I’ve been hanging around developers *entirely* too much lately.
These days I find myself telling myself the story that unless I get back into coding, I’m not going to be relevant in the tech industry any longer.
Hanging out (aka volunteering) at developer-focused conferences will do that to you:
Volunteering on open source projects will do that to you (jQuery Foundation‘s infrastructure team).
Interviewing for engineering-focused Product Owner and Technical Product Manager roles will do that to you. (Note: when did “technical” become equivalent to “I actively code in my day job/spare time”?)
One of the hang-ups I have that keeps me from investing the immense amount of grinding time it takes to make working code is that I haven’t found an itch to scratch that bugs me enough that I’m willing to commit myself to the effort. Plenty of ideas float their way past my brain, but very few (like CacheMyWork) get me emotionally engaged enough to surmount the activation energy necessary to fight alone past all the barriers: lonely nights, painful problem articulation, lack of buddy to work on it, and general frustration that I don’t know all the tricks and vocabulary that most good coders do.
Well, it finally happened. I found something that should keep me engaged: creating a stripped-down search interface into the Marvel comics catalogue. Marvel.com provides a search on their site but I:
- keep forgetting where they buried it,
- find it cumbersome and slow to use, and
- never know if the missing references (e.g. appearances of Captain Marvel as a guest in others’ comics that aren’t returned in the search results) are because the search doesn’t work, or because the data is actually missing
Marvel launched an API a couple of years ago – I heard about it at the time and felt excited that my favourite comics publisher had embraced the Age of APIs. But didn’t feel like doing anything with it.
Fast forward two years: I’m a diehard user of Marvel Unlimited, my comics reading is about half-Marvel these days, and I’m spending a lot of time trying to weave together a picture of how the characters relate, when they’ve bumped into each other, what issue certain happenings occurred in, etc
Possible questions I could answer if I write some code:
- How socially-connected is Spidey compared with Wolverine?
- When is the first appearance of any character?
- What’s the chronological publication order of every comic crossover in any comics Event?
Possible language to use:
- C# (know it)
- F# (big hawtness at the .NET Fringe conf)
- Python (feel like I should learn it)
- ScriptCS (a scriptable C#)
More important than choice of language though is availability of wrappers for the API – while I’m sure it would be very instructive to immediately start climbing the cliff of building “zero tech” code, I learn far faster when I have visible results, than when I’m still fiddling with getting the right types for my variables or trying to remember where and when to set the right kind of closing braces.
So for sake of argument I’m going to try out the second package I found – Robert Kuykendall’s “marvelous” python wrapper: https://github.com/rkuykendall/marvelous
See you when I’ve got something to report.