Do you suppose we'll eventually get access to all of those great station maps we can now only see one at a time, and only as part of the local neighborhood maps that are outside the turnstiles in subway stations? It's as though they're trying to get the least possible use out of them.

Although I'm generally in favour of more access to information, having seen the utter bollocks that most institutions make out of providing access to "anonymized" data, I find myself twitching madly about the idea of releasing something that might contain PII.

That said, there's obviously some data that doesn't carry that risk.

As an argument against ever engaging in public/private infrastructure agreements, it's hard to beat this one: Ad agency blocks free iPhone app used to tell the user where the nearest municipal bike stand with bikes/spaces is. I guess we can anticipate the nickel'n'diming to start in a few months when the agency releases their own, for-pay, proprietary version (that'll only work on IE, of course).

MTA intellectual property lawyers

Well this is obviously the problem. In my (painful) experience¹, having corporate lawyers make decisions about intellectual property is about as productive as having sharks decide on how to deal with endangered species.

1. The lawyers of the corporation I worked for decided that I, as a researcher in the corporate lab, couldn't be allowed internal use (no product intended) of technology that had been developed in my own lab.

Huh. The Washington area transit folks are apparently willing to make information available; I don't know about specific cellphone aps, since I don't use 'em, but they definitely have "next bus" data out there.

There's a trip planner on the WMATA website that includes not just the Metrorail and Metrobus systems, but the adjuncts like Montgomery County's Ride-On buses.

I'm surprised New York's being perverse on this.

Maddeningly, those neighborhood maps were available on the MTA's website as PDFs several years ago, circa 2001. They didn't cover the complete system, but they were genuinely useful for the fraction of the city mapped.

They weren't entirely up-to-date even then, with revision dates primarily back in the late 1990s – I don't know if they were removed for that reason, since the timing also suggests it could've been a post-9/11 "security" excuse.

It's recently become available in Montreal -- the STM website became about a zillion times easier to use once they gave an option of using Google Maps instead of their own crazy system. And it works on iPhones.

Sounds like the Iron Law of Institutions has taken hold of the MTA's legal department.

In the UK, we get a lot of this shit. Among other things, the complete list of postcodes and the locations they map to is secret - or rather, copyrighted and extremely expensive. (Somebody posted it to wikileaks the other day.) And the OpenStreetMap guys discovered that the administrative boundaries are copyright, as well; so they've got a project going to walk around London ground-truthing the borough boundaries by observing the signage etc.

We systematically underfund mass transit systems and beat up on them every time they forgo some revenue opportunity. So why should we expect them to forgo this one, even for a company as sacrosanct as Google? And why should they spend some undetermined number of IT hours to massage their internal data into the formats that Google or some other party requires? To benefit that horrifically underserved slice of the public that carries iphones?

In Washington DC, as I recall, a big part of the kerfuffle was that the Metro folks could point to something like $300K in revenues from people accessing their (partly ad-supported) web site for the information.

Teresa @ 1, I've often wished for a New York version of the UK's A to Z books.

It's a bit better now that I have Google maps on my Crackberry, but that requires a signal.

In reference to Alex #9 - it seems the gvt wants to privatise the ordnance survey as well. I believe it makes a profit, but the gvt would rather get rid of a common good in order to get a one off hit of money, rather than utilise it for the maximum benefit of the inhabitants of the UK.

And why should they spend some undetermined number of IT hours to massage their internal data into the formats that Google or some other party requires?

No, just say *what* format it is and point it into /public_html/.

paul @ 10:
o benefit that horrifically underserved slice of the public that carries iphones?

Most of the apps that work with the Portland system are not iPhone apps at all; some of them are schedule viewers or bus arrival alerts that use text messaging, so are usable on any cell phone.

Why do it? Because anything that makes it easier to use mass transit has the potential to increase ridership¹, thus increasing revenue.

1. I hate that word, but my wife's cousin, who's a consultant for mass transit systems, tells me it's not going away, so I should get used to it.

Dave Weingart @ 11: You mean there isn't an equivalent of the A-Z books? For New York? For any US cities? How do people find their way around???!!!

And wouldn't easy access to data mean more subway and bus riders, which would enhance their revenue? Essentially they're wanting to charge people for advertising, or maybe access to the user manual.

