- tor-dev - lists.torproject.org

moving toward requiring regression tests for bug fixes
by Taylor Yu 28 Jun '17

28 Jun '17

Hi all, I've opened a Trac ticket at https://trac.torproject.org/projects/tor/ticket/22745 about possibly requiring all bug fixes to have associated regression tests. This is aimed at Core Tor (starting with 0.3.2.x) but other Tor-related software might want to consider a policy like this as well. Ideally all bug fixes will have automated regression tests so we can promptly recognize when they've regressed and fix them. I realize that some of our code may be too complex for an automated regression test to be feasible, so we would have a procedure for exceptions from this requirement. (This would become part of the patch review process.) Please comment on the ticket if you have opinions about this idea. (Or respond in email if your reply wouldn't work well in the form of a ticket comment.) Thanks! -Taylor

1 0

prop224: Time Period Overlaps and Blinded Keys
by teor 27 Jun '17

27 Jun '17

Hi, The time period overlap section 2.2.4 in prop224 is under-specified: https://gitweb.torproject.org/torspec.git/tree/proposals/224-rend-spec-ng.t… 1. During the overlap period, does the service use the new blinded key for the new period, and the old blinded key for the old period? I think the answer is yes, but this requires some deduction to work out. 2. If the overlap period starts when a service sees the first consensus with a new SRV, does the service stop using that SRV and blinded key: * at the end of the period? (that is, exactly 36 hours after the earliest the overlap period could possibly have started.) * exactly 36 hours after the SRV was first seen? (that is, exactly 36 hours after the service started the overlap period. For example, if the service fetched the consensus 2 hours after it was created, it would end 2 hours after the end of the period.) * when the first reveal consensus is received with that SRV as the previous SRV? (or some similar consensus-driven event) Does every service on a tor instance start the overlap at the same time? https://gitweb.torproject.org/torspec.git/tree/proposals/224-rend-spec-ng.t… T -- Tim Wilson-Brown (teor) teor2345 at gmail dot com PGP C855 6CED 5D90 A0C5 29F6 4D43 450C BA7F 968F 094B ricochet:ekmygaiu4rzgsk6n xmpp: teor at torproject dot org ------------------------------------------------------------------------

4 3

Re: [tor-dev] Problems on torrc.d-style configuration directories
by iry 27 Jun '17

27 Jun '17

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi teor and Daniel! Thank you so much for your reply! Your instructions are really helpful to me. teor: > You probably need a %include directive in /etc/tor/torrc. I tried to add %include directive in /etc/tor/torrc and /usr/share/tor/tor-service-defaults-torrc separately. And both of them worked well :) teor: > If you get this working, please submit a patch to the Debian bug > tracker. No problem! But please forgive my ignorance, could you please explain a little bit more to me that why I should report it to Debian BTS, instead of tpo? In other words, what is the relationship between packages.debian.org and deb.torproject.org? According to cypherpunks[0]: > The first released tor version with this feature is 0.3.1.1-alpha. > As usual there will be alpha packages on deb.torproject.org > > If you want this feature _now_ you can use the nightly builds: > https://deb.torproject.org/torproject.org/dists/tor-nightly-master-str etch However, > the highest tor version in Debian BTS right now is 3.0.8-1.[1], which means the feature has not been included into Debian? My current thought is deb.torproject.org is the upstream of packages.debian.org in terms of tor package. So a change made in deb.torproject.org will be adopted by packages.debian.org after a while. The following is my testing environment which may be helpful to the problem: I tested the torrc.d feature in both Debian 8 and Whonix13(based on Debian8). Instead of downloading from packages.debian.org, I download tor from: > deb http://deb.torproject.org/torproject.org jessie main deb-src > http://deb.torproject.org/torproject.org jessie main deb > http://deb.torproject.org/torproject.org tor-nightly-master-jessie > main The Tor version I tested was: > Tor version 0.3.1.3-alpha-dev (git-a73d0fe9a87df762+b433dff) Again, thank you very much, teor and Daniel! I really appreciate your help! Best, iry [0]: https://trac.torproject.org/projects/tor/ticket/1922 [1]: https://packages.debian.org/source/experimental/tor -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJZUb3/AAoJEKFLTbxtzdU8jQsP/2B29N3WpkypE7DQ1A2wtjd9 MP3Blz9lbJH8LQJw+juaHhzslokCXwpGJSl7OfyqhRL44VkXnTllfd88HacW5luM PqZ4OeljB2UYrpDM7TypjdA+RXmSLTKNCmFifCESQPcmsu97qxlcvkgtF69fJ8eX p/22PEOjsqnows37rYp0AZiLa7x9I1dZdjzD6tHfKNVylM5AuuyOExAZKlx90rlW lL1nCJEuw/d6HBtVWWBgYDjeAMg5G0TpO2gU1j/A7kEzgZxWID6q/r1TvpDp1pi3 aJhjmyeZn2rVwdGPFX03pp7DWJetbYA5CuhGFucjPtkdXDCh2guuSsivcM6obpiA EmI4QtedGGhy2Xp1ufLAHP89TuuSU1n9VaioHSpLlGkQA9v2PRjW4ChBVJwP6mVt 9XGhY7TgEQMVZWL/brCNI6aeCf8rfaBSHtMwAR6xQ6ofvYAjl7X1BpNOSx4s73B/ CWnGxvMNDPuu1scxkOF0y3M68p+xgT0UuHTx4JjUTQtESFTqtdua5JxLyzE2qGyz pwzFjFzwoPBav5TPY7MB8WN79GImZFUiELA6vOL4CLaXEMCepsXYDhdP/3JLkD0e GQfliduoa2KGUijKrzy+ZVrKwBjF63ylbKGrxJHnX1SB8OftYbdlIQRvhzwHwaOL uSfnWGIg3Fo0fKs1/F/Q =L3H9 -----END PGP SIGNATURE-----

