wcampbell

> Thanks for the update. Probably this needs some refactoring because the compression happens in `data_writer.add_bytes(...)` and `files` is borrowed from self. Feel free to submit a MR, I'm quite...

> What version of the toolchain is being used, compiler outputs being used for tests without pinning a version is what's causing some of these CI failures. I had been...

Quickly wrote up a possible solution here: https://github.com/sharksforarms/deku/pull/441 EDIT: Eh, I need to fix the test failures.

> I also wonder whether it is possible to read an exact number of bytes instead of the loop for Vec. But I haven't figured out how to specialize the...

For reference, not as *perf* as it was ;) ``` + critcmp stable nightly stable-perf nightly-perf group nightly nightly-perf stable stable-perf ----- ------- ------------ ------ ----------- Deserialize/binrw 1.38 2.8±0.06ms ?...

Hmmm, my inline(never) hurt the performance of the count vs read_all. I might have to balance this.. ``` read_all_bytes time: [10.605 µs 10.623 µs 10.649 µs] read_all time: [10.715 µs...

> Hmmm, my inline(never) hurt the performance of the count vs read_all. I might have to balance this.. > > ``` > read_all_bytes time: [10.605 µs 10.623 µs 10.649 µs]...

I'm not interesting in continuing this MR, I don't even think it's really a good idea lol

This is because of `ctx`s such as `Endian` are passed into sub-types. The following works: ```rs use deku::prelude::*; #[derive(DekuRead, DekuWrite)] #[deku(id_type = "u8", endian = "endian", ctx = "endian: deku::ctx::Endian")]...