Bruce and velope requested for it to be raised. I'm tempted to say 'heavy'
based on the number of commits, but prior to March there was no activity at all
for two years so I'm gonna hold off on doing that until we see if this is
maintained or a flash in the pan.
Someone showed up on irc confused, thinking that we were primarily mentoring
through OPW. The two programs have different requirements and GSoC is a far,
far larger program for us.
Andrew reported during the dev meeting that we still have a slot so adding the
OPW section back in, updated for the program running in parallel with GSoC.
This section was removed in r25974.
Spoke with Nick and got both confirmation that they're still relevant and
prereqs for applying. Hopefully this'll ensure a decent quality for
applications and warn away students for whom it's a bad idea.
Tweaking the 'Stem Usability and Porting' idea to include tackling the
remaining issues that prevent us from deploying TorBEL. Thanks to Sebastian for
the list of tickets!
Cross metrics/stem project from Karsten for rewriting Doctor in python, using
stem for the descriptor support and expanding it to be able to remotely fetch
descriptors.
Why are we still bundling dozens of commented out tasks with the page? If we
want to add these back in later then we should... well, add them back in later.
That's what SCMs are for (even if svn sucks at it). :P
This is slowing down page loads, and most ideas are getting to be pretty old
anyway.
When I've spoken with Sukhbir he mentioned that he doesn't think there's enough
torbirdy tasks remaining to fill a summer. We can add this back in if he
changes his mind.
Noting that Google hasn't yet selected orgs (hell, the org application phase
hasn't even *started* yet). I'm not interested in repeatedly saying this to
prospective students so hopefully these notices will be sufficient.
This is a project that I have been batting around with Nick for a bit. It
doesn't include the 'grand vision' he wants for covering inter-network
components, but it would provide dedicated tests for a very substantial portion
of Tor's codebase.
While I think a more complicated network simulator and testing suite would
yield greater benefits, I also think it would be *vastly* harder to make and
hence quite likely to never happen. Presently our automated testing of Tor is
pretty poor and there's no need to let the perfect be the enemy of the good on
this.
Last year I proposed a stem counterpart for PathSupport as a project for GSoC.
This did not go very well since we really haven't a clue *what* we want from
such a project. Since then this has not changed and there's little reason to
leave it on the page (even commented out).
Many of the subtasks for stem's usability project have been done. Replacing
those with other tasks I've come up since then. I'm also requiring a code
sample from applicants to better get them engaged during the application
process.
Ok, I'm adopting the following policy:
"If a project never made it out of alpha *and* hasn't been touched in over two
years then we should stop listing it."
Thandy and TorBEL are getting close to this threshold, but JTor has been
inactive for almost three years. It's confusing for us to keep showing
prospective volunteers such a long dead project.
Removing the remaining stubs as previously discussed (Tor cloud, ExperimenTor,
Chutney, and Torperf). I'm also dropping bridge-guard, see my unanswered
tor-assistants@ email on 12/28 for why. :)
Peter's r25976 reversion was done in the name of expedience rather than making
the page better. Restoring what I had minus the removal of the stubs (which I
gather was the only objectionable bit). As mentioned on tor-assistants@ I plan
to remove those stubs later next week after the CCC crowd has passed.
Besides this made a few small fixes...
* You were labeling bridge-guard as being "Flash Proxy", likely a paste error.
* I've removed the empty links from the stubs. No links are better than broken
links.
* Integrated the new Thandy links with the summary and switched it to use past
tense. The project has been inactive for a long while.
This reverts commit 681ab33b165e9c427401f19f3453fb6ba002eb7f.
Please leave them as Roger made them at least for now. Future updates
deleting lots of content should probably be discussed in advance.
General update for the volunteer page's projects table. This reverts quite a
bit of Roger's additions in r25973 which added a bunch of projects without
summaries or links (please don't do that!). Kept the good bits, though. :P
* Merged Obfsproxy and PyObfsproxy. They both have the same project page, same
bug tracker, and same description. Doing a joint listing is what we did for
Onionoo too which has both a java and python implementation.
* Dropping Anonbib. I asked arma about it on irc but he didn't reply. It's not
a substantial coding project and the summary he wrote is essentially nothing.
* Dropping "tor-ramdisk", "Tor cloud", "bridge-guard", "ExperimenTor",
"Chutney", and "Torperf". They all had no summary and code/bug tracker links.
If the author doesn't care enough about having their project listed to send
those then we shouldn't list it. :)
* Added missing link for torouter's code and Onionoo's bug tracker.
* Removed the flag that Ooni Probe was in alpha. I'm not positive that it's in
production, but it's been under highly active development long enough that I
suspect it's ready.
* Changed arm's description. It's definitely not still under active
development, and the summary didn't really say what it *did*.
* Adjust activity...
* Flash Proxy: Moderate => Heavy
ye god he's doing a ton of work!
