March 2008 Archives

In a video interview I came across on 37signals, Ira Glass, host of This American Life, explains his storytelling process and two points in particular stood out to me:

“The amount of time finding the decent story is more than the amount of time it takes to produce the story.”
“Not enough gets said about the importance of abandoning crap.”

These two ring true with everything we do in our office. It’s not obvious to the visitors of our sites just how much time is spent on doing the initial research, finding the story, building the information architecture, deciding on the back-end technologies or creating the design mock-ups. The actual build-out of a site, the stuff that people end up experiencing, goes much, much quicker.

That’s only half the battle. Even if you do all the initial work, you still have to edit out the “crap.” You have to be ruthless and only show what needs to be shown in order to get your information across with as little static as possible.

Anyone can tell a story, or build a Web site. The difference between a good one and a great one is not about what has been put in, but what has been taken out.

The entire interview: Part 1, Part 2, Part 3 and Part 4

See also: Stories smooth as Glass (Mizzou Wire)

I'm a big fan of Google Analytics and am always on the lookout for information that helps me use this tool to better serve the needs of our online audience. I found a great blog post from the .eduGuru who recently attended a Google conference and it has some great tips for using Google Analytics.

I originally posted this to our internal Web Dev mailing list, but Josh Nichols suggested that I also publish it here for our external audience.

First, Microsoft has released the developer beta of IE8 : http://www.microsoft.com/windows/products/winfamily/ie/ie8/default.mspx

As with previous versions, it overwrites your older IE installation, so you can not have it installed along side IE7. Personally, I would suggest installing it in a VM. What I have done is installed the Internet Explorer Application Compatibility VPC Image for IE6 (http://tinyurl.com/y64upm), made a copy of it, and then in the copy installed the beta of IE8. That way I have IE6 and IE8 running in separate VM’s and IE7 actually installed on my computer.

*update* - Microsoft has released a separate VM of IE8 : http://www.microsoft.com/downloads/details.aspx?FamilyId=21EABB90-958F-4B64-B5F1-73D0A413C8EF&displaylang=en

The biggest change I have noticed so far is the way it handles forms. From the sites I visited, it renders legends (with no additional CSS applied) inside of fieldsets differently than in previous versions. It also seems to be adding extra padding to button elements. They've added further support for CSS and will, supposedly, completely support all aspects of CSS 2.1 by the time IE 8 goes final. They've also improved Javascript speed and increased IE's compliance with the DOM spec.

Microsoft has a complete list of changes available here: http://msdn2.microsoft.com/en-us/library/cc288472(VS.85).aspx

PCMag has a complete review of IE8 available here: http://www.pcmag.com/article2/0,2704,2274126,00.asp

Second is that microsoft has flip-flopped on IE8's rendering mode and instead has decided that IE8 will render according to the DOCTYPE declaration in the page. Initially, Microsoft had announced that IE8 (and all future browser versions) would default to IE7's rendering mode. So even if your page had a doctype of XHTML 1.1 strict and the code was completely compliant, IE8 would render the page the same as IE7. If you wanted your page to render in the new IE8 standards compliant mode, you would have to add a custom meta tag to your page to tell IE8 to render as IE8. Supposedly, it was to ensure that pages across the vast Web wouldnt "break" once IE8 comes out.

I, for one, applaud microsoft's decision to look forwards towards standards compliance instead of looking backwards to what has been a frustrating mess.

Third is that the Web Standards Project has released the Acid3 test: http://www.acidtests.org/ If you aren’t familiar with the Acid tests, they are test pages designed to test a browser’s ability to render HTML and CSS (and with Acid 3, ECMAScript and the DOM) according to the W3C specifications. Acid3 features 100 different tests and “grades” a browser by how many of the tests it passes. The following page charts a large collection of browsers (and multiple versions) and how they fare with Acid 3 : http://www.anomalousanomaly.com/2008/03/06/acid-3/