That was a waste.
Oct 25th 2006Wallabynerd alert
After I defended my dissertation in May, there was a gap of a couple months before starting my postdoc down the hall. This was a bit of a surprise when “I’ve already got funding” turned — of course — into “I just need to apply for one more grant…” And my landlady, nice as she is, was not willing to accept a freshly bound copy of my dissertation on why bones are strong in lieu of rent.
Thankfully, my old advisor was generous (or enlightenedly self-interested) to keep paying me on a temp basis, to tidy up some loose ends with the software in the lab. Well, really to rewrite as much of the software as I could, since some of it was 4-5 years old and nobody really knew how it was supposed to work anymore. And then to write a new piece of software to control the laser and the camera for the Raman scanning system. That was a fun little adventure, but I got it working reasonably well, wrote up a couple pages of instructions, made sure I’d commented everything for the next sucker, and said “tada!”. And my advisor said, “it’s not half bad!”
Today I got an email. “The scanner program isn’t working. Could you come take a look if you have a chance?” This is the disadvantage of having an office just down the hall from my old lab, instead of across the country. Must remember this for the next job: move far far away.
So I stroll on down, expecting it to be a loose connection or something equally simple. Chatted with my advisor for a minute; he mentioned that someone had found and fixed a bug in the program we’d been using to subtract baselines from spectra. Oh, it was the same bug I’d found and fixed in my update of the software back in June. Which apparently nobody had taken any notice of.
Loose connection? Nope. Instead, the postdoc who’d sent the email turned out to be using the wrong version of the program; the one I’d written for the earlier camera. (The one with the correct number of control lines, instead of the “improved” version with virtually no capability to interface with other hardware.) So I went looking for what I thought was the correct version — not “scanner2″, you need “scanner2_newton”. Well, that didn’t work either. “Wait a minute… I wrote a whole separate thing entirely in LabVIEW to avoid just this kind of problem. Where’d that go?” Search search search. I turned up two versions with good names, both of them about three weeks behind the finished version.
It seems that someone managed to delete both copies of the program I’d spent the whole month of July writing and polishing. From two separate computers.
And so, instead of putting in a good two months’ work to earn my keep this summer, the lasting effect was… well…