rust retrospective

critical things to do before deciding about rust

rust builds on windows

platform support

  • takeaways: their support goals is quite like our support goals
  • need to make sure that rust will work well in mobile!!!
  • get in touch with tor mobile people so that they can try rust in mobile!!!
  • (maybe not all parts of tor need to run on mobile?)

need tor team members experienced with rust

need non-trivial rust code in tor

need reproducible builds / reproducible rust?

  • ximin working on this for debian
  • tor browser team will have to figure this out anyway, since ffox 59 ESR (late june 18) will include rust
  • maybe we should try helping tor browser team?
  • next dev meeting is march 18 so before ffox 59. so that's good.
  • who else from debian can help with this?

status update

chelsea's protover code reviewed by rust people. a few review pooints need to be addressed

rust people found the ffi code pretty good


optional off-by-default rust in 0.3.3

also, get a few more modules going in 0.3.3

  • chelsea's protover as a first candidate
  • dirauths in rust?
  • use isis ed25519-dalek code in tor (needs audit by people who know how to edit that sort of code!!!)

also, get more feedback from people to try rust in various platforms (variety of OSes, mobile, compilers, etc.)

also, speak to our linux packagers and inform them that we are doing rust and let us know if they envision any problems with it

also, write a blog post about rust support and our platform support policy + our timeline

need some sort of verification mechanism to ensure that C files and rust files are in sync

pair up with different peopple everytime when working on rust stuff to propagate knowledge

how to fit the rust stuff easier into our sponsor roadmap?? ==

  • the modularization proposal?
  • the dos resistance grant? (no crashes!)
Last modified 2 years ago Last modified on Oct 12, 2017, 4:05:48 AM