Hacker News new | past | comments | ask | show | jobs | submit login
Bay Area Regional GTFS Feed: 31 transit agency feeds together in one API (interline.io)
136 points by wuster on Jan 9, 2020 | hide | past | favorite | 49 comments



On one hand, this is a great effort and I'm sure will help commuters a ton.

But on the other, why on earth does the bay area have 31+ different transit agencies??


If it actually worked well, people might use it, especially the undesirables.

It’s also because the Bay Area is stupidly split up into dozens and dozens of small-medium sized municipalities spread over multiple counties. So every little group acts in their own locally optimal, shortsighted interests. Most of the people affected worst by the lack of public transportation and housing can’t vote in the areas that can actually fix the problem.

I’ve been thinking the past couple of years you should be able to vote in local elections in the area you work in, even if you don’t live there.


> I’ve been thinking the past couple of years you should be able to vote in local elections in the area you work in, even if you don’t live there.

In many areas this would effectively lead to the suburbs being able to dictate urban policy to the inner city, where suburban commuter interests like parking override the interests of the actual residents. The 1998 amalgamation of Toronto is a good example of this happening, and I think a lot of Old Toronto voters are quite unhappy with this, given how it led to mayors like Rob Ford who would never have been elected with the old boundaries.

Other areas try to solve this by creating an uber transit authority which theoretically directs and coordinates the smaller agencies for the greater regional good. A good example of this is the Regional Transportation Authority in Chicago. It often doesn't work out the way it's intended (in Chicago, the CTA [city proper transit] and Metra [commuter rail] still have very poor coordination, although Pace [suburban buses] and CTA do have somewhat good coordination).

A third approach is to (try to) make the whole region's transit the responsibility of one single agency. Picking Atlanta as an example here (MARTA), it tends to lead to affluent suburbs (Cobb and Gwinnett Counties) trying to stay out of the system because of concerns like "transit brings crime" and "it's too expensive and no one will use it" and other assorted nonsense. So that approach has its problems too.

In short, this is a very thorny problem and there honestly aren't a lot of places in North America that do it very well, although some are worse than others (the SF Bay Area may seem like a mess, and it is, but it's inarguably better than the dozens of barely-funded, not-at-all-sufficient systems that exist in most large and medium cities in the U.S.). I think the right solution is probably specific to each region and it still won't solve every problem, at least not without a level of funding that transit simply does not get on this continent.


Perhaps a better example than MARTA is Dallas’s DART: it’s one transit agency, in total control of all the busses, rail, and streetcars within the service area, which includes all of Dallas, almost all of the suburbs, and even the wealthy enclaves like the separate township of Highland Park.

From a customer perspective, it seems to work well: one map, one timetable, one fare payment program, one branding, regardless of what city or suburb or wherever you’re in.

It does have the occasional political snag though of suburbs occasionally spreading core service in Dallas proper thinner than they should be (e.g. in order to get DART representatives from the suburbs to sign off on a second subway alignment through downtown, urban Dallas representatives had to agree to fund a suburban-only E-W line that could have been better located also in the city proper). But, each city does contribute an equal proportion of funding (1% of all sales tax revenue) to DART, so it seems fair enough.

The result? From having no rail network in 1996 to having the US’s longest* single light rail network by 2020

*by track mileage; that isn’t to say the breadth of the system doesn’t have its own flaws, mainly that some high density urban areas are woefully underserved in the effort to have a single station in every (very car-oriented) member suburb


> In many areas this would effectively lead to the suburbs being able to dictate urban policy to the inner city

Then perhaps "urban policy" is an unnatural category, combining conflicting interests.

Why not just have two municipal governments—one elected by those whose business interests lie within the city, which would be in charge of the city's business policy (e.g. corporate taxe and grants, arterial infrastructure, commercial zoning); and a separate one, elected by the city's urban residents, which would be in charge of the city's civic policy (e.g. estate taxes and VATs, non-arterial infrastructure, residential zoning, etc.)? These are essentially orthogonal problems that don't really "run into" each-other much; you could have two separate sets of people working on solving them without those groups needing to communicate all that much.

Municipal government is already somewhat factored this way, insofar as e.g. school boards and park boards are separately elected rather than being appointments of the municipal executive; and some of those elections are defined by different political boundaries (e.g. catchment areas for schools) than the election of the municipal executive is. Why not just go one step further?


Because there is no division whatsoever between the policies that those groups may apply.

* Should a lot be used to build a park or a parking lot? The park is advantageous to the school next door, so it is a civic issue. The parking lot is advantageous to the businesses nearby, so it is a business issue.

* Arterial infrastructure as purely business? Utter nonsense. Placing a highway means that all houses within a block are now significantly noisier, a civic issue. It decreases walkability, as there is now one direction that cannot be walked, unless you take a mile-long detour.

