Corin Anderson (magellanic) wrote,
Corin Anderson

Notes from work

My tech lead and his co-conspiritor gave a talk today about how the system they built worked. After about an hour, we members of the audience decided what they built was just like X, only with minor changes delta. My friend K said what we all were thinking: you guys should have just said "we did X + delta" and we'd have understood in an instant instead of an hour (they had presented the work in some very obtuse terms). I had felt this way about the work for the last 6 weeks (having worked with them for that long, I got this explanation that long ago), but just thought it was my own slowness in understanding, not strictly their obtuseness. K's comment made me feel much better.

I found several small bugs in my worker script, but squashed them. Slowly, my script is becoming just like existing script Z at work. Too bad I didn't know about Z before I wrote my own. Oh, well. My script does things Z doesn't. Interesting that I chose some of the same options given the same design decisions to make.

I can stare at this work for only so long. Despite distributing the work over many machines, some tasks take a long time (30 minutes), and I can surf the web only so much per day. And, there's the frustration of herding many machines all at once. Invariably, one command on one machine will wedge, and, if it's a critical command, it may be necessary to restart the whole process. With luck, though, I should be able to avoid most of those cycles for the next several days (finally got a reasonable set of data to play with).

As part therapy, part actual work, I slurped down a buch of code from my grad work to look at and work on. Today, I began writing some code for an upcoming paper deadline based on that grad work. It felt sort of like I wasn't doing work, but not too much unlike work. I mean, I have an official okay to spend work time on the paper, which is great. And, it's fun! It's fun because, for this bit of code, I get to design and implement my own algorithm, which is a lot more fun than just adding minor features X and Y to an existing system. Or writing a hacky implementation to solve some problem. And it's fun because I'm building what I think is a neat system. Granted, my lion's share project at work will be much cooler (if we can pull it off) than my paper. But, I'm hoping my paper will fill a niche. And be a further reason to cement why I'm attending the conference.

So, today wasn't too bad of a work day. I wrote some code I'm sort of proud of. I sat through a 3 hour meeting and decided my opinions aren't too far off base. I got to write some slick python code. And I left work with no open bugs with my name on them. If I can close off the week with another day like today, I think I may feel pretty good this weekend. 8)
  • Post a new comment


    default userpic
    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.