘giving up being an evangelist’. I am still a web standards advocate. I’m promoting pure CSS based design and modern DOM scripting like any other ‘evangelist’.
  
  The point I was trying to make is the fact that I don’t see much use in tinkering with totally unimportant things such as small character issues. The ‘result’ just isn’t worth the effort. No visitor will ever see the difference and the site won’t ‘break’ in any browser if there’s one or two ‘wrong’ characters in it. I could spend this energy better on improving the site, writing articles and doing other things nice to the site. I’m quite sure my readers will appreciate that a lot more than a 100% clean xhtml source.

  
  

• At 19 January '06 - 11:41 bosmeneer wrote:

  Well, may be Pivot isn’t 100% XHTML-valid? The ampersand-problem is really easy to solve: just use htmlentities() !?
  And I think this log is valid when you change the ‘triple dot’-char (in the commentary-tab) to the Ascii-code 8230 (

  
  

• At 19 January '06 - 12:32 Eljo (Morpurgo.nl) wrote:

  

  Ik moet bekennen dat het er werkelijk schitterend uit ziet en het waanzinnig mooi werkt!

  
  

• At 20 January '06 - 03:04 Ray mosley wrote:

  

  I am 100% behind you….all too often this last year I have produced sites at my company that conform as much as they can. I actively seek to make my pages user friendly and valid but I have also seen other firms come in to try and take the business by running the sites through W3C compiler to find two

  
  

• At 20 January '06 - 04:14 Khevor wrote:

  

  I gave up making sure that my site was always valid. With a blog, it’s almost impossible to stay on top of total validity simply because of its dynamic nature. I had also given up using tables at all for a time but those are back now simply because of weird rendering issues I had with Firefox a few versions ago. By the same token, my site is really only viewed by a very small number of friends (both online and not) so I don’t worry too much about it anyway. Of course, you have a much larger audience than I do but the the principle remains the same. As long as it works the way it’s supposed to in all the target browsers you design for, why should it matter if your code is less than 100% valid?

  
  

• At 20 January '06 - 04:56 Marco wrote:

  

  Wellll… I still am strongly in favor of pure CSS based web design Khevor! I just had a look at your site and I figured your layout could be created without any tables quite easily.
  
  The advantages of pure CSS design are numerous. Less code, better SEO, better accessibility, the list goes on and on…
  
  But then again, the saying ‘whatever works for you’ goes as well of course. If your site was made just for a couple of friends and they can all properly view it I guess you’ve already reached your objective and you shouldn’t bother indeed.

  
  

• At 20 January '06 - 05:57 Andy wrote:

  Dude your site has 0 errors, but 92 warnings. I’m a perfectionist, so this botheres me.

  
  

• At 20 January '06 - 06:10 Marco wrote:

  

  @Andy> In fact it has two errors. Pivot inserts the ‘...’ character in the recent comments section and I still haven’t found where it does that. I didn’t see 92 warnings anywhere unless you mean the CSS.
  
  The CSS throws some errors due to an Internet Explorer proprietary thing I used and due to the cursor:hand thing. Both of these don’t cause ANY problems in ANY browser which is why I find them perfectly acceptable. Then there’s the CSS warnings telling me I have duplicate definitions for background colors etc. I could eliminate all those but for what? Tell me what use that has. The CSS is valid apart from the above mentioned ‘errors’.
  
  Anyway, all this is the kind of stuff I stopped caring about. Anyone can see from my sourcecode that this site hasn’t been built by some braindead amateur which is all that counts when it comes to the few that look at the source in the first place.

  
  

• At 20 January '06 - 06:40 Veracon wrote:

  

  There’s no text/xml, believe you mean application/xml.

  
  

• At 20 January '06 - 06:59 Marco wrote:

  

  There definitely is text/xml as a content type!

  
  

• At 20 January '06 - 07:36 Ronald wrote:

  

  But not all errors for a tidy document are necessary (like using different names and id’s for forms)... tidy is of course the least you could do :)

  
  

• At 20 January '06 - 07:42 Marco wrote:

  

  I’ll definitely tidy up the CSS. At this moment I’m still ironing out small glitches. Once things have fully settled down I’ll definitely go through the CSS to group things, comment things etc. and I’ll skim through the templates as well for some cleaning.
  