* Shadow: Moderate => Heavy
again, almost daily improvements - I'm jealous
* Txtorcon: Moderate => Heavy
not as much as Flash Proxy or Shadow, but still unusually active
* Ooni Probe: Moderate => Heavy
not surprising, with multiple people dedicated to it
* Tor2web: Light => Heavy
lots 'o commits
* Vidalia: Light => None
chiiph has been gone for four months
* Orbot: Moderate => Light
no commits since October and I didn't find much activity on their github
page... though things are so spread out it's hard to tell
* Torouter: Light => None
basically no activity since August
* Torsocks: Heavy => Moderate
no activity since October
* Thandy: Light => None
no commits since 2011
* Atlas: Moderate => Light
essentially nothing since September
* Compass: Heavy => Light
nothing since September
It has always irked me that we had a task 'priority'. Who decides these
priorities? Against what? Sure I think that stem projects are the most
important thing since the discovery of pepperjack cheese but does that really
mean I should label my projects as 'ultra super omega-purple level important'?
Dropping the priority from task ideas and sorting everything alphabetically.
Couple of revsions suggested by Marina including...
* Noting the application deadline date.
* Saying that the ideas listing is for OPW in addition to GSoC.
As theatened on tor-assistants@ I'm trimming our volunteer page's task listing
to only include things with an engaged mentor.
The listing has grown stale which is confusing to newcomers. Some projects are
done, and others reflect plans that are a year out of date. My devious scheme
was that if I required some effort from mentors to keep task ideas around then
we'd only get buy in for those that were really still relevant. Looks like it
worked - we're down to three.
As Ravi discovered during the last GSoC application phase our PathSupport idea
is too poorly defined to make a good GSoC (or GSoC-like) project. Replacing it
with far better defined tasks for improving stem's usability.
Dropping the 'torbutton for thunderbird' project idea since it is now... well,
implemented. Replacing the entry with a project that Sukhbir would like to
mentor.
Stem is far enough along that it's useful to users, and should have a spot on
the projects page. Also revising the volunteer page links to point to its
readthedocs page, which is presently our homepage.
As I understand it Aaron has adopted the torctl/torflow codebases. They also
haven't been touched in months so noting both.
Ravi's still working on stem and Sathyanarayanan is working on Onionoo's python
counterpart (it should probably be listed as a separate project, but he hasn't
provided an entry for it yet).
Arturo stripped all trailing whitespace from the volunteer page wml while he
added his project. This bugs me and, since I make 95% of the changes to this
file, I'm gonna put it back the way I prefer.
The volunteer page wasn't updated after the last mass trac migration, causing
most of the project page links to give an error. Caught thanks to breton.
Nathan hasn't responded about GSoC mentorship, and his last email indicated
that it was unlikely. Jake gave a blanket 'yea, I can mentor any of those
projects' which is nice, but doens't fill me with much confidence that we have
a primary mentor for anything Orbot related.
I'm on the fence about if we should remove them or flag as simply not being
eligible for GSoC and mostly decided to go this way due to lack of
communication. If Nathan replies later then great.
Over a week back I asked everyone listed as potential mentors if they could do
it again this year and if their projects were still relevant. Most people
replied but a few, understandably were too busy to do so. This included Erinn,
Nick, Nathan, and Robert Hogan.
Removing them as being potential mentors and dropping the projects that they
were the sole mentor for. This included...
* Improve our unit testing process (Nick, Erinn)
* Simulator for slow Internet connections (Nick)
* Make torsocks/dsocks work on OS X (Robert Hogan)
* Orbot integration library and community outreach (Nathan)
We can add them back later on if they change their minds.
Mike is juggling too many things right now to be a primary mentor. At the
moment we're trying to figure out if there's an alternate mentor for the
thunderbird idea - after that I'll remove him from that one too.
I should stop flagging everything that I work on as 'medium'. If some of these
other projects can be flagged as being a 'high' priority then I can flag mine
that way too. It's certianly high priority to me. ;)
Couple projects sent by intrigeri...
* Petname system for Tor hidden services (ague)
* Tails server: Self-hosted services behind Tails-powered Tor hidden services
(intrigeri, anonym)
Roger wants to focus on helping tor people rather than head a GSoC project, so
removing him from the potential mentors of the censorship resistance project.
We previously had a note saying that this was a no-go for GSoC but it had been
commented out. If this was git then I'd use 'git blame' to figure out why, but
not worth the effort to try to get this information from svn.
Suggested revision by Robert Ransom.
Adding a project that I'd really like to mentor this GSoC: a stem port for
TorCtl's PathSupport module. Hopefully this idea will get a few interested
students...
At the moment I'm trying to focus on stem related work, and would rather not
mentor an arm related project. The gui made by krkhan still needs work and the
client mode use cases are up for grabs so I'm leaving these bits intact, just
commented out.
We'd prefer for new volunteers to hack on the new django codebase rather
continue the php counterpart. Changing the entry in on the volunteer page to
indicate that both exist.
Updating project activity as per the last couple months of commits, and adding
Stem to the project table (people gsathya talked with were confused since it
wasn't there).
- More clearly marking TBB, TAILS, and Torouter as being sys admin tasks.
- Changing the activity on TBB (misread the commits - most activity is version bumps rather than development)
- Suggestion by karsten:
- Replacing the 'etc' with Sebastian for Tor
- Listing kaner with GetTor and nickm with BridgeDB
- Revised description of TorRouter's activity
- Removing the 'General' metrics repo
- Dropping the metrics project idea