I bet a substantial number of people who fail to find the data they need give up and either walk or take a taxi, especially if they're in a rush or are tourists, thus cutting down on said revenue ...

dcb@15, I don't know about New York, but Los Angeles and other California cities have the Thomas Guide books -- spiral-bound books, a little larger than a standard sheet of paper, with detailed maps (every street is shown) of the entire metro area. It looks like these may be the same sort of thing, and they're an absolute lifesaver -- they'd be a bit bulky for carrying around for cycling but nobody can get around without a car here anyway.

#17
Well, most of the streets ... and it's not nearly as up to date as they claim. I want the ones for year-after-next, because where I work, that's the stuff we're looking at now.

I heard an item on CBC radio yesterday -- for the next month or so, you can listen to it here -- about the Toronto transit system running at a deficit. And because the system is partially funded by government subsidies which aren't based on ridership, as the ridership has been increasing, the deficit has been increasing because of the need for more vehicles, drivers, etc.

The system's managers are in a bit of a bind. Yes, it's great that more and more people are using public transit, but if that makes the entire system financially unfeasible, the whole thing falls apart. They can only increase fares so much. They're hoping for more government funding.

Dave @11, lorax @17, That'd be like our street directories, Gregorys, UBD & the 'ways series (Melway, Sydway, Brisway, Ausway, etc). Great historical documents, too, charting growth & changes. Until GPS mapping (based on data from these – they sell e-versions now) how else would you navigate a city?

Cold War memories: Reds are nasty, see, they don't publish reliable city maps.

Epacris@20, Lorax@17, dcb@15:

There are US street directories, but the Thomas Guide does not compare to the wonder that is Melway (they look even better on paper, where the colours are less saturated and the resolution is higher). Also, Melway was (perhaps still is) sufficiently ubiquitous in Melbourne that a page and grid cell was a common way of giving directions, like a pre-internet hyperlink.

I do wonder if the relatively easy availability of basic street data from the Census Bureau in the US discourages publishers from investing in really good maps.

Novalis, who I think comments here on occasion, probably has a few choice words on how the MTA chooses to share their data.

I remember a hilarious half-hour in a taxicab in Beverly Hills when the poor driver (yes, the stereotypical new immigrant) couldn't figure out how to get to the hotel at the street address I provided, because it was across Wilshire Blvd. and on a T-intersection one-way the wrong way. Or something like that. He finally turned the meter off, pulled off to the side of the road, stared at the Thomas Guide, and said "Sir, I can't get there."

All I had was a carry-on overnighter, so I paid him and walked across the street and checked in.

Josh Jasper, I am the impetus for that Streetsblog post. I'm also one of the cofounders of New York Transit Data, a working group trying to help open up the MTA schedule data. We've had a couple of great meet-ups so far, and I think there's some real momentum in the community about the idea of better and more open transit data. I would love to see some of Making Light's readers at the next meetup.

I can also give a better link for "bus routes down Broadway".

Re #10, there is no need for the MTA to do extra work to massage the data into GTFS (the standard format) -- I have done it already for bus schedules, and should complete subway schedules in the next week or so. My code is released as free software. Also, the MTA apparently uses a proprietary package called HASTUS for at least NYCT bus schedules, and HASTUS can supposedly generate GTFS natively.

I'm going to partially disagree here, and then thank Novalis.

The MTA wants a share from programs and services charged for. Yes, the MTA should absolutely 100% provide, as far as it is able, up-to-the-minute routing and service information for free to the public, including free re-packagers and re-providers. On the other hand, I see no problem with the MTA insisting on a fair share of revenue made directly from its information. (This is, by the way, the structure I also advocate for legal information and reporting. I have no problem being charged for the value-add of Lexis or Westlaw, but I'm thrilled that there are, more and more, appropriately free sources for directly pulling a cite).

And thank you, Novalis -- whatever happens to the for-pay sources, I know that the free-track sources will continue to grow and improve.

BSD@25: This means that you agree that MTA effectively cutting off access to that data by refusing to negotiate in good faith is a problem?

Starting with the link Patrick provides, it's not hard to find at least one case where the MTA refused to negotiate pricing for access then accused developers, perfectly willing to pay for this information, of working in bad faith. In the case of StationStops, the MTA claimed that the disclaimer "Not Affiliated with the MTA" implied "an initial and ongoing relationship with the MTA" and claimed that "There is no such language!" that could actually disclaim such a relationship. (The blog at www.stationstops.com is interesting reading.)