3 2

Names for your onions
by heddha 25 Jun '17

25 Jun '17

Good day everyone! A friend and I are looking into implementing a few of the ideas listed in [0] as part of a university project. We saw quite a lot of discussion on idea 1, but we're mainly interested in working on ideas 2.5 and 3, and additionally either 2.1 or 4. Has there been discussion and work on these yet? Where can we find it? Is there anything else that you can tell us that we should know? Thank you very much for your time! Kind regards, heddha [0]: https://blog.torproject.org/blog/cooking-onions-names-your-onions

3 3

When to exit 0 on error
by f55jwk4f＠vfemail.net 25 Jun '17

25 Jun '17

I noticed tor sometimes exits 0 on error, specifically in monitor_owning_controller_process in src/or/control.c: if (owning_controller_process_monitor == NULL) { log_err(LD_BUG, "Couldn't create process-termination monitor for " "owning controller: %s. Exiting.", msg); owning_controller_process_spec = NULL; tor_cleanup(); exit(0); } and in try_locking in main.c: r = try_locking(options, 0); if (r<0) { log_err(LD_GENERAL, "No, it's still there. Exiting."); exit(0); } When to exit 0 on error?

3 5

Problems on torrc.d-style configuration directories
by iry 22 Jun '17

22 Jun '17

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hello everyone! I met some troubles when testing the new torrc.d-style configuration directories feature (https://trac.torproject.org/projects/tor/ticket/1922) Could anyone share some ideas on the problem, please? I will be really appreciated. The following is what I did: 1. add the following line to /etc/apt/sources.list.d/torproject.list: deb http://deb.torproject.org/torproject.org tor-nightly-master-jessie main 2. update Tor to nightly version: sudo apt-get update 3. tor --version: Tor version 0.3.1.3-alpha-dev (git-a73d0fe9a87df762+b433dff). 4. sudo mkdir /etc/tor/services-available 5. sudo cp anon-connection-wizard.torrc /etc/tor/services-available/ 6. sudo mkdir /etc/tor/services-enable 7. sudo ln -s /etc/tor/services-available/anon-connection-wizard.torrc /etc/tor/services-enable/anon-connection-wizard.torrc 8. reload tor 9. since bridges are used in the anon-connection-wizard.torrc, when we use arm to check the connections, tor should connect to one of the bridges, if torrc.d style configuration worked. However, it didn't work. I also tried to remove the /etc/tor/torrc, and tor could not find a torrc file anymore. The latest related discussion on the ticket is as follows: > Weasel and I (aka Hans) sketched out how we would use it in the > Debian package, closely following the Apache pattern but with > naming that is more appropriate in Tor: > > /etc/tor/torrc:: %include /etc/tor/services-enabled/*.torrc > %include /etc/tor/instances-enabled/*.torrc > > These dirs are present for the actual snippet files: > /etc/tor/services-available /etc/tor/instances-available > > These dirs include relative symlinks to *-available: > /etc/tor/services-enabled /etc/tor/instances-enabled > > For example: /etc/tor/services-enabled/sparkleshare.torrc --> > ../services-available/sparkleshare.torrc > > The sparkleshare package would include: > > /etc/tor/services-available/sparkleshare.torrc > > The davical package would include: > > /etc/tor/services-available/davical.torrc So I assumed it still worked. Could anyone please help me point out what I have done wrong? Or could anyone please point me to some docs about the new feature? I can see the source code of core Tor and write some docs on this, if this will be helpful, btw :) I really appreciate your help! Thank you! Best, iry -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJZRvsbAAoJEKFLTbxtzdU86OkP/REdnsdcJJHoGyTdyXYoqJ6n NgW3cKlEGjCZF5bxIQ+lIG6q85j3EfflPDov97VrVyCE8xr6TFG0dDs/sE4rPE/h GjYhBzM1WC5o55vyvdehTh80wl3/hSj/LehCJLOpBKAjoxuzeDHtyrkd3LG22y8i mi93wgfxBuEbqvstDVipWfJt/5JbBLtXS/QainQqFwwFHRxJpIZcSgvWR5UAXAyS IQLW7m91lwl9W/irQvhM8d4nnk0ciTGMPKoyEt3FWHs9jP9ewKvIvWlxJVYZEefW LlLHMVzbyqWZWnkLkCBNyBfZZ0Rk16YFrUuMkTUH83LO/kBSlWFOj3idN5gYwda/ aanJhtopLH1u1nP10TqF6c4e6Bu6tfCOUh8sdOp8s7iB7MJHvaPU3lJM1vDk55u1 +fSxcDOm7NuN7xrJ3elUKS+oeOU8Ah4h43L/9lJgrIZ6zdIdI1f0E4QXlIWbKIbf dkQJrG760R/6jG6JMnbiOFPwoFelLna1hX6USpbnVEfhh33n3t/t5EpK5kH/q2m9 /lCkMcBpa5XXVquGaeCCQ+2X4wQ31wlhka12GADtVkq3SohOnfKzRauJYN8OcHZH eNZlDMbO8j3y9IHEdM+mhT8a7XnnZokFINSjmjnOnAQdk8A8QjB5V/TIZQB1uJwd 1uUyayVwl9SQHy3P6yuH =SLHC -----END PGP SIGNATURE-----