  However, all this is still not the true point of this article. It’s about whether it’s useful or not to keep track of any tiny little problem that may or may not occur over the months / years while running a weblog. It’s not about the importance of delivering a clean product when you launch it. On that issue I fully agree tidiness is important, if only for maintainability reasons.

  
  

• At 20 January '06 - 16:13 Volkher Hofmann wrote:

  

  Very nice redesign! Like it.
  
  I tend to agree with you. I always have problems with those pesky Amazon associate links and the ampersands. I’ve stopped bothering and the next redesign will see the disappearance of the “valid XHTML” button as well.
  
  It’s good enough to know that I’ve tried my best and will try to improve on it with my next design.
  
  There are more important things to lose sleep over.
  
  Keep up the great work here!
  
  Cheers!

  
  

• At 21 January '06 - 13:48 Christian Montoya wrote:

  

  Marco and Veracon, to correct you both:
  
  XHTML should be served as application/xhtml xml.
  
  text/xml is optional, and not what you meant.
  
  If you served your weblog as application/xhtml xml, it would not display. If you served it with an html 4.01 doctype, you seriously would not have to change anything, because self closed tags with spaces are exactly the error in html tag soup parsing that allows us to serve xhtml as text/html. I’m pretty sure I told you all this already through some discussion list, and I never did get your response on any of it.

  
  

• At 21 January '06 - 17:32 Andrew wrote:

  

  You’ve said the things that many people have secretly been thinking about.
  
  Your redesign looks pretty good! (...Except for some strange Google ad flickering in FF when I hover over links in the right column.)

  
  

• At 22 January '06 - 07:30 Hong Hui wrote:

  

  Hmm…i don’t know for sure. I wondered if the extra links at the bottom half of the main navigation panel should be there in the first place.
  
  I got confused for a while and thought that those links, with the same typos and font size, was linking to pages within this site.
  
  And how come the url to this entry is at the last of the navigation panel? :)

  
  

• At 22 January '06 - 07:59 Marco wrote:

  

  Hong Hui> that’s a VERY clever remark. I will think of some changes that make it clear which entries actually make you leave the site.
  
  As for that other remark: I’m not sure what you mean?
  
  Thanks!

  
  

• At 22 January '06 - 17:03 Koray wrote:

  

  sigh
  
  When Idiots join forces.

  
  

• At 22 January '06 - 17:06 Thame wrote:

  

  Marco, you have a wonderful design here and it was bound to get attention. At least you can find comfort in the fact that even if one of these losers rips your design, they don’t get any visitors.

  
  

• At 22 January '06 - 18:43 David wrote:

  

  this is ridiculous, i get tired of people dumping on others because they for what everreason have found that complete validation is not worth the expense in time. you would think the web sould break or something. the average user is not going to view source or have a little validation button in FF. while i try to do my best with validation its not the end all be all, sometimes i think people who have a perfectly valid site forget whats important. the content. so F*ing what if there is a non encoded ampersand. a few unencoded charachters isnt going to doom the net. and does not make you non professional.
  
  there i have thus invalidated his site even more ;)
  
  edit—well it seems i was tring to be a smart ass put intentional ampersands in to make a poitn but the comment did not let me d it. as you can see it blocked out half of my prevoius post. well there goes the neighborhood

  
  

• At 22 January '06 - 21:49 Creative Common License wrote:

  

  You are free:
  
  * to copy, distribute, display, and perform the work
  * to make derivative works
  
  HYPOCRITE!!!!

  
  

• At 22 January '06 - 21:57 Hong Hui wrote:

  

  Marco> Hi there, i got to think about this last portion of my comment and it indeed does not sound very clear.
  