If anything this seems to be a case where a developer is perfectly willing to give the MTA money, but the MTA won't take it. It doesn't seem fair to make the argument that the MTA deserves "a fair share of revenue made directly from its information" without also pointing out how they have approached doing this.

In Minneapolis it was Hudson maps for years (local company; their office is on Nicollet Ave.). Then King came in from outside and started doing map books of the area, and suddenly the Hudson maps sprouted color too :-). I've kept a Hudson map in my car since 1977, but I don't update very often, I think this is my fourth. And these days I use Google maps more.

In the Washington area the map books are published by a company called ADC. They're good.

Most cities have a local company that does the equivalent. In every market that I've been involved in, realtors are required to put those map coordinates into the MLS system (and, in MSP, whether you got them from King or Hudson) as well as directions. It's a bit outdated, with many agents now having GPS in their cars or on their phones, but invaluable if the starting point of the directions is unclear, or if you're not sure of where that house is relative to this house (is it north and a little east of me or north and a little west? And isn't there a back entrance to this subdivision 2 miles closer?). I still keep my city/county Wunnenberg in the seat pocket and update it about every 3 years. It being paper, I happily pay for the privilege of doing so.

We don't have a robust enough transit system to warrant actual updates to transit maps. Unfortunately.

In North American cities outside Boston, while street maps are occasionally useful, they're not a total necessity the way an A-Z is in London, because of the grid system. This is hard for Europeans to understand, but in a North American city three lefts make a right -- you can literally go around a block, any block. And if you are a couple of miles north, or east, of where you normally are, half the street names are the same and you know what street will be coming next, because the grid is big and extends. Also, because the streets stay the same, numbering gets high but it changes by block, so if you have a street address you can tell how far up that street it will be. So it's really easy to find places, and you only need a map for the odd exceptions, not every single time you go anywhere new, as in Europe.

There's a bit in Delany's Triton where Bron is on Earth and he keeps saying "Don't think in terms of urban units, there aren't any!" which I thought was science fiction until I realised it was the way Americans are with blocks in Europe.

Jo @ 30
Well, three right turns to go around the block mostly works.
Los Angeles has a lot of intersections with no protected left turns. It also has areas where streets have gaps due to freeways, flood control channels, or railroads, or streets at strange angles, or otherwise requiring some fairly roundabout routes to go around the block. (Usually it's a matter of an extra mile or so. Sometimes a U-turn is much shorter and faster.)

I get a laugh, some days, as I come home from wherever along a street which has a gap of about a block, in between two major streets: someone will come along that section, expecting it to go through, and run right into the dead end, with no idea how to get out except to go back to the last major intersection.

Jo @ 30: Maybe off the east coast, or more to the point, in cities where the urban core underwent serious re/development within the 20th Century. The east coast cities with which I am familiar, with the signal exception of thoroughly rebuilt and overbuilt Manhattan, are European-style hub and spoke waterfront layouts, the hub typically being the original wharf/harbor/transportation center of vital importance.

What gets really entertaining is when there's a grid slapped right on top of the hub and spokes. Intersections take on fascinating shapes that way.

Philadelphia's center city was designed as a grid, but with spokes leading out from it. The spokes are still there, for the most part, and in the northeast and northwest parts of the city, the grid bends to follow the spokes (which then form the backbones of localized grids).

This can get a bit confusing when the grids have bent enough from the origin. In much of the northwest, the streets marked "east-west", by grid reckoning, are really more north-south by geographic reckoning, and vice versa. It took me a while to get used to this.

Pittsburgh's topography in much of the city is far too three-dimensional for grids to work, or for there to be any guarantee that three lefts will get you anywhere near where a right would take you. After a few lefts, you may well cross under your original street, on a bridge hundreds of feet above you.

Mark #32:

West coast cities other than LA also have problems. San Francisco has at least two grids, at roughly right angles; they sort of meet around Market Street. In Seattle, it's much much worse, as there are at least three grids colliding downtown, and there appears to be no such thing as a legal left turn anywhere along those fracture lines. Add in the fact that street names (usually numbers) repeat in different sectors of the city and that the only way you can tell them apart is by the sector designator (NW, N, etc.) and you have a recipe for permanent lostness, particularly if you are one of those poeple whose sense of direction gets ruined by proximity to water. I've spent at least a couple of months in Seattle over the last 30 years, and I still can't get a coherent overall picture of what I'm doing.