* Residential zoning a purely civic issue? Nonsense. How would you ensure that people are close enough to reach a grocery store? How could you have mixed-used development, where the ground floor are shops and the higher floors are residential?

The only way this idea makes sense is if a city has already decided to separate out business and civic areas, which requires a significant investment into car-only infrastructure.


What happens when the business govt wants low corp tax, and the residents govt wants high corp tax to fund programs? What happens when they disagree over zoning, ie which side gets to build somewhere?

I don't think you can segment that way.

One related idea that might work a little better (I think it comes from libertarian circles, can't remember exactly where) is representation by profession. So eg x% of the legislature is elected by residents, and other percentages by other groups like business people.


Representation by industry has worked really poorly for Hong Kong, its a key part of why they have the majority of thr population protesting and demanding real democracy, not the corporate controlled legislature that currently presides over HK.


Thank you for letting me know, I didn't realize it had actually been tried. It appears to have also gone poorly in Taiwan. I absolutely didn't intent to say it was a good idea, merely that it was a similar idea to the parent's that might work a little better.


Regional transit policy needs to be what's best for the region. You can't let city specific special interests stall that because it does not work out well in the long term.

Look at Boston. They canned I695 to stop a couple neighborhoods from being demolished and in doing so basically required all the major surface roads to handle significant commuter traffic (killing how many people over the years?) because they became the de-facto main arteries and horribly compromised future subway lines at the same time (in addition to a laundry list of other negative downstream effects).


Think of all the people who live closer to the urban core in Boston now, due to this decision not making long distance commuting easier.

Portland started ripping out freeways after I-205 destroyed multiple neighborhoods, and Seattlites organized to stop the central district freeway after watching I-5 vivisect the city, paving parks and neighborhoods.


Short term framing of "destroying neighborhoods" is exactly why we can't get anything done. For some reason everyone wants to try to freeze development so that things stay the way they are or revert to some past state. This is frankly stupid and in the long term has terrible outcomes. Look at SF to see what this does. Yes, any time land is taken by the government it sucks for the people there. Nobody is disputing that. What we are saying is that this short term pain is necessary for longer term and bigger picture improvements.

Look at the wealth disparity between the Worcester area (tolerable commuting access by rail and car to Boston) and the Fitchburg area (far less tolerable commuting access by rail and car to Boston) and then say that. As regrettable as it may be the best way for people who do not live in Boston to move up the economic ladder is to commute into Boston where you can make the big bucks and for practicality reasons doing so in a car is the least worst choice available to most people. Literally no party involved likes this arrangement but forcing every commuter that has to go somewhere between I93 and I95 onto surface roads ASAP does nobody any good. I'd call it cutting off your nose to spite your face but spiting one's face implies a level of forethought that was clearly absent.

Any major city is going to have a huge number of car commuters no matter what and decreasing the distance between the highways and their destination makes all the infrastructural changes at the surface road level that make a city less car centric more palletable.

The thing that really pissed me off about I695 isn't the specific highway or lack thereof specifically but that is represents a turning point prior to which it was possible to step on a few toes in order to get something that benefited everyone done and after which everything had to be compromised to hell and back in order to not run afoul of even the most microscopic stakeholder. This has just as big of an impact for public transportation infrastructure as it does for private transportation infrastructure and one would have to be a very near sighted fool to cheer that.

At the end of the day it comes down to regional interests vs local interests. The way I see it is that if Boston had built good highways in the 50s and 60s they would have been building subways in the 90s and '00s instead of digging overpriced tunnels. By caving to special interests they screwed themselves onto on a path that led us to where we are today.


Aren't there enough of us to leverage power by voting as a block?

I know Palo Alto Forward was trying to do this a while back - it'd be nice if we could block vote in people that could massively upzone housing (or make whatever change is necessary to increase housing supply).

Myself and a lot of my friends live in these towns with roommates - I'd think if there was a clear group to vote for and clear policy to push to increase housing we'd have a lot of votes.

It's hard to even find what the existing law around building housing is, is it online anywhere?


You should look into supporting https://www.transformca.org which works with Bay Area governments already to solve transportation policy. They have already contributed immensely to transportation accessibility in the region, and have been at the forefront of the BRT fight, but it's an uphill battle and they need more support.


>I’ve been thinking the past couple of years you should be able to vote in local elections in the area you work in, even if you don’t live there.

Makes sense to me, especially if you are paying local income taxes in that area. I seem to recall from high school that 'taxation without representation' caused a big kerfluffle a couple hundred years ago.


I don’t think there are local income taxes (at least visible to the employee) but they are IMO a part of the community, pay taxes indirectly (property taxes where they work, sales taxes, etc.) and should have a say because of that.


