... There are Chocolate-Covered Bacon-Wrapped Smoked TwinkiesA few weeks ago I got a Vertical smoker for Father's day. Over (smoked) dinner later that week we somehow got on the topic of disgusting fair food, things like chocolate-covered Corn Dogs and fried Twinkies. And then the topic of Smoked Twinkies came up. Of course everything is better wrapped in Bacon, so it became Bacon-wrapped smoked Twinkies. And then chocolate-coated, bacon-wrapped, smoked Twinkies.
As my daughter will tell you, perhaps the scariest thing about these were that they actually didn't taste that bad. And from this experience (never to be repeated), we also learned that smoke-flavor imparted into cake can be good (Vanilla ice cream on smoked pound cake turned out pretty well as a follow-up).
As an experiment, it was completely ridiculous, very easy to perform, and resulted in some interesting results that had been completely unexpected. The cost was virtually nil, and I learned a great deal about my smoker just doing it. This is much better than spoiling a $60 piece of brisket on a first attempt (that didn't happen either fortunately, it was wonderful).
The point of this post is that sometimes you have to be willing to try an experiment on something complete stupid. There's value in that because there is no danger that anyone will ever tell you to ship it, and you can still learn a lot just by doing. And you can throw a bunch of things at it just to see how they would work together. And if it fails, you never expected it to work anyway, so there's no real disappointment.
[Aside: How many developers have had the scary experience of showing a weekend prototype to a product manager only to be asked how soon you could ship it?].
The trick is to do the best job you can anyway, just to see whether there is something in this completely idiotic idea. When I smoked the Twinkies, I left a couple of them bacon-unwrapped, just so I could see the difference the bacon made. When I chocolate covered them, I also covered only the bottom half, again to compare with Bacon alone vs. Bacon and Chocolate.
When I design a form (e.g., for an EHR), I often throw in some features to see how they work together with the form. Chocolate coating as it were. But I also leave myself an easy way to see how the form works without them. Profiles and standards often have some optional bits to, because we aren't sure they'll be needed all the time. When I implement a standard, implementation guide, or IHE profile, I try to include the optional bits if I can. It's only when it becomes challenging (doesn't taste good) that I ignore the option. As it turns out, all variations of the Smoked Twinkie were edible, and some (like me) even liked them, although others (like my daughter) would never admit that.
And if, as in my Twinkie experiment, you actually learn something other than how to work with the tools, it was worth more than you expected. If it doesn't, it was at least something you could chalk up to experience and pass the story along to your colleagues amid gales of laughter.
P.S. How does FHIR fit into this? Attending my first FHIR Connectathon was just one of those throw-away experiments. Now I'm a believer. You should try it at least once.