[Thali-talk] Thali daily status - 8/12/2015

[Yaron Goland]

Matt - He figured out how to fix our retry behavior so things don't stop working when Bluetooth goes belly up. He submitted a PR for the fix for review and suffered through listening to me drone on about story 0.0.1 for awhile. Today he will submit a PR for the mux tests and start working on story 0.0.1. He also is going to talk to Jukka about issues we are seeing in Bluetooth where retrying failed connections takes a really long time.


Srikanth - He spent yesterday fixing various issues I found with his PR. Today he has a few more issues to fix and he needs to talk with David to figure out if it's worth doing the drop down work or if he should just wait for David's new version of the post card app.


Yaron - Spent a bunch of time reviewing Srikanth's checkin. Had a call with our customer and reviewed a doc they wrote on how they want to use data. Had a long chat with Sreejumon and Matt about the work needed for story 0.0.1. Had a call with Doug. And spent more time on Srikanth's code review. Today I need to review Jukka and Matt's PRs and try to get Jukka's BLE code running on my devices.


David - He is making good progress on the new version of the post card app. He is right now just squashing bugs. He should have the story 0 logic up and running by the end of this week.


Luke - He has finally stopped working on his old project and is now working on ours. Yesterday he fought with issues with PouchDB not properly compiling LevelDown (answer: Install Python 2.7) and is writing test code to get replication flowing so he can better understand it.

Jukka - He re-wrote the wi-fi direct layer to instead use BLE and is testing that out. He is a bit hamstrung because he only has one device that supports being a BLE peripheral. We are mailing him two more devices but due to international shipping regulations dealing with batteries it will take a little while for the devices to reach him. In the mean time he will start doing testing work on it. I'll also try to help out by testing his code with my Android phones which I believe do have the appropriate hardware. Jukka will also investigate the Bluetooth timing issues we are seeing.

Sreejumon - Yesterday he finished up some issues with Winston (the JS logging framework) and was forced to listen to my story 0.0.1 speech. Today he should submit the PR adding Winston and will then start working on Story 0.0.1.

Oguz - He is still working on the GitHub web hooks and hopes to have them working by the weekend.

Toby - He is running into a bunch of fun issues with iOS. What is happening is that in multi-peer connectivity there is the concept of a session. And within a session there are streams. The problem is that a session can exist without streams and sessions can actually exist without a peer. This is because iOS caches session data and just sort of assumes a session is still there for some time even if the peer has disappeared. None of this is a show stopper but it makes our retry logic fail because our existing code makes incorrect assumptions about how long sessions can last. So Toby is having to put in smarter logic that understands things like "Just because iOS thinks a session still exists, it might not." Nothing earth shattering but it's the kind of detail we have to master to make the system production ready. I'm sure Toby will get a great case study out of this.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist10.pair.net/pipermail/thali-talk/attachments/20150812/6dcdf734/attachment.html>