3 2

[release] metrics-lib 1.9.0
by iwakeh 21 Jun '17

21 Jun '17

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Hi there! Another planned release of metrics-lib is available: https://dist.torproject.org/descriptor/1.9.0/ The last release preparing the non-backwards compatible changes to come in version 2.0.0. The top three changes: * Simplify DescriptorReader by returning Descriptor instances rather than DescriptorFile instances containing Descriptors, and add a File reference to Descriptor. * Introduce a new UnparseableDescriptor to be returned by DescriptorParser and DescriptorReader if a descriptor cannot be parsed. * Let DescriptorParser return an Iterable instead of a List, which prepares parsing large descriptor files descriptor by descriptor. For more details see the changelog [1]. Please direct comments and questions to the metrics-team mailing list [2]. Cheers, iwakeh [1] https://gitweb.torproject.org/metrics-lib.git/plain/CHANGELOG.md?id=descrip… [2] https://lists.torproject.org/cgi-bin/mailman/listinfo/metrics-team -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBCAAGBQJZSnx7AAoJEPeEx9Sa/gviQ80QAK9m1muFju1dFqdPM6k6R76z O5XyYWUndNcWQhvZ8wYNM1UjAWalhRnuPBVBFp1wkBtywOnH7YWAerXwtK8s/t2s 3QSmR9eMPSjmDKD07QzRxaiwSmW276qUFvkRt71EGTCI0Cz4PF5RnBCuhUxbjtQa c38k6v2l7ZfYpfMPLhuPgtuLZCFDIOXAp48q1xCReHAltsOEskOh8zUarDF9+Ivr BD0o8Wc4GSEl7rQcVTEz+iaheYWJOV5C02oMSBPltdCxP7ug0y+OiNlucRe29OeG 7dUfDF1PrufCQY1Bs1Q5/ny07xpW5ARG07goalcRz2KhJCUkj/fVUAK+kOqN+s2t NrzVSgkMLyp7NEEKwZ/ZdEZQ/hv14ksLA0L3P8ru1pvGSa/RPV+LFwu4AzAvfWHJ m3JNcfJwM0CpWZPtQFjSYlseYh9Ogr9VAFENV4F+8YwgCUGnX02BSFV+B0T0AVw4 okwzY4W8jjaSjBDkp+lNE+0VEq3We5vSXKJTaHXV4Eic9kXxNEs7jqsbY2DsA0gp 4/JNSziXCmTCchajEKUDFUyBGhrbQ153klff8kyC/OkZ8Z6mben8Y8isSgC9Qydc li1anXWlXMklf6wj0cGQCP4Ovi07rUcMrx/6gTngyIQ5xOmGTUHbE3y0T1w+VTw0 G+MpWFQxiGWnVoZG20vj =NFWS -----END PGP SIGNATURE-----

1 0

Tor master branch is now 0.3.2.x
by Nick Mathewson 19 Jun '17

19 Jun '17

Hi! Quick heads-up: development on the 0.3.1.x series is now in the maint-0.3.1 branch. The master branch is now for the 0.3.2.x series. So if you're fixing a bug in 0.3.1, base your patch on maint-0.3.1 (or some appropriate earlier maint-x branch) ... and use master for new features. Confused? See doc/HACKING/CodingStandards.md for info on how we use git. cheers, -- Nick

1 0

Are we planning to use the "package" mechanism?
by Nick Mathewson 19 Jun '17

19 Jun '17

With proposal 227 in 0.2.6.3-alpha, we added a way for authorities to vote on e.g. the latest versions of the torbrowser package. It appears we aren't actually using that, though. Are we planning to use it in the future? -- Nick

4 3

Extending Tor stats to cover anon OSs?
by bancfc＠openmailbox.org 18 Jun '17

18 Jun '17

@TPO devs Since you do a great job safely collecting useful stats on the network, would you be open to adding a self-identifying anonOS distro option to the protocol? Would this be OK or is it mission creep? On the flip side it would be much more accurate than anything we can do to estimate active user base. Some ideas: *The distro name options could either be hard-coded values (for example TAILS, Whonix...) or a custom one chosen by downstream maintainers. *Distro maintainers would enable it for Tor clients via torrc.d

2 1