LA has lots of non-grid quirks (as P J mentioned). In the San Fernando Valley (where my folks lived, and I was for about a decade) street names are semi-persistent. ie Kinzie will be the the third street S of Lassen, for several miles (about 7, as I recall), but the longest continuous piece of it is about 1/2 mile. There are places where the even/odd sides of streets are in cul-de-sacs, on opposite sides of a cross street (my folks live in such a place. They are in the odd-numbered cul-de sac, go to the end of the block, and look 50 yards S, that is the even numbered cul-de-sac.)

A number of streets don't connect to the major cross streets, so you have to go in a block, take a minor street to connect, and then drive along. Odds are it will be a bit less than half a mile long (because that't where the major streets are), and most of the residential cross streets won't connect; you have to know which one's grant access.

Thos. Bros. is a really nice system, but their are (of course) a lot of introduced errors (as a means of copyright protection). I recall being late to a rehearsal because I was on foot, had check the map and tried to go in a manner impossible.

Having maps lie to one is always interesting.

There is a place in the north-eastern part of Halifax, NS, where, for some reason beknownst only to city planners, two streets meet a third in a rather peculiar tangle. The map we were using indicated that all three streets actually intersected at some point, and one could turn off one on to the other.

Naturally, we were a little surprised to find that the street we were on dead-ended in the parking lot of a doughnut shop. Fortunately, there was an exit from the parking lot onto the street we wanted, but I still wonder to this day whether this unexpected result was due to a deliberately introduced error in the map, a real mistake on the part of the surveyors, or the propensity of Tim Hortons doughnut shops to sprout up unexpectedly, like autumn mushrooms, in most large Canadian cities.

Terry @ 35
There were some interesting quirks that were used to great effect in skill-gimmick rallies. Like the one block of Dorrington stuck literally between two blocks of Gullo.

Atlanta's streets are not even remotely grid-like, either. Actual straight streets are a distinct minority; I can name several trios where all three streets cross each other at right angles. Then there are the repeated street names. And the similar street names (far too many variations of Peachtree). And the streets that change names several times. A good map is a vital necessity in Atlanta, and map-reading a skill that one learns early.

eliddell, #36, some maps purposely put in streets that don't exist to make sure other people aren't stealing their data. We have two streets that come about four feet from each other horizontally, and about eight feet from each other vertically, and you have to go to the highest definition to see that they don't connect on the online maps.

The key difference isn't that the streets are laid out at right angles--they aren't all, even in Manhattan--it's that the names don't keep changing. Which is why I find Mass Ave. in and around Boston disconcerting: the name is continuous, but the numbers change as you cross from one city or town to the next, so it's easy to think you're almost at your destination, and then wonder where it went as the house numbers continue to increase (or decrease), and then suddenly you're in Somerville and there's a discontinuity, and another when you get into Arlington....

Even in Manhattan, it's not all grid and right angles; at the north end it goes 204, 207, 211, 214th [a block of staircase], 215 [part of which is a staircase], 216, 217, 218: but 216th is a single block, and 217th is a single block, and they don't intersect the same streets. We manage, though people working only with printed road maps periodically run into problems when they find that 215th Street doesn't go through if you're using wheels.

But I can't blame the city for Rand-McNally's unwillingness to indicate staircases as distinct from roads suited to things with wheels.

Portland consists of several grids, jammed together at the edges, plus a few places where changes in elevation result in straight lines not being geodesics as you'd expect on a plane. In fact, I think some knowledge of differential geometry might help a lot in getting around here: a map of Portland is really an atlas of a differentiable manifold; each grid is one Euclidean chart of the atlas, and the weird places in the hills and on our pet volcano are some rather warped spaces that are tacked on.

But the nasty problem with Portland streets is that even in the grids, they don't always continue. I used to live on SW California St., of which there were four segments: 5 blocks between 55th and 60th, 3 blocks between Capitol Highway and 27th, and unconnected 1 ½ (really just a couple of cul-de-sacs), and 4 block segments on the hill between I-5 and the Willamnette River. There is no simple route that connects either of the first two segments with the latter two, though you'd think it would be easy to get from one part of a street to another. The streets parallel to California (all named after states) do the same thing in different places.

