2012.06.14 Weekly Check In - camsys/OpenTripPlanner GitHub Wiki
13:31 <demory> good afternoon/morning/evening everyone
13:32 <demory> grant_h mele, is Frank joining today? should we wait for him?
13:33 <mele> Hm, no reason why he wouldn't be that I know of
13:33 <grant_h> Yeah, I think he'll join, but I doubt he'll mind if we go ahead and get started
13:34 <demory> ok
13:34 <demory> we'll let's go ahead then
13:34 <demory> my update -- opentripplanner.com live! otp.org continues to point to wiki for technical reasons but eventually it will redirect as well. also, working on some deployer workflow / demo issues and announcement to transit developers list. and beginning to look at logging issues discussed last week.
13:35 <novalis_dt> FYI, otp.org has some apidocs which we don't want to lose
13:35 <demory> i know, that's part of why we're waiting to do anything w/ that domain
13:35 <demory> i want to sit down w/ Evan sometime and go over the whole domain siutation top to bottom
13:37 <demory> any other updates?
13:37 <novalis_dt> I restored the TSP code, which seems to work
13:38 <novalis_dt> And fixed a few other bugs, including the turn-restriction-circumvention bug
13:38 <novalis_dt> Which was fun times
13:38 <mele> yeah thanks for that, it seems to be running perfectly now
13:39 <jcwong86> The CaBi/transit analysis work is complete with posts on GGW and OpenPlans and a rep on github. There was someone on earlier @cathal who was doing similar work and was curious about whether or not OTP could handle multiple simultaneous requests. If I get bored, there are some comments on the blog with other things to do with that data.
13:40 <novalis_dt> mele, awesome. I'm just thrilled that my semi-joke proposal of using DFAs to model path restrictions turned out to make sense
13:40 <novalis_dt> jcwong86, are you on the mailing list?
13:40 <jcwong86> i don't think i am
13:40 <novalis_dt> jcwong86, there were also some responses to cathal there
13:40 <novalis_dt> So you might want to check that out.
13:41 <abyrd> I'm mavenizing the previously hard-coded analyst many-to-many mode a bit
13:42 <abyrd> and using it to prepare some plots for a seminar tomorrow
13:42 <abyrd> erm, I mean springifying
13:42 -!- FrankP [d819d197@gateway/web/freenode/ip.216.25.209.151] has joined #opentripplanner
13:43 <abyrd> thanks David for restoring that TSP code. sorry you got stuck with that.
13:43 <novalis_dt> abyrd, eh, it happens
13:44 <abyrd> jcwong86, are you still interested in feedback on your methodology paper?
13:44 <jcwong86> sure
13:45 <jcwong86> i don't have a real plan for it, but would like it to be as accurate/correct as possible so it makes sense to someone else if they go to the repo
13:46 <kpw> also: we're talking about a response to the iOS 6 developments from this week
13:46 <abyrd> novalis_dt, actually there are several papers suggesting use of DFAs for blocking nonsensical mode transitions
13:46 <abyrd> so it's taken seriously by others as well
13:47 <novalis_dt> abyrd, Good to know.
13:47 <kpw> current plans include: developent of an open source iOS client
13:47 <demory> yay!
13:47 <novalis_dt> jcwong86, I am curious to see what results you get on DC Bikeshare time vs OTP planned time given the new slower bikeshare speeds I put in.
13:47 <kpw> and examining ways to provide comprehensive coverage for public GTFS feeds
13:48 <mele> YAY!
13:48 <kpw> lookign at some creative ways to route searches to the right otp instance
13:48 <kpw> based on lat/lon
13:48 <kpw> should be fun
13:48 <kpw> our line is that iOS is the best thing for transit since GTFS
13:48 <kpw> er iOS 6
13:49 <demory> yeah this is a great opportunity for the app developer community
13:49 <demory> and we want OTP to be out in front
13:49 <demory> what's the timeframe on iOS 6?
13:50 <mele> I'm so happy to hear this, guys. I agree that OTP has a great opportunity here. I think it's sometime in the fall, right?
13:50 <novalis_dt> Apple typically doesn't announce release dates IIRC
13:50 <novalis_dt> Apparently, "Fall" is as good as we get.
13:50 <demory> ok
13:50 <kpw> fall
13:53 <FrankP> Is there a foundation to start on the iOS app? PdxBus code is open source, and it's the most popular app here...
13:53 <mele> it's a really excellent app, I'd love to see a lot of the features carried over
13:54 <FrankP> but I said jinx, so no talking till someone says your name, Mele...doh!
13:54 <novalis_dt> kpw, you were going to be the one writing the ios app, right?
13:55 <demory> what does PDXBus currently use for trip planning?
13:55 <demory> your OTP API by chance?
13:55 <FrankP> my current ATIS wrapper.
13:55 <demory> ah
13:55 <novalis_dt> Which is very similar to OTP
13:56 <FrankP> I have it on my plate to wrap OTP with the same api (so that abandoned apps that are still used by customers don't break)
13:57 <FrankP> (e.g., don't break when we turn off ATIS)
13:57 <demory> right
13:58 <demory> but i guess the thinking is that actively maintained apps would migrate to the OTP api?
14:00 <FrankP> Yes, I hope so. That said, the wrapper is different than OTP, in that you can specify just a name in the request, and the wrapper will geocode that for you...so that might be a disincentive to switching (just having a bridge might be a disincentive)
14:00 <FrankP> Honestly, I'd rather get out of the API hosting business...but a member of the press asked me specifically whether "we'd continue to support our customers" (him), so I feel like I'm on the hook here.
14:02 <demory> hmm, ok
14:04 <demory> well, there will be plenty more to discuss about the IOS stuff going forward. we're just getting started w/ it now
14:04 <demory> anything else for today's check-in?