Ahh, maybe in CA. In Ohio you have to pay income taxes where you work and where you live. If they are different places then there are usually some kind of weird commuter credits that never actually offset the difference.


literally everyone who commutes into a city for work wants to also vote there. good luck with that.


Because there is no Emperor of California who can come through, with a single stroke and merge agencies together. Once those agencies were created, they were never going to be merged with other agency.



Or, perhaps https://www.transformca.org which already has a track record of working with local governments and successfully implementing transit policy.


To ensure that we will never have a viable public transportation network that can make high-density mixed-use walkable urban environments possible.


Google Maps and the Transit App already ingest GTFS feeds for trip routing and predicting arrivals.

A regional feed would make things easier for developers, not so much commuters, who are already relying on aforementioned products.

For anyone interested in playing with GTFS feeds in their area, I recommend checking out TransitFeeds' list: https://transitfeeds.com/feeds


- Nine counties

- 7,000m^2

- 7.75 million people

- 101 municipal governments

Shocking that we only have 31+ transit agencies. The place is dense!


The number of goverments is the only reason though. London also only has one transit agency with a similar population (and maybe size).


It's a lot larger than 7,000m^2 :)


ha, the metric system strikes me again!


This is exactly the problem. It's so inefficient. 31+ executives (and top-paid people), 31+ budgets, 31+ locally-optimized goals and preferences. It's completely nuts.


The Bay Area (once a generation of boomers passes) will HAVE to become more like NYC - with "cities" (using that term loosely for places like Cupertino, Mountain View etc as these are really sterile suburbs with trip malls) becoming more like "boroughs" with an overarching authority.

We're talking a 30 year timeline, though.


It should be noted that San Francisco is smaller than San Jose by quite a bit, and all the cities in the south are suburbs of San Jose.

If that magic combination ever happened, it would be San Jose as the main city. And somehow I don't think San Francisco would be happy being a borough of San Jose.


You should read "Cities without Suburbs" if you want to know the politics around consolidation and how hard it is in most of the states.

A lot of it boils down into tiny municipalities being able to avoid a cripple larger metros that the littler metros hem in.


Agreed. They need to merge and coördinate agencies. Someone like abag could become the MTA or the ny-nj port authority.


Good point. And I would like to also point out that this is another example of the Valley fixing its own problems instead of problems we all share. All we ever hear outside of the Valley is how the Valley is going to save us.

And yes, there is a lot of troll in that comment :P.


The article didn't explain what a GTFS feed was, and my first thought was it's a "Get The Fuck Somewhere feed". Turns out it's General Transit Feed Specification.


Fun fact: GTFS used to be Google Transit Feed Specification:

https://sf.streetsblog.org/2010/01/05/how-google-and-portlan...


It has all the bus stops and times for applications to use. Metros usually generate them to send to google for their app but they sometimes also publish them online for people to use.


I like yours better


This partnership came about through a streamlined procurement mechanism called challenge-based procurement. Instead of months to issue an RFP it takes days and for bidders only hours to respond. If you want to learn more about how we’re helping cities work with innovative startups https://cityinnovate.com (we’re hiring)


I just tried to sign up and got directed here: https://511.org/open-data/transit

It looks like the bulk data feeds are available but the APIs are not. I am particularly interested in:

    Real-time Vehicle Monitoring (SIRI)
    Real-time Vehicle Monitoring API provides information about current location and expected activities of a vehicle in XML and JSON formats.

    Endpoint: http://api.511.org/transit/TBD
    Allowable parameters: api_key (mandatory) and TBD
Also, looks like there's an error in "GTFS-Realtime Vehicle Positions" description - it claims:

    GTFS-Realtime Trip Updates provides real-time updates on the vehicle positions for an agency in the Protocol Buffer format.


I note that OpenStreetMap apparently has tags for mapping between OSM nodes and GTFS. I wonder how much GTFS data gets imported into OSM.

https://wiki.openstreetmap.org/wiki/General_Transit_Feed_Spe...


No ACE train?


Sharp eyes!

ACE is included in the feed, but we did not include it in the animation. We don't currently have geometries for the route alignment (the train tracks), so we'd be drawing straight lines between the ACE stops — didn't look that great in the GIF, so we left it out.

[Note: I'm an Interline staffer]


If you want a quite fast and nice solution for aligning routes, pfaedle[1] is a great tool that creates a gtfs shapes.txt by using open street map train track / street data.

[1]: https://github.com/ad-freiburg/pfaedle


Nice. Thanks for sharing.

For what it's worth, this would bring ODbL-licensed data into an output GTFS feed.



Send an intern with a logging GPS?


Curious why the line to Sacramento (upper right) is considered the Bay Area.


The Capitol Corridor runs through the East Bay to San Jose, so it's used by a number of bay area commuters.


I think because it's an everyday commuter line. Unfortunately there are really limited options but I do know a few people that make the commute.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: