Today was slightly slow. I re-remembered why it’s a good idea to leave things incomplete at the end of the day or week: it gives you somewhere to start. I consider lsaddr to be more or less feature complete at this point, at least for a ‘1.0’. That was the state on Friday, and is still the state today. If something was left, I would have had an immediate place to start this morning.

I wasn’t feeling like going back to IPC benchmarking immediately, so I ended up deciding to run a version of my Let’s Build a Shell! workshop. I’ve been wanting to do this for a while, since I think it’s a pretty good way to get a deeper understanding of how our programs are run. I need to file some issues to improve the documentation a little, but I remain pretty pleased with how it’s set up. Good job past me!

After the workshop, I spent a bit of time deciding what to do next. I think I’ve settled on adding a memory mapped message builder to Cap’n Proto’s C++ library. This was requested in a recent mailing list posting. It’ll get me poking at the innards of Cap’n Proto a little, which would be good for one of my potential bigger project ideas for my time at RC: a shared memory transport for Cap’n Proto RPC.