Arranged Delirium

One great thing about web development over the past couple of years, is that we’ve finally seen a breakdown of the IE hegemony! IE is still over 60%, but at least most folks are running version 8. (Really? Why are you still using IE?) But its made testing and developing web sites a bit more of a chore. Add in the recent explosion of varying mobile devices, and suddenly you’ve got a couple dozen viable operating systems and browsers to support. (Or, at least to think about supporting.)

As a professor at a small college, its become beyond my abilities to support and maintain multiple testing kits - even with dual or tri-booting machines, it’s too time consuming. So, I’ve decided to outsource. There are two compelling services that I know of off the top of my head: Litmus and CrossBrowserTesting. I decided to test both, and both are quality services. I think that if you were running email marketing campaigns, Litmus might have an edge for you. For general web development, I found both services adequate. In the end, I decided to go with CrossBrowserTesting, primarily because: 1) it was cheaper, 2) did everything I needed, and 3) had a more robust free trial period which made for a better evaluation.

Best of all - this is going to speed up certain elements of my grading in my web development courses…

Other Notes: Most everything works on the Mac OS with each browser. Linux/Unix boxes running Firefox are all good; some of the ancillary browsers are problematic, but do a good enough job of rendering most valid CSS and constitute such a small market percentage it’s not worth addressing. Windows machines are much more problematic; but if you are running a recent version life is better. IE versions before 8 remain the biggest problem. The question now is: do I add browser detection and route folks to a different style sheet, or just forget about it? IE 6 & 7 make up about 24% of the general web viewing population, and about 8% of the folks that access my content.

As classes start back up, I am reminded of some of the particular content organization challenges that college web sites face; and I’m reminded of this comic from xkcd:

There seems to be some consensus regarding the disconnect between the messages being sent and the information requested. I wonder if there is some kind of data - perhaps we could call them “analytics” - that could tell us which pages are accessed, and with what frequency, and for how long? It might do wonders for helping us generate web content that serves identifiable and actual needs.

And I’m not talking about the W&J web designer here either - who actually does have analytical data. It still comes down to the folks who make decisions about content and whether those decisions are based on data or other inexplicable intuitions. Intuition can actually be valuable in terms of design - but content creation should be based upon data.

A few weeks ago, I presented on web development at the AIIP conference. I had presented there before, and I’ve always found it to be a really friendly group, and this year was no exception. After the conference I received several emails (and I apologize for my slow reply - it’s the end of the academic year!) with some really thoughtful comments and very considerate commentary. The event was very useful, because it has solidified my intent to write a text for web development courses. Currently, I need to employ several different texts in my intro course, and I’m convinced now that that can be done comprehensively in one title (sans technical reference).

As students in Web Design and Development start wrapping up their group project work, now would be a good time to reflect on all the little details that need to be worked out before the project is actually finished. Smashing Magazine to the rescue!

Here’s a link to an article that covers a nice lists of checks to complete before you can really consider your site to be “finished.” Four our class, the sites are a bit more basic - and no-one is using a CMS to build pages - but many of the 15 suggestions still apply. Folks working on more advanced projects int he capstone course might also find this to me information of value.

PS: Make sure to take the time to peruse the contents listed on the Smashing Magazine homepage. There are some great articles up there right now addressing issues of usability, an interesting Jquery list, and some nice detail on IE and the challenges for writing CSS3 code

I’m frequently talking in class about what a designer can count on, and what’s out of our control. In web development, fonts are a good example. Even as you specify fonts, you cannot necessarily assume that the viewer’s machine has those font installed. Or, that they will render similarly - particularly given the differences of platform and browser. I think the image below from a post at The Small Wave makes the differences clear. Both of these render form the same code, but obviously appear different:

Now I know that some will argue that I’m just trying to applaud the Mac here - but I’m really not trying to do that. This is all about being able to see the fact that fonts render differently on different machines and/or browsers.