Crickets Chirping

Steve Yegge’s Blogger’s Block #3: Dreaming in Browser Swamp is a stream of concsiousness post that hits on a lot of things that I think about: programmers’ (often snobby, and frankly shortsighted) attitudes towards new languages, how dreams affect your creativity, the social side effects of working with people who know you blog, the recent ascention of JavaScript, and AJAX hype vs. utitlity.

Behold WikiCalc (check out the screencast*). It’s a cross between a wiki and a spreadsheet, written by Dan Bricklin. He was one of the creators of VisiCalc, which was the first spreadsheet application -before Lotus 123 or Microsoft Excel. I think WikiCalc (or something like it) has HUGE potential.

I don’t know how many times I’ve run across critical applications that written as excel spreadsheets. From humble beginnings, they grow in importance to the company and eventually somebody wants to put it on the intranet. And this is where I usually say, “man, if only somebody had a way to just upload this damn thing to a web server and turn some of the cells into form fields.” I’ve done some unnatural, ugly acts such as calling out to Excel COM objects from HTTP request handlers just beause that was easier than porting the spreadsheet code into C# or java. Often enough, nobody knows exactly what the spreadsheet is doing becuase so many different people have edited it over the years. Plus they want to be able to edit that spreadsheet after the app had been deployed to the web!

Excel has to be the most common Decision Support System in use today. I would be suprised to find out that Sharepoint didn’t support some kind of WikiCalc-like functionality. I’ve played with Google Spreadsheet, which is underwhelming in its current form- just a spreadsheet on a web page. Here’s a CNet article offering some context for WikiCalc amongst Excel, JotSpot and the like: Software pioneer Bricklin tackles wikis.

* Speaking of screencasts I was thinking of starting a Screencast Review site. I (and anyone else) would watch and rate screencasts, and write reviews. Like Siskel and Ebert for screencasts. this is sort of in that direction but it’s just a message board at the moment.

Thinking back on it, I’ve already pointed out a few screencasts on this blog: DabbleDB and SQL on Rails. There are lots of others that I could write about, like Heavy Metal Umlaut: The Movie, and of course the original Rails Movie.

Screencasts are growing in popularity, and like any other video content that demands your attention, having a filter to sort through what’s worth watching and what isn’t would help out a lot.

There are some problems with AJAX that lots of people complain about. For instance, it’s very unfriendly to people with accessibility needs and browsers on mobile platforms. It also breaks the back button (although that’s a pretty much solved problem at this point). Adam Alex Bosworth made a good list of other tricky issues.

But there are some others that are pretty tricky and not for obvious technical reasons.

AJAX and Ads

If you’re building a business around a web site that offers a free service you’re probably relying on at least some income from banner advertising. Guess what? The banner ad network thinks you’re only getting one page hit per user session even though you’re getting 100′s of clicks. None of those clicks triggers a new page load since your site is all AJAXified so the ad servers don’t know about them. You get paid for a single impression even though the user’s been pounding away on your sever resourcs for six hours.

Aint that a bitch. You want to use AJAX but it kills your ad impressions. It took like ten years for internet ad firms to come up with standards like CPM and banner ad sizes. How the hell are they going to figure out what to pay you for a “tab switch” or “scroll down” click? An AJAX action that repaints 25% of the visible pixels in the browser window should get you 25% of a page view? Good luck trying to bill them for it.

• Forbes: Nielsen/NetRatings doesn’t know how to measure your AJAX traffic.
• AJAX Counting Nightmares – This guy Eric Picard has been writing about the problem for a while.

There’s another problem which componds the first but it might be easier to deal with since it doesn’t require the internet advertising industry to agree on something.

AJAX and SEO

You want to drive lots of traffic to your site from search engines. The best way to do this is to doll up your site for when googlebot comes strollin’ by. He’s got strange taste so that means doing some odd things to your html. Odd things that most people don’t want to see or know about. This is known as the black art of Search Engine Optimization or SEO for short. (Put on your boots because the BS gets thick around the “SEO Expert” crowd. It’s like real estate investment seminars or multilevel marketing, but with a highly technical twist.)

How is AJAX problematic for SEO?

Your javascript widget generates html inside the browser window. You want google to know about this html content so it will point people to your site. But GoogleBot doesn’t run the javascript that generates the html content in your widget, so it never sees that content and never indexes it. You must make this content appealing to Monsieur GoogleBot, and he likes static links to static-looking resources. Javascript ain’t so static looking to him so you have to slap some makeup on it to make it look that way.

Some interesting thoughts on how AJAX and SEO might play together nicely:

• SEO Considerations for AJAX Development
• AJAX and SEO- includes a demo of how to make AJAX links look more static