tim: Tim with short hair, smiling, wearing a black jacket over a white T-shirt (work)
[personal profile] tim

I can barely bring myself to write anything today, both because I'm annoyed about some work-bureaucracy stuff (but isn't that always the source of annoyance?) and because I spent today on work that, while highly enjoyable (to me), is especially interesting (to anyone else), but writing every day is a habit I want to have, so I'll try.

Today was bug triage day! So I triaged ALL the bugs -- no, not really, I triaged bugs 2396-2526. Not a particularly notable number of duplicates or obsolete bugs today, but a fair number that were in need of labels and milestones. I stopped assigning bugs to people as much as I was before, so that there are some unassigned bugs for interns and new contributors to work on.

Then I moved on to the ongoing task of opening issue tracker issues for each FIXME comment in the code. It's hard for me to explain exactly why I enjoy doing this so much, but I suppose it's the joy of taking something chaotic and making it slightly more orderly. It's totally tempting to strew random FIXMEs all over the codebase (and I've done it plenty), but that means there's no easy way to ask "what are the open issues right now?" When each FIXME corresponds to an issue in the issue tracker (not necessarily a one-to-one correspondence), it's a lot easier to do queries on the open issues.

Moreover, it's especially fun when I can see a FIXME that I know how to fix, and I fix it -- it's like the triumph of human communication, in that I can go ahead and understand someone's cryptic comment and modify code that I don't know at all, and have the test cases still work so I know it's okay to check in. An example today was a comment on a foreign function saying "change the argument type to u64 when the FFI supports that"; (unsigned 64-bit ints) -- it was 32-bit ints. It turns out the FFI does support that now, so I fixed it, and everything worked.

Plus I got to do some trivial refactoring on the first significant Rust code I wrote last year (the typestate pass) that had been sitting around waiting to be done for a long time. Of course, that might be pointless because Rust might be getting rid of typestate... but I really hope it doesn't, because no one ever really tried to develop it or improve it after the initial prototype written by me (and I in particular didn't try because I didn't want to continue the apprehension that I was the only one who cared about it), and it seems sad to abandon what I think is one of the cooler features of Rust's type system just because no one's tried to make it more practical.


tim: Tim with short hair, smiling, wearing a black jacket over a white T-shirt (Default)
Tim Chevalier

October 2017

8910 11121314
15 161718192021

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags