Hey everyone. This morning my grandma passed away peacefully in her sleep at 6am. She was ninety years old and ready to go. As the saying goes: “old age ain’t for sissies”.

These last few months supporting my family has been a full time job, but my schedule should now free back up.

sunset


CollecTor Support

The only technical thing to note is CollecTor support, a new module within Stem that makes gathering Tor’s network topology at prior points in time simple. For example…

import datetime
import stem.descriptor.collector

yesterday = datetime.datetime.utcnow() - datetime.timedelta(days = 1)

# provide yesterday's exits

for desc in stem.descriptor.collector.get_server_descriptors(start = yesterday):
  if desc.exit_policy.is_exiting_allowed():
    print('  %s (%s)' % (desc.nickname, desc.fingerprint))

Behold, my glorious domain! Weep at its splendor!!1!

granny's attic

This month returned to the island I love, and having a blast volunteering locally. Both non-profits are amazing.

Granny’s is a second hand store that fundraises for our community. Volunteering there is a fun mixture of serendipity with the preeminent form of recycling. I really love matching excess stuff with homes that can make good use of it.

As for our food bank, it feeds fully a seventh of our island’s population. That’s ~150 people! Twice a week I’m unloading trucks and re-bagging groceries. After a decade at Amazon I gotta admit – there’s something deliciously cathartic in non-technical work.

Tor is taking a back seat right now, but I have a feature branch for CollecTor that should be ready to announce soon.

Aside from that lots of runs, plays, and other fun stuff. This is kicking off to be a very fun summer!

Freedom! Sweet, delicious freedom! I went dark since March for crunch time at Amazon, but then decided one decade there was enough. Amazon was a great first job but time for a break.

Vacation aside, this summer I’m volunteering with our island food bank and Granny’s Attic to get more balance away from the keyboard. That said, knowing myself I’ll still need my open source fix.

Oh! And of course, most important of all: found a new puffy friend!

mini puffer

Damn I’m a sucker for puffers.


Tor/Stem CI

Stem’s tests were designed to verify Tor/Stem’s interoperability, as well as provide a tool to test Tor itself. This month teor integrated Stem into Tor’s Travis CI and we’ve been working out the kinks…


Smaller things these last few months include…

Greetings wonderful tor community! This month I pretended to be a web development again for kicks and giggles.


Tor People Page

Hiro’s putting together a new website for Tor. As the maintainer of our people page I ported it over. Personally I think the new page looks a lot better.

Before

people page before

After

people page after

Hrio declined my patch, and that’s fine. Learning Lektor was fun. The new site will launch with Hiro’s version of the page with her as its maintainer.


Personal Website

Beside the Tor website I also put the finishing touches on my own! For comparison my old site is available here.

personal website

Happy January, onion enthusiasts! Made it through another spin of the globe. January I wrapped up a couple projects that were in flight…


Nyx 2.1 Release

My biggest news for users this month is Nyx 2.1! No features, but this release includes a full year’s worth of fixes that smooths Nyx’s rough edges.

If you run into any issues please let me know!


Bandwidth File Support

Bandwidth files are weighting heuristics used by Tor for relay selection. Stem can now read these documents. For example…

import stem.descriptor
import stem.util.str_tools

# Bandwidth files are not yet available, but soon will be...
#
#   https://trac.torproject.org/projects/tor/ticket/21377

with open('/home/atagar/Desktop/bandwidth_file_v1.2') as bandwidth_file:
  desc = next(stem.descriptor.parse_file(bandwidth_file, 'bandwidth-file 1.2'))

print('Median relay bandwidth measurements are...\n')

for fingerprint, measurement in desc.measurements.items():
  # Not quite sure what the units are. The spec seems to indicate KB/s?

  bandwidth = '%s/s' % stem.util.str_tools.size_label(1024 * int(measurement.get('bw_mean', '0')))
  print('  * %s (%s) averaged %s' % (measurement.get('nick', ''), fingerprint, bandwidth))
% python demo.py
Median relay bandwidth measurements are...

  * PompoRelay (319AC0126125DA6A557A416C1BB643E52E272AD0) averaged 187 MB/s
  * oberon (B4486871EA062BAF2354F070FECC344CF3C20880) averaged 196 MB/s
  * ICSIL1xDEDISx1 (7CBBAB7867F939763D94FC6C2819E129971CB0E4) averaged 54 MB/s
  * pollypurebred (EF37D150F3A851B32D9E843605D0EA8A1B938117) averaged 86 MB/s
  * Dementor (46D79CCB83639718177B4BDB8AED6BE9378B1D0F) averaged 158 MB/s
  * wormhole (8B863F5CE8676EFE5BB070CE3CAF54A76AE74D1A) averaged 90 MB/s

Hope everyone had a fun for the holidays! Like many of us December I glugged eggnog with family so this will be short.

  • Stem 1.7.1 hotfix release. Tor adjusted its bootstrap log format in a way that broke Stem, requiring a hotfix release for continued compatibility.
  • Nyx 2.1 incoming! No sexy features, but many long overdue stability fixes. January will be going toward this.
  • Stem failed to read large payloads from Tor’s ORPorts.
  • Writing a new homepage for myself. My current site dates back to my college days and it looks the part. Still in progress but this month I finished drafting its content, passed W3C validation, and made its github repo.
Old Homepage
old homepage
New Homepage
new homepage

Happy holidays everyone! This November Iain and I worked together on Stem’s descriptor capabilities. Though this won’t be of interest to a general audience, for metrics folks like us these are pretty neat.


Descriptor Digesting

Tor relay information is provided by multiple documents. Signed descriptors transitively validate others by inclusion of their digest. For example, our consensus references server descriptor digest, and server descriptors in turn cite extrainfo digests.

To illustrate, here’s a diagram from Iain…

descriptor digests

Stem can now calculate digests from server, extrainfo, microdescriptor, and consensus documents. For instance, to validate an extrainfo descriptor…

import sys

import stem.descriptor.remote
import stem.util.tor_tools


def download_descriptors(fingerprint):
  """
  Downloads the descriptors we need to validate this relay. Downloads are
  parallelized, providing the caller with a tuple of the form...

    (router_status_entry, server_descriptor, extrainfo_descriptor)
  """

  conensus_query = stem.descriptor.remote.get_consensus()
  server_desc_query = stem.descriptor.remote.get_server_descriptors(fingerprint)
  extrainfo_query = stem.descriptor.remote.get_extrainfo_descriptors(fingerprint)

  router_status_entries = filter(lambda desc: desc.fingerprint == fingerprint, conensus_query.run())

  if len(router_status_entries) != 1:
    raise IOError("Unable to find relay '%s' in the consensus" % fingerprint)

  return (
    router_status_entries[0],
    server_desc_query.run()[0],
    extrainfo_query.run()[0],
  )

if __name__ == '__main__':
  fingerprint = raw_input("What relay fingerprint would you like to validate?\n")
  print('')  # blank line

  if not stem.util.tor_tools.is_valid_fingerprint(fingerprint):
    print("'%s' is not a valid relay fingerprint" % fingerprint)
    sys.exit(1)

  try:
    router_status_entry, server_desc, extrainfo_desc = download_descriptors(fingerprint)
  except Exception as exc:
    print(exc)
    sys.exit(1)

  if router_status_entry.digest == server_desc.digest():
    print("Server descriptor digest is correct")
  else:
    print("Server descriptor digest invalid, expected %s but is %s" % (router_status_entry.digest, server_desc.digest()))

  if server_desc.extra_info_digest == extrainfo_desc.digest():
    print("Extrainfo descriptor digest is correct")
  else:
    print("Extrainfo descriptor digest invalid, expected %s but is %s" % (server_desc.extra_info_digest, extrainfo_desc.digest()))
python demo.py 
What relay fingerprint would you like to validate?
3BB34C63072D9D10E836EE42968713F7B9325F66

Server descriptor digest is correct
Extrainfo descriptor digest is correct

Detached Signatures

Somewhat obscure, detached signatures are exchanged among directory authorites to establish the next hour’s consensus. They reference the agreed upon digest the consensus should have, and are only available five minutes of every hour (minutes 55-60). Stem can now read these documents

import stem.descriptor.remote
    
detached_sigs = stem.descriptor.remote.get_detached_signatures().run()[0]
    
for i, sig in enumerate(detached_sigs.signatures):
  print('Signature %i is from %s' % (i + 1, sig.identity))

When available (minutes 55-60 of the hour)

% python demo.py
Signature 1 is from 0232AF901C31A04EE9848595AF9BB7620D4C5B2E
Signature 2 is from 14C131DFC5C6F93646BE72FA1401C02A8DF2E8B4
Signature 3 is from 23D15D965BC35114467363C165C4F724B64B4F66
Signature 4 is from 27102BC123E7AF1D4741AE047E160C91ADC76B21
Signature 5 is from 49015F787433103580E3B66A1707A00E60F2D15B
Signature 6 is from D586D18309DED4CD6D57C18FDB97EFA96D330566
Signature 7 is from E8A9C45EDE6D711294FADF8E7951F4DE6CA56B58
Signature 8 is from ED03BB616EB2F60BEC80151114BB25CEF515B226
Signature 9 is from EFCBE720AB3A82B99F9E953CD5BF50F7EEFC7B97

When unavailable (minutes 0-55 of the hour)

% python demo.py
Traceback (most recent call last):
  File "demo.py", line 3, in 
    detached_sigs = stem.descriptor.remote.get_detached_signatures().run()[0]
  File "/home/atagar/Desktop/stem/stem/descriptor/remote.py", line 476, in run
    return list(self._run(suppress))
  File "/home/atagar/Desktop/stem/stem/descriptor/remote.py", line 487, in _run
    raise self.error
urllib2.HTTPError: HTTP Error 404: Not found

Hello fellow onion lovers! Winter is here, and with it gloomy overcast days and seasonal depression. Yay!

