[tor-dev] Has Core Tor Development Slowed? Or Are We Moving To Rust/arti?

[Alexander Færøy]

Hello Neel,

On 2021/01/08 16:34, Neel Chauhan wrote:
> I hope you all had a great holiday season.

Likewise!

> Sorry if I have been less active in the Tor community as opposed to the
> past. I've been more focused on FreeBSD Ports as of now, but I still want to
> post at least the occasional Tor patch here and there.

Absolutely no need to apologize, Neel, you're one of our more persistent
contributors over the years :-) Hope you enjoy hacking on the FreeBSD
ports tree as well.

> Core Tor/Tor hasn't seen an update since December 21, even when we are
> already a week into 2021.

TPI ("The Company behind Tor") gave everybody some additional amount of
vacation days for this holiday because last year was quite intense for
us all. Some people returned on the 7th and the rest will return on
Monday the 11th. That might explain why nothing have happened yet here
in 2021 :-) Expect operations to normalize next week, hopefully with
recharged batteries ready for 2021 :-)

> Is this related to the layoffs which happened last year? Or is Tor moving to
> Rust (via arti) and all the development is happening there? I'm guessing
> both.

The amount of work done is definitely related to the layoffs that
happened in April, but also related to our deliverables for 2020 for the
teams and thus the people left after the layoffs.

When the layoffs happened, most of April and quite a bit of May was used
for re-structuring the organisation and figuring out what to do next. We
decided to move some people around to do tasks for different teams than
their usual ones, and we created some new tasks we wanted to resolve
here and now that didn't have to do with any specific codebases or
teams, such as looking at our infrastructure for doing software
development.

If we look back at the second half of 2020, for Tor's Network Team, then
a lot of the tasks we have had has been longer term deliverables. We
started up the new project on network performance work where a lot of
cross-team interaction have happened together with the network team, the
metrics team, and our new network health team where the teams works
together on analysis/experiments or in codebases that are not
necessarily tor.git. Efforts such as looking into the next generation
bandwidth scanners (both sbws and flashflow) have taken some time as
well.

When Nick had shown the prototype of Arti at some point in 2020, we
tried to figure out a way where Nick could focus a bit more intensely on
this project, because it was looking *very* promising. This of course
have an impact too, but we think it's worth it because Arti will be
solving some of the architectural issues that will take a very long time
to resolve in the current codebase.

Additionally, some time have been spend helping our friends from the
guardian project on some profiling and Tor integration work for their
mobile applications.

It should be said though, that we don't hope that you have been impacted
by this as a volunteer over the year. We have tried to ensure that all
the normal tasks we have over the weeks regarding Tor maintenance have
remained the same (tasks such as doing code reviews, rolling releases,
backporting, etc.)

> If it's the latter, I guess I have to start learning Rust sooner rather than
> later. Well, who am I kidding, I had to learn C# and PowerShell for my job
> (Disclaimer: I work at Microsoft, not on security or Windows/Azure however).

I think exploring Rust now might be a good idea. The Arti project that
Nick started does seem to solve some very specific issues that we have
been trying to deal with for years, and everybody in the Network Team in
Tor have been very impressed with how quickly things are able to move
forward in the Arti project.

If you take a look at the TODO file[1] in the root of the Arti
repository, and take a look at its history, you can get a feeling on how
quickly Nick have been able to move this project forward. It's very
exciting, and people are starting to submit patches to it as well :-)

A lot of us inside of Tor's network team still need to become better
Rust programmers, but hopefully we can speed that up by working
together.

All the best,
Alex.

[1]: https://gitlab.torproject.org/tpo/core/arti/-/blob/main/TODO

-- 
Alexander Færøy