[ismarc]

I think this is really well made and only have a few comments on the speed test portion.

I'm not sure if you scale the dial showing your speed, but at 100 words per minute, it's not maxed out, but it is at 123 words per minute (I'm assuming the max is 120 words per minute). I'm assuming so that people can target 60 wpm? I would actually suggest it go either up to 100 or have it go up to where 50% would be the target or average, and maybe scale past that.

Also, for the accuracy on the test, typically it is measured by whether the word was correctly typed after the spacebar is pressed to move to the next word, not a character by character as it's typed (as in, type tou<delete><delete>hough<space> and it's correct). I've always thought it odd that it was done this way, but I naturally correct while I'm typing and that had negative consequences on your test.

That said, you've got something really, really solid here, worth way more than $9.95. It almost lends itself to a subscription model.

[kaffeinecoma]

The WPM gauge is set to go from 0..120. I didn't expect to have many customers that can type > 100WPM, but perhaps I was mistaken. If anything, I thought that I might end up dialing that down to 0..75, so that most users at the lower-end of the range feel like they're getting a positive result. Maybe I could make that a configurable option in a future release.

As for measuring typing accuracy, that's surprisingly complicated. I wrote up a blog posting on this:

http://quickbrownfrog.wordpress.com/2010/11/22/measuring-typ...

Thanks for the feedback!

[europa]

I second the subscription idea. Probably you could charge like $5(or more) per month with a minimum commitment of 6 months.