could the wars be in our heads
it's been two weeks on the job and I'm already feeling great about it. my research is, as my current favourite description puts it, "slightly more intelligent than pure randomness" race condition detection on kernels, being implemented as a simics module, hopefully to be used in 410. the infrastructure paralysis that i was most worried about has hardly been a problem at all - like all large codebases, simics's api is tragic in some ways, but I've just been blazing through it somehow, and I'm coming up on the point where I get to make interesting design decisions.
today I accidentally made my research meet its ultimate goal (that being to expose concurrency bugs) - or, at least, in one particular case. after my code grew the ability to generate keyboard events, I tried telling my student kernel to run "mandelbrot\n" ("very quickly" and several times at once), and this happened:
compare that to what the reference kernel does:
anyway, I decided to have a rule that I only do work on campus on the weekdays, so I stay motivated while at work and stay in high spirits during the off time. this plan is working great, and also helped by my other policy to run three weekdays at 0830, which itself is great not only not only because it avoids the oppressive heat but also avoids me having to motivate myself to go later in the day but also because I feel fantastic when I sit down afterwards to start a fresh day of work.
mullownium is interested in training with me (to a very impressive degree), and showed up for my speedwork on wednesday, which was tons of fun. I only hope I can have company more consistently as the summer goes on.
garth offered to arrange for me to have a key to an upstairs lab in which to do research, but I don't know if I'd ever want to use it - working in the cluster is actually already nice enough. people flow through it often, so I have enough low-effort socialisation to keep from getting lonely while at work, and also I have company with whom to grab lunch (especially several people I don't get enough chance to interact with during the year). and when it gets too busy I can always throw on my headphones and do the "passive people-watching" "i'm in a big crowd but it's okay" thing that I always seem to enjoy.
I've had strong pangs of missing for the bay area - the artificial "all of us interns are here so we should hang out all the time" atmosphere, the absence of unbearably hot weather, all the little ways I'd get taken care of (google paying for my food, for example). some of these pangs have been for strange things such as restaurants that I only enjoyed for the company or conversations I used to have there rather than any inherent enjoyment of the place, which seems suggestive of deeper unresolved feelings about where I ultimately want to end up. for now, though, I'm trying to make the most of what's at hand, which can be more than I give it credit for (farmers' markets: yessss).
I've vague notions of going to california for a week or more sometime later in the summer - not sure for how long, or when, or where I'd stay, though I'd like it to coincide with capture the flag in the park, and obviously with seeing people who want to see me.
today I accidentally made my research meet its ultimate goal (that being to expose concurrency bugs) - or, at least, in one particular case. after my code grew the ability to generate keyboard events, I tried telling my student kernel to run "mandelbrot\n" ("very quickly" and several times at once), and this happened:
compare that to what the reference kernel does:
anyway, I decided to have a rule that I only do work on campus on the weekdays, so I stay motivated while at work and stay in high spirits during the off time. this plan is working great, and also helped by my other policy to run three weekdays at 0830, which itself is great not only not only because it avoids the oppressive heat but also avoids me having to motivate myself to go later in the day but also because I feel fantastic when I sit down afterwards to start a fresh day of work.
mullownium is interested in training with me (to a very impressive degree), and showed up for my speedwork on wednesday, which was tons of fun. I only hope I can have company more consistently as the summer goes on.
garth offered to arrange for me to have a key to an upstairs lab in which to do research, but I don't know if I'd ever want to use it - working in the cluster is actually already nice enough. people flow through it often, so I have enough low-effort socialisation to keep from getting lonely while at work, and also I have company with whom to grab lunch (especially several people I don't get enough chance to interact with during the year). and when it gets too busy I can always throw on my headphones and do the "passive people-watching" "i'm in a big crowd but it's okay" thing that I always seem to enjoy.
I've had strong pangs of missing for the bay area - the artificial "all of us interns are here so we should hang out all the time" atmosphere, the absence of unbearably hot weather, all the little ways I'd get taken care of (google paying for my food, for example). some of these pangs have been for strange things such as restaurants that I only enjoyed for the company or conversations I used to have there rather than any inherent enjoyment of the place, which seems suggestive of deeper unresolved feelings about where I ultimately want to end up. for now, though, I'm trying to make the most of what's at hand, which can be more than I give it credit for (farmers' markets: yessss).
I've vague notions of going to california for a week or more sometime later in the summer - not sure for how long, or when, or where I'd stay, though I'd like it to coincide with capture the flag in the park, and obviously with seeing people who want to see me.