Half kidding aside, September and October have been a great couple of months. Saw a great performance of Jane Eyre with Shari, Cirque du Soleil’s Volta with my family, took care of my grandma a few weeks, fun Amazon team outing that was littered in Henry’s art (the same artist that decorated our tor office)…

flatsticks art

… and of course tasty, tasty code.


Stem 1.7

No surprise the big ticket item of these last couple months has been Stem 1.7. In the works for almost a year, this release included numerous features and fixes.

Check it out!


Membership Audit

Tis that time of year again! Per our membership policy and my duties as our membership secretary I’ve sunk several weeks into reviewing what folks have been up to. Lotta neat things!

Nothing too notworthy to announce at the moment. The process is chugging along reasonably smoothly. I’ve attempted to tailor it the best I can as an opportunity to re-integrate fringe members of our community that are drifting away. In some cases folks indicate that work, life, and other hobbies have made it time to move on. In others we’re brainstorming fun projects they might enjoy.

Curious to see how it goes!


Smaller things these last couple months include…

Peeking out from the engine room, I’m delighted to announce Stem 1.7. A full year’s accumulation of fixes and improvements.

What is Stem, you ask? For those who aren’t familiar with it Stem is a Python library for interacting with Tor. With it you can script against your relay, descriptor data, or even write applications like Nyx.

https://stem.torproject.org/

So what’s new in this release?


ORPort Descriptor Downloads

Stem can now download descriptors through ORPorts just like Tor!

reply = stem.descriptor.remote.get_server_descriptors(
  endpoints = (stem.ORPort('128.31.0.34', 9101),),
)

This is just the tip of the iceberg for the ORPorts capabilities we hope Stem will have. Whats next, python tor clients? Relays? Stay tuned!


stem.directory module

In collaboration with teor, Stem now provides authority and fallback directory information through our new stem.directory module

import stem.directory

COLUMN_FORMAT = '%-17s%-20s%-10s%-10s'

try:
  authorities = stem.directory.Authority.from_remote()
except IOError as exc:
  print('%s\n' % exc)
  authorities = stem.directory.Authority.from_cache()

print(COLUMN_FORMAT % ('Name', 'Address', 'ORPort', 'DirPort'))

for authority in authorities.values():
  print(COLUMN_FORMAT % (authority.nickname, authority.address, authority.or_port, authority.dir_port))
% python demo.py 
Name             Address             ORPort    DirPort   
maatuska         171.25.193.9        80        443       
tor26            86.59.21.38         443       80        
Bifroest         37.218.247.217      443       80        
longclaw         199.58.81.140       443       80        
dizum            194.109.206.212     443       80        
bastet           204.13.164.118      443       80        
gabelmoo         131.188.40.189      443       80        
moria1           128.31.0.39         9101      9131      
dannenberg       193.23.244.244      443       80        
Faravahar        154.35.175.225      443       80        


Descriptor Compression

Stem now supports Tor’s new ZSTD and LZMA compression. ZSTD and LZMA’s higher compression ratio comes at a CPU cost and are less available, so when using these to save bandwidth you should provide a fallback…

reply = stem.descriptor.remote.get_server_descriptors(
  compression = (
    Compression.LZMA,  # higher compression but often unavailable
    Compression.GZIP,  # decent compression and always available
  )
)

Greetings fellow onion enthusiasts! Like the rest of the summer August has been less tor-ry than usual. This month went toward eviscerating my home in an act of utter apartment seppuku.

apartment mess

Truly it’s amazing how much stuff closets accumulate. Dumped my every worldly possession in a pile and now dealing with the mess. I’ll be at this through September as well but so far so good. Eight full grocery bags of stuff to rid myself of and counting. Fun project!


Stem 1.7 Release Prep

As for the technical front… *sigh*. I told Matt I would release Stem 1.7 when he asked for it but six weeks later and still no release. Dealing with several pre-release issues…


Tor Summer of Privacy

This month Dave concluded his Tor SoP project. Honestly we didn’t get too far. I’ll probably advertise this project again in next year’s GSoC/SoP. That said, Dave sent quite a bit of code at the end so much of August went toward reviewing that.

PS. Already half way through September? Yikes. Between the tardiness of this report and spring cleaning I’m gonna skip my September update. I won’t be in Mexico but hope you all have fun!