-
Notifications
You must be signed in to change notification settings - Fork 1k
Weekly check in 2012.08.16
Andrew Byrd edited this page Dec 17, 2014
·
1 revision
- 13:33 <demory> hey, this a good time for the check-in?
- 13:33 <abyrd> sure
- 13:33 <demory> don't want to cut short any conversations
- 13:33 <demory> ok
- 13:33 <mattwigway> yep
- 13:34 <demory> mele grant_h, you know if Frank is joining today?
- 13:34 <grant_h> haven't heard that he's not
- 13:34 <mele> he's on vacation for a while starting today actually, so i don't think so
- 13:34 <grant_h> oh that's right
- 13:35 <novalis_dt> I've been working on a possible way to handle search bounding in RAPTOR with some precomputation . It's going kind of slowly, but I think I'll eventually work out all the bugs. Then we'll be able to see if the precomputation can be done in a reasonable length of time, and if so, how much it helps.
- 13:35 <abyrd> I checked in and merged some updates to realtime stop times updating that address concurrency, as well as code that indexes patterns over a certain size and performs binary searches for depature lookups.
- 13:36 <demory> My main focus has been follow-up on the bikeplanner.org launch. Lots of good coverage for OTP -- working on a blog post now w/ rundown of that. Also, bikeplanner should be working fully on IE8 now, still looking into some Android issues. Let me know if anyone catches other issues/platforms that need attention.
- 13:37 <mattwigway> I checked in a lot code to master, which accomplishes two things. It is now possible to generate shortest-path trees that are already reverse-optimized, and I also merged PatternBoard and PatternAlight.
- 13:38 <abyrd> (details in recent dev list post). I will be adapting the stop times updater to use that code, and look into fetching several consecutive departures at once.
- 13:38 <mattwigway> The reverse-optimization in conjunction with the clampInitialTime option allows one to control how wait time figures in to Analyst results.
- 13:39 <abyrd> Also want to try otpprofiler and Sonar on a rackspace openstack server that could be triggered nightly
- 13:41 <demory> abyrd, so is otpprofiler ready for general use/testing?
- 13:41 <demory> sorry i haven't really been keeping up w/ that like i should be
- 13:43 <abyrd> not really, it's in its very first iteration. I'd say it's ready for general use when it's running nightly on a server, and has at least a minimal front end to request profiling runs by specific commit hash
- 13:44 <demory> ok, thanks
- 13:45 <demory> grant_h mele, any news from TriMet? Everything ok w/ the new ride.trimet.org?
- 13:46 <grant_h> Everything's going well for the most part, mostly positive feedback
- 13:46 <mele> yeah and it's petered out so people seem to be adjusting to it just fine
- 13:46 <mele> the feedback has, that is
- 13:46 <mele> getting usual traffic & everything
- 13:47 <mele> personally grant and I are just waiting on a few bug fixes and improvements from y'all and Frank, keeping on with our editing
- 13:48 <novalis_dt> Yeah, I would like to get to those, but I kind of want to wrap up this raptor thing first
- 13:48 <grant_h> btw novalis_dt wanted to let you know the deadline for session proposals for State of the Map is August 31st
- 13:48 <mele> no problem, i don't think frank will be back for 10 days anyway so nothing could be pushed up on ride until after that
- 13:49 <demory> yeah, i intend to do some more work on the elevation chart in the coming days
- 13:49 <novalis_dt> I just submitted yesterday, grant_h.
- 13:50 <grant_h> nice
- 13:50 <mattwigway> Oh cool, you're doing a session on routing using OSM?
- 13:50 <mele> demory we've had a couple different emails lately about wanting that close button to be more obvious-- they didn't realize it was there at all
- 13:51 <novalis_dt> mattwigway, Yeah
- 13:51 <novalis_dt> Well, if they accept it
- 13:51 <demory> mele, which close button are you referring to?
- 13:51 <mele> for the elevation panel
- 13:51 <mele> hide/show button
- 13:53 <demory> oh yeah -- I actually pushed some changes w/ that a while ago but Frank said he wasn't able to use them yet -- something about how it interacted with his add-on code for the bus schedule panel, i think? (which isn't a part of the core OTP UI)
- 13:53 <demory> I need to get w/ him on that when he is back
- 13:53 <mele> yeah i saw that it works better now but i'm worried people still won't see or understand the button as-is
- 13:55 <demory> ok. that button is part of the Ext widget -- i'll see if there's a way to make it more prominent
- 13:56 <mele> thanks! I think it just threw off some older users that just take transit and wanted the elevation out of the way so they could get map space. they couldn't figure out how to hide it.
- 13:58 <demory> yeah, another option would be to incorporate a more obvious "hide" button/link into the panel itself. but i'll see if there's anything quick w/ Ext first
- 13:59 <demory> so, anything else?
- 14:00 <novalis_dt> Nothing here
- 14:00 <demory> Oh, one quick plug -- we're in the home stretch of the OTP/iOS kickstarter campaign. please share w/ your circles if you haven't already!
- 14:00 <demory> http://www.kickstarter.com/projects/228865951/transit-app-for-ios-6-and-beyond
- 14:00 <mele> yep, all over it :)
- 14:01 <demory> awesome, thanks!
- 14:02 <demory> ok, sounds like that's it for today. have a good week everyone!
- 14:02 <novalis_dt> Later!
unless you are intentionally working with legacy versions of OpenTripPlanner. Please consult the current documentation at readthedocs