On Mon, Jul 1, 2013 at 6:40 PM, Meredith Whittaker < meredithrachel@google.com> wrote:
Hi Aaron,
Thanks for these notes! From vacation, I'll answer your M-Lab questions quickly (also adding Stephen).
The collector policy implementation is a blocker for deployment on M-Lab. To complete deployment, it will be necessary for M-Lab to configure this collector policy, to ensure that the broader M-Lab privacy/openness policies are adhered to. Thus, I believe it should be a part of this feature-complete milestone, since it is fundamental to launch.
I can review the M-Lab policies, if that's helpful. However, I don't think the collector policy should be designed for M-Lab, but should be designed to allow M-Lab and any other collector to implement their own policies, per their legal/organizational requirements.
Let me know if you have any questions here. Happy to discuss further.
Ok, that's what I thought and believed was part of the discussion that we had in Berlin.
Tom, can we agree that this is an issue we must address for release 1.0 and factor this into our release agenda?
In the meantime, on the topic of changing the meeting from Hangouts to "other" -- I have a phone bridge with international dail-in numbers. If it's helpful we can use it.
Cheers, M
Thank you for taking time from your vacation to reply. Hope you are having a blast :-)
--Aaron
On Mon, Jul 1, 2013 at 12:58 PM, Aaron aagbsn@torproject.org wrote:
Present in call:
Tom, Aaron, Arturo, Johannes
What we have worked on this week:
Arturo worked on twisted socks client support and NetTest specification Aaron worked on specification and implementation of the test deck and test input api for ooni-probe and ooni-backend (117, 115, 109).
Are there any items blocking?
No blocking items were raised.
Are we on track to make the first milestone (Fully Specified)
Aaron and Arturo say yes.
What are we working on this week:
Aaron and Arturo will be working on issues 115, 117, 91, and 90 this week.
Issues that we discussed on the call:
Tom states that collector policy implementation (issue 109) is not part of the feature complete milestone and instead should be in future versions. Aaron and Arturo believe that collector policy was a blocking issue for MLAB deployment as discussed at the Berlin meeting. We would like to request feedback from MLAB on this point.
Arturo asked for clarification on issue 107 (do we need to specify each unit test?) The resolution was that the unit test spec should be replaced with a code coverage report as per Nathan's suggestion. Johannes suggests coveralls.io which writes a code coverage report into the README.md.
We discussed alternatives to using Google Hangout for our weekly meetings Johannes suggests: Mumble. Tom suggests: phones. Can we expose a phone number for call ins into the Google chat? No, but it can dial out worldwide (US and Canada free). Aaron Tom and Arturo agree that video chat is nice. Mumble now has a voice-activation feature and Johannes has a mumble server that we can use. Google Hangout has some issues on Johannes computer (system load > 4). We agree that the four of us cannot reach a consensus without others present and will take this discussion to the mailing list.
See you all next week!
--Aaron
ooni-dev mailing list ooni-dev@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/ooni-dev
--
Meredith Whittaker Program Manager, Google Research Google NYC