[tor-dev] txtorcon 0.2

[Damian Johnson]

> It seems to me that within the Tor community, for everything that's not
> the "tor core c code", there's a lot of duplication, duplicated code
> around or abbandoned projects.

We do have quite a few abandoned projects, though the core codebase
certainly isn't the only one that's actively maintained...
https://www.torproject.org/getinvolved/volunteer.html.en#Projects

Stem aims to be a replacement for TorCtl which has been collecting
dust for years due to maintainability problems (long story short, its
codebase lacks tests and its maintainers are fearful of any
substantial changes to it).

> Additionally you might consider rewriting, based on STEM, the ooonioo
> software that is now written in heavy Java.
>
> Based on your parsers it should take probably few hours to provide the
> Atlas's REST interface.

That's actually already our plan. Stem will eventually replace
metrics-lib (a java library for descriptor parsing), and be used for a
python Oninooo. This has been discussed on occasion in the 'Python
metrics-lib' thread...
https://lists.torproject.org/pipermail/tor-dev/2012-May/003504.html

Cheers! -Damian