  Allow me to rephrase myself. :)
  
  Let me guess, the last link entitled “Almost valid xhtml” is actually to inform ppl, like many blogs on the net with those cute little “valid “ buttons, that the site has conformed to certained standard but in ya site context, it is not yet valid. Kekeke…:)
  
  If that is the case, then may i suggest, in my previous post that you apply another set of styling to allow users to know that these links are not part of the navigation panel but just some notes of sort. :)

  
  

• At 23 January '06 - 14:06 Christian Montoya wrote:

  

  David, who in the world are you talking about? Who is dumping on who? I’m very confused here.
  
  Marco: still waiting for a response.
  
  Anyway, this is something I’ll be writing about in the near future… a return to HTML 4, a call for completely valid CMS’s and better information from the standards community. For the time being, I have to mention that Wordpress plus a text editor such as Markdown or Textile is very effective at ensuring that the markup does not break validation, though if you are already using some other platform there isn’t much you can do.

  
  

• At 23 January '06 - 14:46 Marco wrote:

  

  Christian> I’m not sure what the question is you want me to answer?

  
  

• At 23 January '06 - 16:30 Christian Montoya wrote:

  

  If you served it with an html 4.01 doctype, you seriously would not have to change anything, because self closed tags with spaces are exactly the error in html tag soup parsing that allows us to serve xhtml as text/html. I’m pretty sure I told you all this already through some discussion list, and I never did get your response on any of it.
  
  I guess the implied question is, why don’t you use an html 4 doctype?

  
  

• At 23 January '06 - 16:31 Christian Montoya wrote:

  

  Well, that was meant to be a blockquote.

  
  

• At 23 January '06 - 22:45 Marco wrote:

  

  Hmm well try validating this site at w3.org.
  When overriding the doctype and setting it to 4.01 strict you’ll get a lot more stupid errors than with the current doctype (which only throws two wrong characters).

  
  

• At 24 January '06 - 01:06 David B. wrote:

  

  Christian Montoya => i wasnt talking about you if that’s what you where getting at. on another note. anybody notince that some scriptaculous effects dont work in IE, like pulsate for example. also Marko IE is throwing this error: line 317 char 13 document.getElementByID(...).piv_remeberinfo is null or has not been set. something crazy like that. I guess its talking about the remeber this info checkbox.

  
  

• At 24 January '06 - 01:48 maxzilla wrote:

  

  Misschien is non valid XHTML niet zo erg…maar er zit wel een dikke javascript error op deze page ;)

  
  

• At 24 January '06 - 13:24 Marco wrote:

  

  @David B: I don’t know what IE version you’re using but the pulsate effect works fine here!
  
  I did notice that the script.aculo.us SlideDown effect crashes IE when I use it in my tabs in the right section. This is how I ended up replacing it with fades. I liked sliding better but IE dies like a squashes bug every time something finishes sliding down.

  
  

• At 24 January '06 - 19:44 david wrote:

  

  I have IE 6.02900 WIN XP. I thought it was just something i was doing wrong on my site so i looked arround and even on the scriptaculaous site itself , for me anyway, some of the effects dont work.i have tried restarting and getting on my lap top, but no dice. i use FF anyway but it worries me for most of the target audiences i design for. it all degrades well but all that work for nothing is what i am left feeling like.

  
  

• At 25 January '06 - 09:05 Tobias wrote:

  

  Off topic: about the scriptaculous slide ‘problem’ I came up whith a workaround in a project I am/was working on. I created an class that was visible to all browsers except IE and placed an

  
  

• At 25 January '06 - 09:06 Tobias wrote:

  

  this code should have had appeared in my preivous post <>.

  
  

• At 25 January '06 - 09:07 Tobias wrote:

  

  Ahh, forget it. Point is clear.

  
  

• At 21 February '06 - 15:01 Jero wrote:

  

  I agree with you on removing the buttons. First of all, HTML has grown from an SGML application to something quite different: tag soup. This forced modern user agents to support the biggest trash you can imagine, so the small validation errors you’re talking about obviously have little to no effect on how the page is displayed.
  
  However, I do support using valid markup, but not in a way as most other would. As I wrote on my website in one of my articles, valid markup has no meaning. I could write a page heavily abusing tables and divs and still be able to put those ugly “Valid whatever” buttons on my website, but it’s the quality of the markup (mostly called semantical markup) that matters to me.

