2012.07.19 Weekly Check in - atlregional/OpenTripPlanner GitHub Wiki

  • 13:32 <demory> good afternoon/morning everyone
  • 13:33 -!- novalis_dt [[email protected]] has joined #opentripplanner
  • 13:33 <novalis_dt> Hi! Sorry I'm late.
  • 13:33 <novalis_dt> I found an interesting paper on an alternate routing algorithm (RAPTOR), so I'm implementing it (while taking breaks to fix other bugs)
  • 13:34 <demory> novalis_dt can you share the link to that paper?
  • 13:34 <demory> i haven't seen it yet
  • 13:35 <demory> My check in -- lately i've been working on cleaning up the elevation front end issues in #776 / #649. FrankP I'm getting ready to check in some code that will hopefully resolve those for RTP.
  • 13:35 <demory> Other stuff this week: helping w/ data prep for the new Analyst work, general catching up on Deployer issues, etc.
  • 13:35 <novalis_dt> demory, http://research.microsoft.com/apps/pubs/default.aspx?id=156567
  • 13:35 <FrankX> Thanks David...I'll be on the lookout.
  • 13:36 <demory> novalis_dt, thanks
  • 13:37 <FrankX> My update: we had a good bug in the router, which Dave fixed yesterday. I also 'improved' the itinerary option formatting, as well as the trip details. let me know if there are objections...
  • 13:37 <FrankX> http://maps5.trimet.org/opentripplanner-webapp/?submit&fromPlace=45.517638,-122.530111&toPlace=45.515713,-122.722372&maxWalkDistance=2000#/
  • 13:38 <novalis_dt> Yeah, the itinerary looks nice.
  • 13:38 <novalis_dt> Maybe "0 transfers" should just be "direct"
  • 13:38 <mele> FrankX , look at the student and senior fares :)
  • 13:39 <abyrd> this week I've been analyst working on the OTP Analyst batch/aggregate framework. this serves to evaluate accessibility metrics over large areas (walk/transit heatmap style)
  • 13:39 <novalis_dt> mattwigway, did yo uget a chance to see if my fix worked on #782?
  • 13:39 <mattwigway> yep, I closed it just after you left.
  • 13:39 <novalis_dt> mattwigway, oh, awesome
  • 13:40 <abyrd> using some data prepared by demory from the us census and the initial wait time fix prepared by mattwigway I've got some preliminary results as a geotiff
  • 13:40 <FrankX> Student & senior fares are a way to close our budget shortfall, mele (seriously, a config issue on the trimet side I haven't checked in yet)
  • 13:40 <mele> haha
  • 13:41 <abyrd> but the real point is that the analysis is set up via spring XML and uses pluggable csv/shapefile/raster loaders and aggregate functions so it can be readily adapted to other studies
  • 13:41 <kpw> my update: we're just about to launch our kickstarter campaign for an iOS 6 app and a north american OTP API end-point
  • 13:43 <demory> great
  • 13:44 <demory> oh one other thing: maybe this is a good time for a 0.7.8 release?
  • 13:44 <demory> sounds like there have been some significant fixes since 0.7.7
  • 13:44 <demory> i'd like to include the elevation fix which should be ready this afternoon
  • 13:44 <novalis_dt> I'm a bit nervous, because of the mail we just got about a possible regression on before-plan arrival times
  • 13:45 <demory> hmm
  • 13:45 <demory> are we sure he's using the latest version?
  • 13:45 <novalis_dt> Nope.
  • 13:45 <demory> that issue reappeared in 0.7.5 or 6 i think, then was fixed again
  • 13:45 <demory> in 0.7.7
  • 13:46 <novalis_dt> Ah, I'll ask about that too
  • 13:47 <FrankX> BTW, is someone replying to Sean B. about a new OTP group for Android? Should it be a 'mobile' group for both Android and iOS?
  • 13:48 <novalis_dt> Well, what's the plan on the iOS transit support?
  • 13:48 <novalis_dt> Does Apple have an API? Or are we expected to develop our own client?
  • 13:50 <FrankX> p.s., forgive me...the Sean B. email was from March 29th .... it's just gotten a reply today, so it appeared back atop my email
  • 13:50 <novalis_dt> Well, it's a valid question anyway
  • 13:50 <novalis_dt> I guess I would be happy with another mailing list if someone else wants to admin it
  • 13:51 <demory> kpw and I are chatting w/ Sean about another project next week, i'll see what his latest thoughts are about this
  • 13:55 <demory> going back to releases -- assuming we don't hear anything about the arrival time still being an issue, let's plan on cutting 0.7.8 before the weekend?
  • 13:55 <novalis_dt> OK.
  • 13:55 <demory> there's at least one broken Deployer request waiting on that
  • 13:55 <demory> ok, anything else for today?
  • 13:55 <mattwigway> I'll just mention that I'm working on making Analyst requests more closely match requests to stock OTP.
  • 13:56 <abyrd> demory, no problem for the release, whenever we're confident about the code quality
  • 13:56 <demory> ok, thanks
  • 13:56 <grant_h> The redaction bot perfroming the OSM license change passed through Portland
  • 13:56 <abyrd> what differences are you eliminating
  • 13:56 <abyrd> ?
  • 13:56 <grant_h> With little damage
  • 13:57 <novalis_dt> grant_h, excellent!
  • 13:57 <mele> there was a bit of a mess in yamhill and clark counties and with the columbia though
  • 13:57 <mele> i'll be cleaning up for a bit
  • 13:57 <mele> it's generally done with the US, so you guys will want to check your areas for obvious missing pieces
  • 13:58 <novalis_dt> Do you have a suggestion for how to do that check?
  • 13:58 <abyrd> mattwigway, which differences between trip planning OTP and Analyst OTP are you referring to?
  • 13:58 <mele> there seem to be a lot of unconnected, untagged nodes about
  • 13:59 <mele> that you can check for in josm by filtering out everything but nodes
  • 13:59 <mattwigway> The elimination of initial board time (done, but will need to be revisited once my other changes go in) and transfer time.
  • 13:59 <mele> you can also do a find in JOSM for 'user:"OSMF Redaction Account"'
  • 13:59 <abyrd> oh, ok, so only the new API params were different
  • 13:59 <mele> but mostly just look at major roads for errors that pop to your eye
  • 13:59 <mele> missing parts of freeways and ramps, etc.
  • 14:00 <novalis_dt> OK, thanks. If I have time this ewekend, I'll look into that
  • 14:00 <grant_h> Columbia, South Carolina and the Southern California are talked about as place where damage was the worst
  • 14:00 <FrankX> mele, will the portland area's OSM be ready for a production Graph.obj? (I want to release a new graph next week with new OTP code) Or should I stay with pre-redacted data for awhile?
  • 14:01 <mele> stay with it for at least another few days
  • 14:01 <mele> i should hopefully be able to get everything i want to fix fixed by the end of the weekend
  • 14:01 <FrankX> okay...I'll ask next week ; old data saved off if need be...
  • 14:02 <mele> yeah i'll send you an email when i think it's ready. we should definitely run the tests too
  • 14:02 <mele> i'll be out of town at the esri UC next week but will have my laptop
  • 14:04 <mattwigway> novalis_dt, got a minute?
  • 14:05 <demory> ok i guess we're done w/ the group check in, thanks everyone