I had a busy week with my parents visiting. I was leaving RC a bit earlier than I would have otherwise, and also getting to bed later—they were staying in midtown, and I’m in Flatbush. As a result, I let my posting slide a bit, and this is me catching up. I’ll also plan to write lab notes such as these each normal RC day, ie Monday through Thursday.
Ostensibly, the main project I’m working on right now is investigating Linux IPC. The eventual plan is to try and implement a share memory transport for Cap’n Proto.
As a start, I implemented a couple of small programs to measure latency for pipes and eventfd. It looks like the round trip time on my machine is on the order of a couple of microseconds. It varies a little depending on if I pin the communicating processes to the same core or not. This was nice to see, since it would require going over the L3 cache. However, the difference was on the order of a couple of hundred nanoseconds, which is more than I expected. I’ll aim to get some data and draw some histograms early this coming week.
Eventually (or maybe soon?) I want to start learning JS. Or ECMAScript 2015 to
be more specific. Or maybe even some ES7 features, like
was planning to learn React for some nifty front-end isomorphic buzzword
compliance, but I’m not sure now. Maybe Polymer? Maybe some other virtual DOM
I did a refactor of my IPC programs to allow them to share a
handled command line arguments and other stuff. This required me to rejigger
Makefile to have both programs depend on the new source file, and to link
it in. I asked Twitter the wrong question:
is actually learning make a worthwhile pursuit?— Kamal Marhubi (@kamalmarhubi) June 6, 2015
The responses lead me to expend small but non-zero time on this. I learned that
$< are meaningful variable names. Also that it’s
possible to build a static site generator driven by
@kamalmarhubi I love make! It saved my life so many times. I'm writing a static site generator in a Makefile.— linse (@linse) June 6, 2015