  
  

• At 22 February '06 - 12:21 Christian Montoya wrote:

  

  Marco, if you still care you might be interested in: this article

  
  

• At 28 February '06 - 18:45 h3 wrote:

  

  Pretty frustrating that Microsoft doesn’t comply to W3C’s standards, a group which this company is part off .. I can’t believe that micro$oft ignores an entire coder communauty’s frustration just because an overpaid marketing dude there decided that xhtml 1.0 would be an IE7 feature rather than a IE6 logic update. Their creative CSS implentation is unpardonnable too.. if only we could boycott them :|

  
  

• At 24 October '06 - 13:25 The Faeriekeeper wrote:

  

  Excellent article or “vent” if the same had been written by myself. Having a web presence since 1986 and in the mid 1990’s working toward valid HTML and WAI compliance, I often feel the same.
  
  Yet in some small way hopefully an extra effort now and then will encourage others to see that the beauty and promise of the world wide web is really not for ourselves and our own vanity, rather for those who rely on the web, our disABLED.

  
  

• At 25 October '06 - 14:24 h3 wrote:

  

  I must correct something in my previous comment, now that IE7 “stable” is out.
  
  In fact, things are even worst.. XHTML is planned for IE8.
  
  Don’t hold your breath :|

  
  

• At 12 November '06 - 09:41 Bhavik Vora wrote:

  

  well yes i agree..
  nyways even the most reputed sites..
  google and microsoft.com arent even CSS valid !!!!

  
  

• At 15 December '06 - 02:22 Joe Volpe >> testing services wrote:

  

  the last link entitled “Almost valid xhtml” is actually to inform ppl, like many blogs on the net with those cute little “valid ” buttons, that the site has conformed to certained standard but in ya site context, it is not yet valid.

  
  

• At 20 December '06 - 01:31 Scott Brison - weight loss support wrote:

  

  I thought it was just something i was doing wrong on my site so i looked arround and even on the scriptaculaous site itself , for me anyway, some of the effects dont work.i have tried restarting and getting on my lap top, but no dice. i use FF anyway but it worries me for most of the target audiences i design for.

  
  

• At 25 December '06 - 05:49 Gerard Kennedy => Employee management wrote:

  

  This is something I’ll be writing about in the near future… a return to HTML 4, a call for completely valid CMS’s and better information from the standards community. For the time being, I have to mention that Wordpress plus a text editor such as Markdown or Textile is very effective at ensuring that the markup does not break validation, though if you are already using some other platform there isn’t much you can do.

  
  

• At 20 February '07 - 09:30 David Bannon wrote:

  

  You could use php to manipulate the finished page production..
  by scanning the http ACCEPT header, you can find if application/xhtml xml is allowed and if its q rating is higher than text/html. You can then output the header content-type.
  
  Also you could simply run the finished page through an output buffer to replace “ />” with “>”.
  
  Exactly what I do on my site!

  
  

• At 26 January '08 - 19:55 Sean wrote:

  

  Use a script to filter through any data. PHP provides it’s easy-to-use htmlspecialchars() function to easily run through any text that should be rendered as text and replace any “HTML characters” with their encoded equivalents… you really have no reason to HAVE any HTML characters that aren’t encoded. As I look at this page, it doesn’t even seem that the comments that users post are filtered, which could lead to XSS injections?
  
  Not only is filtering a great way to abide by certain rules (w3), it’s a great way to make sure your site is rendered in all browsers the same way, and to keep text as text.

  
  

• At 27 January '08 - 05:15 Marco wrote:

  

  The input through comments on this site is definitely being filtered.

  
  

• At 27 January '08 - 05:42 David Bannon wrote:

  

  Wow, thats quite a bump sean =P.

  
  

• At 24 March '08 - 07:37 Briongloid wrote:

  

  I think it’s important to make sure that if your site is not 100% valid, then at the very least the number of errors should be a single figure.
  I agree though, that this is not always possible.
  
  The whole thing gets thrown out the window anyway if you want to add any third party code, such as a Google Map or a Paypal button.
  
