The first thing of the day was fixing a super edge case in a JTWC tropical weather outlook for Cyclocane.
Bizarrely, their outlook included a sentence that almost exactly modeled the sentence structure they use to describe the potential for some tropical disturbance area to turn into a tropical cyclone. More bizarrely, the sentence included the word ALLOW, which was then getting picked up because it matched my regex for
/LOW/. The fix was to add something similar to this, to the regex:
/( |^)LOW/. So now, it looks for a space character or the start of the line before matching LOW.
The big work of the day though, was the awesome BMI calculator.
I finally launched the calculator publicly.
Originally, I was going to try to figure out how to build a site using the
Opal::Server method that I had been using for testing. But I just ran into too many snags where I would've had to add manual workarounds to make it all play well together. So instead, I decided to just give up and make it a Middleman project.
Actually, it was likely to end up as a Middleman project eventually anyway; I just wanted to see if I could make it work the other way first.
Originally, I put the calculator on a subdomain of this blog's domain name, but later I decided to actually buy a domain name for it.
To be honest, I feel kind of silly spending money on a domain name for this project. It's tiny and given my previous experience with this sort of thing, the project is most likely going to fade into obscurity. Or more so, it will remain obscure.
See, the BMI calculator space is full of plenty of competition. And competition from big sites who have built-in audiences to drive traffic to their calculator. And of course, a calculator (with the way most of them have been implemented anyhow) is pretty straightforward to implement. There's a low barrier to entry, in other words.
I bought a domain name anyhow. For a few reasons.
For one, the calculator has a better chance of actually making a name for itself if it has a brandable name rather than a name like "bmi.unrelateddomain.com".
And there's the fact that, not only have I spent so much time on it already, but I have plans for future work as well. To me, giving it its own name "respects" the work that I have/will put into it.
But really, it's like playing the lottery. There's that hope. The hope that it could become something big. "Big" could mean traffic; it could mean money. Preferably both, of course.
Of course, given my track record, I'll be happy to just get the domain to pay for itself.
glaring omissions from the BMI calculator
The biggest thing that's missing right now is that the calculator just gives you your number. It doesn't tell you anything about what category that puts you in, nor does it tell you how much you'd need to gain/lose to get to other categories.
The experience on mobile could also be greatly improved.
The upside to how I wrote things is that it can respond immediately to what you're typing in, rather than having to wait for you to fill everything and hit submit.
The downside is that this "breaks" things like mobile devices that are expecting the form to work in a certain way. Specifically, on my ios devices, a "Go" button is available, but it does nothing. If you're using an ipad, it's not too bad, because you are hopefully noticing that your BMI number is changing when you input your numbers. On an iphone, it's much worse. Though there's a somewhat subtle transition between red/green for the input boxes indicating when your input data is invalid, the BMI number is "off screen" and you have no idea that you should hide the keyboard and zoom out.
At this point, I'm not sure how I'll go about making things obvious, but I know that I definitely should.
Otherwise, it's not a very awesome BMI calculator.