2009/03/20

(null) requests from IE8

I've noticed since the release of the IE8 Beta 2 some strange request for non-existing pages: the requests are something like this "http://www.mydomain.com/(null)"

The only thing in common between those request is the user agent, always IE8 using the Trident 4 engine:

Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0)

It might include other toolbars or add-ons, but that's the common part for all of them. Most are using XP, but there is also some Vista requests.

The only page that I've been able to find so far is this thread that is focused just in how to ignore those request, instead of trying to understand the reason why are generated.

I'm afraid that now that the final IE8 has been finally released the number of such bogus "(null)"  requests will increase at a very fast pace, so the only solution will be to set a rule to ignore it as just another bug in IE8 and don't get flooded with the 404 error.

2009/03/15

The pain of supporting IE6

In this day and age it's a pain the fact that IE6 is still out there. Every web developer that have spend a little time trying to make a page or script work cross-browser and include IE6 in the set of supported browsers knows it.

I guess that if everything goes right and MS doesn't make any big mistake the release of Windows 7 will be a tremendous help in order to see IE6 go away. We all know that too many companies have been holding on XP due to the bad words about Vista, but for the moment they have managed to get good press for W7, so it won't be strange that not too long after it's released more and more companies that have been delaying the upgrades to their workstations or that have bought new ones but immediately replaced Vista with XP will be more positive to get Windows 7 and that way the included IE8.

Movements like the one started in Norway are a step forward, most of the average user doesn't know anything about internet or a browser, so when they see a similar message in every page that they visit they might get a clue and finally update to IE7 or (if we are really lucky) to any of the other browsers. Of course this isn't the first movement, there are previous ones we just need to send the message out, it doesn't matter where you did read it. The statistics show that following the current trends IE6 should slip to something like 10% usage at the end of this year. And I guess that the release of IE8 will help that number to get even lower.

As stated too many times, the usage between home and office is very different and the message usually only will be able to force the change at homes, but what happens when they go back to the office and their browser there doesn't behave the same way that it does at home?, when they can't use tabs? when the pages look slightly wrong? At some time they will be able to force the change of policies and get the IE7 update and that's good.

But this isn't just about the users and their browsers. If you are a web developer you can help every body: don't create pages in Quirks mode. As soon as you add the correct doctype to your pages you have done the right thing, it will be much easier to create pages that will behave the same way in different browsers and will make some of the IE6 problems go away. Of course it isn't perfect and IE6 must die, but you will be able to forget about it for a little while. You can read about the goodness of Standards mode by a CSS Expert

What does it mean that you use standards mode for anyone else?

As long as you use any third party library, a lot.

If the developers of any javascript library can forget about supporting IE6 and its Quirks mode it means that they have one browser/mode less to test on. Not only that, every other browser also have slight differences in Quirks mode, so it means a lot of time saved just by a little change to your pages. As you see, you don't need to force the change on the users to make the web go forward, just say no to Quirks mode.

If you want to use an updated script for your page, why can't you update your page so it follows the standards?

More voices are starting to cry out "that's enough" and some have suggested setting Dec 31th, 2009 as the end date for IE6 support. I think that's a good movement forward: we can't force people to change their browsers, but we can't waste our time and money supporting such a beast that doesn't allow us to focus in newer features, simpler code,and in general look forward instead of being forced to code the same way that we had to do eight years ago.

I will be glad to hear from everybody that they won't be supporting new developments on IE6+QuirksMode any longer, the sooner the better.