  I have, in the past, gone through the code and changed

  
  

• At 21 November '08 - 10:55 Marc K wrote:

  

  http://www.i-marco.nl/weblog/archive/200..
  
  Well, I’ve just viewed the source of this page and it made me smile when I saw the scrap of Google AdWords code inlined but not escaped with CDATA[...] — all this fussing around frustrates me too. I often think of modifying the XHTML Strict DTD very slightly just to allow what I perceive to be perfectly rational things like and elements automatically being set to CDATA inside by default (why not? I don’t want the same parsing engine dealing with it, damnit, only the CSS or JS one. Seems like an obvious oversight, unless there’s something I’m missing.)
  
  To touch upon another debate: I think the same applies to XML character entity escaping for URLs in attributes. Why is this necessary at all? That bloody default PCDATA declaration again… Seems to me like the very quoting of the attribute value itself should instruct the browser (sorry, user-agent) to hold it back from the XML parser and let the guts of the application handle the URL, which is invisible in the document anyway (unless it’s actually placed as node inside the Anchor element, perhaps

  
  

• At 21 November '08 - 10:57 Marc K wrote:

  

  Nooo you didn’t tell me about comment length truncation! I’m glad I dumped it into a text editor… Argh there goes my tab-heavy instance of Google Chrome messing with the text field again. I wonder if it’s a noted bug…

  
  

• At 02 July '09 - 04:43 Manuel Graf wrote:

  

  True.

  
  

• At 28 October '09 - 18:51 Jessica Brown wrote:

  Oh, how I wish there was a “valid at the time of design” button.

  
  

Got something to add to this?

Feel free to leave a comment on this site. You can use Textile and Emoticons. Your email address is only used to show a gravatar. Please stay on-topic and use common decency. Spammers will be shot in front of a live studio audience.

If you plan on posting code, use pastebin please and post a URL to the code. The comment processing doesn't deal very well with code. Sorry for the inconvenience.

Human comment spammers: don't bother posting your crap here. Comments are moderated and I won't let any of your shit through.

Trackbacks

• Earlier musings
• Mass Effect - The best Space Opera of all time (2)
• Yup, Yet another JQuery Accordion Plugin (81)
• Week In Links: a WordPress plugin for link digest posts (1)
• Technorati: What happened to you? (1)
• Some random thoughts about the iPad (1)
• Bookalicio.us version 2.0: an attempt to design the prettiest book blog on the net (3)
• Me, version 2.0. Released! (15)
• PreDevCamp SF Bay Area at Palm HQ - a videoblog (6)
• Me, version 2.0, almost done (2)
• Why I choose the Palm Pre to be my next mobile phone (6)
• What's up with the blogosphere lately? Tired of list-posts! (11)
• Create your own Twitter image badge with PHP (13)
• Five reasons why I hate Facebook (with a passion) (176)
• 5 things MySpace could do to improve Profile 2.0 (1)
• MySpace looking mighty fine with Profile 2.0 - How I built a MySpace profile that doesn't suck (3)

• Tumblr
• Popaganda
• "I come across as miserly because after nearly dying I have realized that life can be taken away at..."
• "I refuse to believe in a god who is the primary cause of conflict in the world, preaches racism,..."
• Photo
• somethingintellectual: This too, kay? The zombie apocalypse is...
• Search terms in the access log of Bookalicious.
• Photo
• "Moderate drinking, which is defined as one to three drinks per day, is associated with the lowest..."
• *SIGH*
• Hope. Delivered.
• Gran Turismo 5 porn
• The best mobile OS on the planet is getting even better!
• stfuteabaggers: revolutionnow: technipol: Clay Bennett does...
• Photo
• visit me on Tumblr

• Tag Universe
• ajax blog bloggers blogging blogosphere blogs browser browsers clock cms code comment-spam comment_spam css design dutch firefox flickr google hype internet islam javascript mac mint pagerank performance photography php php5 pivot pivot-blacklist pivot_blacklist powerbook programming prototype rails referrer+spam referrer_spam rss spam spammer spammers taggerati tags technorati template trackback typo variable web+2 web_2 web_development weblog weblogs website wordpress xhtml yahoo yui (all)