Internet Explorer 8 (IE8) freezes at connecting to site

I was really liking internet explorer 8 until I started having a random issue with opening new tabs and opening new windows freezing or hanging up.  I would sometimes happen when I first open internet explorer 8 or other times when I opened tabs the page would just sit at “connecting to site” and never do anything.  I searched the web off and on for several weeks and finally came across a blog post here at cornerpixel.  It was just a minor registry setting related to the loosely-coupled feature in IE 8.  I was getting desparate so I tried it.  Sure enough not an issue since.  The scary thing is I have seen this particular issue on 5 separate machines with IE8. 

MS you need to take note of this!

Internet Explorer 8 is out, Webmasters be ready!

I was intentionally avoiding loading up internet explorer 8 simply because I knew it was going to break some of my pages.  But today I couldn’t wait any longer since it was the official release.  Sure enough it was worse than I expected.  Internet Explorer 8 broke nearly every site I run!  Some to worse degrees than others, but most notable was the sites running the dojo script library.  It seems there is an issue with all versions except 1.3 RC1, even dojotookits site doesn’t work right.  The relevant error is below:
Webpage error details
Message: Member not found.
Line: 20
Char: 48151
Code: 0
URI: http://www.dojotoolkit.com/sites/all/modules/dojo/dtk_build/dojo/dojo.js
Hmmm.. Bummer, so after some digging it seemed the IE 7 compatible mode fixes the issue totally, so it must be a dom rendering issue, because I find it hard to believe the scripting engine is also changed that significantly (I could be wrong).  At any rate, my headache went away quickly once I discovered there is a meta tag you can use to force IE8 directly into compatible mode when it hits your site.  Add this meta tag in your head section:

<meta http-equiv="X-UA-Compatible" content="IE=EmulateIE7" />

This probably shouldn’t be considered a permanent fix as there is probably a good reason as to why you are having rendering issues, odds are it is the same in safari or firefox.  However, script errors that are IE 8 unique that will take some digging to figure out, but I intend to fix them and remove the tag when able. 

On the plus side, IE 8 does seem quite a bit faster and cleaner.  So hopefully it will be a positive move. I’m all for standards compliance too!

It is cuil to be ironic

So I made a post a few weeks ago basically ranting about Cuil search engine and how it pretty well was a failure.  Anyway, I thought it was funny and ironic the search result that hit my blog post a few days ago.  I have attached a screen shot.  Take note of the first result, the phrase searched for, the image displayed, and the text of the result.  Be sure and read my post about cuil and proceed to get a good laugh.

cuil-sucks

Preload images in IE6

I spent a significant chunk of time today diagnosing an image preload bug with IE6.  Bug probably isn’t a fair word, but for a browser that is one big bug why not.   In short preloading worked as it should in every other browser I test against except for IE6.  IE6 would preload the image, onload event fires, I set the src for the image and it continued to download the image again! 

Doesn’t work

pic1= new Image();
pic1.src='/image.gif';
pic1.onload = function(){docimage.src='/images.gif'};

Works

pic1= new Image();
pic1.src='http://site.com/image.gif';
pic1.onload = function(){docimage.src='http://site.com/image.gif'};

*sigh*, who would’ve thought.  You MUST have the full path in the  src link or IE6 won’t preload it.  It really crossed my mind today to just bite the bullet and not support IE6 at all.  It still appears that about 25% of our visitors are IE6.  Don’t even get me started on the hoops I have to jump through because of no png support.

Tagged: , ,

Cuil isn’t so cool anymore

So for those that aren’t familiar Cuil (pronounced Cool) is a start up search engine founded by some former google employees. With a pocket full of venture capital funds they set out to compete with the big G.  My initial response to this was, “Great, finally some really competition to google”.  Sadly the opposite has been true since Cuil has hit the map.  July 2008 I gave Cuil an honest rundown and was simplynot impressed with the results.  Most notably is Cuil’s attempt to provide images next to search results.  What I found amazing about this is they simply would show images that not only didn’t reside on the site the link pointed to, they didn’t have anything to do with the result at all.   This fact alone is a mammoth failure.  Even today (nearly 6 months later) nearly every page has an image that really shouldn’t be there. 

Getting past the borked image feature I took note of what results the engine actually returned.  Some keywords return relevant results, but it appears the fundamental way Cuil works is inferior to the likes of Google and Yahoo.  For instance, a simple search on Widgets google typically leads to a diversified listing of sites all focused around the Widgets you search for.  However, Cuil seems to feed multiple deep links to the same website, much more so than Google.  Cuil also returns the same results on multiple pages.  This appears to be widespread with all of my searches.

There is also the issue of hacked, faked, and general clutter sites listed on Cuil.  www.xuxkdmf.com/skdflwldkf/akdjf-widget showing up in the first page results for widget search makes me sick.  Mal ware spread would be unstoppable if google didn’t keep up with these listings.  With the all to often issues such as MS08-078 it is even more important to keep these scum sites from being found.

Cuil could have been a real contender to Google, who knows they still may be down the road.  My advice to Cuil, ditch the image feature because it doesn’t work and do like MS and Yahoo (copy off of Google).  Filter results down to be more unique, less of the same sites.

Tagged: , ,