Developers of the World, Unite! IE6 end-of-life is near.

This is a call-out to web developers of the world:

End support for Internet Explorer 6.0 on all new sites!

Several norwegian sites have already lead the way and are warning IE6 users that they should upgrade. The more developers and sites who join in on this the better. Christopher Blizzard proposes on an actual deadline:

Support for Internet Explorer 6 ceases on 31st. of December, 2009.

Should your clients want support for IE6, advice them that Microsoft already supports this campaign and that several free alternatives (i.e. Firefox, Safari, Opera) are available for the users. Asking them to upgrade is not just a question of supporting your site, but just as much as improving security and user experience.

We need to spread the word as much as possible so you could blog about the campaign, join the facebook group or just talk to your friends.

You can get more norwegian news on

How to begin

In the coming days, I will create a project page containing a guide on embedding a non-obstructive warning message to IE6 users, so all you need to do on your sites is embed this and you're playing an important role of the campaign.

Nailing the coffin

Microsoft just discovered a potentially dangerous bug. Heise-online has a thorough explaination of the issue.

So what's the bug? In short terms: IE6 and 7 handles images based on a mixture of MIME sniffing, content-type header and byte-header sniffing (alias signature). When extension, content-type and signature disagrees the MIME-type kicks in. The issue arises when the MIME-type indicates HTML and it tries to act on it.

So what's the danger? Websites that naively displays images from users may be vulnerable to cross-site scripting, phising attacks or other indirect attacks. If your websites doesn't check the image for consistency (which it should) you may make it vulnerable to these "attacks".

What's dangerous is not the fact that there's a bug but that Microsoft only intend to fix it for Internet Explorer 8 leaving a lot of people vulnerable. In all fairness they are trying to promote IE8 but one might argue there are better ways of motivating people.

Luckily the bug isn't that easily exploited. Let's hope the crackers of the world are feeling lazy for a few years…

IE8 RC1 announced. Silent cheers seem too loud.

IE8 RC1 has been anounced but where are the joyful cheers and the popping champagne?

First, IE6 is still big
Well we still have IE6 to support with and since they decided not to support any new CSS3 features at all we still lack being able to make a modern site without the use filters (something as simple as the opacity property is still missing).

Second, IE7 spread slowly
We're still waiting for customers to take IE7 more seriously than IE6. In Denmark the tendency is still not leaning enough toward IE7. Knowing how long it took for IE7 to spread we can expect at least 3-4 years before IE8 reaches the same level as IE7.

Third, A new browser to support
IE8 introduces more IE-specific selectors which mean making opacity and fading effects now requires more lines of code. Good thing we got jQuery to do that ­čśë

But on the bright side IE8 will support CSS2.1 completely. This means that Microsoft will hopefully dedicate more time on bugfixing. IE8 is definitely a step forward but Firefox 2.0 would still be considered leaps ahead compared.

New blog launched

Finally. I did it.

I switched to WordPress instead of Drupal. Much can be said about using Drupal for flexible websites but when it comes to usability and blogging, WordPress is the undisputed weapon of choice. Admittedly, drinking ginger tea from freshly chopped ginger and blogging at the same time has never been easier. Kudos to the people who belived in the ideas of WP.

I'm one happy camper.

And for those who are searching for my old projects: I will be putting up redirecting pages in the comming days so noone will bounce on a 404.

New Mac OS X Plugin for jEdit

After switching to Java 6 I encountered a couple of problems using the Mac OS X plugin for jEdit. After waiting a couple of months to get annoyed with the missing mac integration I finally got my act together and re-coded the plugin to support the API changes.

I've managed to get most of the functionality you would expect from a regular mac program working and have now decided to release an alpha version for any testers that might want to get their hands dirty

Here's the changelog and features:

  • Initial release
  • Support for drag and drop onto the dock icon
  • Support for Open With for files that are associated with jEdit
  • Application menu integration
    • "Preferences" will open Global Preferences
    • "About jEdit" will open the Help -> About jEdit… dialog
  • CMD-click in titlebar now displays the current buffer path in a drop down. Clicking on a folder will reveal that folder in the finder.
  • The red close button now indicates the entire view's modification status. This way it will be consistent with the "save modified buffers?" dialog.

Kris (the original author of the mac plugin) is not completely without credit on this one since I used his plugin as a reference of what needed to be done.

Download plugin

Please give your comments

Greasing up potential MyC4 investments

Being an investor through I spend some time browsing through potential investments/businesses to find some eligible ones.

One aspect of the investing is how high you set the rate of the money you invest. It's all relative to other bidders but as a general rule: The higher the rate, the higher risk you take of someone else under-bidding you.

Another aspect of choosing a suitable rate is how much money you put in and how much others have put in at different rates. The less money there is above (at a higher rate) the more likely you are to be under-bid. So usually you weigh these variables together to form an opinion that suits your mood.

In order to better get an overview of these variables I've created a crude GreaseMonkey script that accumulates the money "above" each bid in the bidding-page.

Download script