#40
Rand McNally also has trouble recognizing that railroads exist and are worth noticing, at least on many of their maps.

17 ::: lorax @ 17; Lawrence @ 28, sisuile @ 29: Thanks for this - I was getting worried about a lack of city maps.

Having read the last few posts I'm now getting worried about what the maps don't show (or do show that ain't so). "Roads" that are flights of stairs are a read problem if you're not in a stunt car.

I do remember having problems back in Cambridge (UK) when a driver asked me the way somewhere. The only routes I, as a cyclist, knew, involved going between bollards - not possible in a car. I'm afraid I was in the classic position of saying "well, if you're trying to get there, I wouldn't start off from here..."

All true, but think of the most confusing bit of your city for which strangers really need a map, and consider that that's what every square foot of London is like. Which is why everyone who lives in London tends to have an A-Z in their bag all the time, and why London taxi drivers have to pass a very hard exam where in North American cities taxi driving is something new immigrants who can drive can do in their first week.

Re John Mark Ockerbloom@33 and Pittsburgh: there are locations downtown for which the appropriate driving instructions are "Take the Parkway west three exits and then take a down."

I've also found locations (in Squirrel Hill) where the traditional "go around the block" methodology -- turn right at every intersection until you return to your starting point -- produces startling results. Such as a multi-mile cloverleaf pattern that crosses over the highway twice, crosses under and over itself, and halfway through the "block" you pass your starting point on the other side of the street.

Marilee@39: I know--I just like the mushroom-doughnut-shop explanation better.

Andrew, 45: "Three-Cornered and Secure," yes.

Vicki #40 :
Which is why I find Mass Ave. in and around Boston disconcerting: the name is continuous, but the numbers change as you cross from one city or town to the next, so it's easy to think you're almost at your destination, and then wonder where it went as the house numbers continue to increase (or decrease), and then suddenly you're in Somerville and there's a discontinuity, and another when you get into Arlington....

Especially since Mass[achusetts] Ave doesn't actually enter Somerville ;)= (it is less that half a block away at some points (and looks like the border was set based on that, there is an otherwise inexplicable bulge in the Cambridge border)).

Marilee #39:

eliddell, #36, some maps purposely put in streets that don't exist to make sure other people aren't stealing their data. We have two streets that come about four feet from each other horizontally, and about eight feet from each other vertically, and you have to go to the highest definition to see that they don't connect on the online maps.

Since map data is a compilation of public data without unique properties or organization*, the data is completely public domain. Except of course, the pseudo-streets which are proprietary. The only copyright violation should therefore only be for that street(s). I suspect the map companies claim that that makes all of their map data proprietary.

*At least from the point-of-view of making your map from their map. Certainly the image is under valid copyright, as well as all the behind-the-scenes stuff on a computer-based map.

TexAnne, #47: *giggle*

Andrew Plotkin (45): My siblings and I once figured out--with the help of a map--that the "block" we grew up on (in suburban Atlanta) was not only multi-multi-sided, but about five miles around--longer if you went in and out of all the dead ends to avoid having to cross the street. So taking a walk around our block would have been a major undertaking. The "block" across the street, however, was quite small, and easily walk-around-able.

#45: Ah, you've been to my old neighborhood!

On my first visit to Pitt, to apartment hunt and drop off paperwork at CMU, I used a campus map that had North on the bottom. I figured things out quickly enough, but getting back in the direction I wanted meant going around the block. I chose Craig. My three right turns put me on the road through Panther Hollow, with Forbes Ave. on a bridge high overhead.

I was still finding new neighborhoods within a few minutes of my apartment when I moved out 18 months later.

Mark @32: The east coast cities with which I am familiar [..] are European-style hub and spoke waterfront layouts, the hub typically being the original wharf/harbor/transportation center of vital importance.

Rochester NY has a hub and spoke layout, although the center of the city is not its lake port, but its river waterfall; the location was selected to use the waterfall for a mill. Not an east coast city though; at one time it was considered the western frontier. The addition of the Erie Canal provided the means to transport the mill's production. One of the notable features of the city was an aqueduct carrying the canal over the Genesee River.

Later the canal was relocated, and the original canal bed was repurposed for a subway (and even later, parts of the subway system were incorporated into expressways currently in use). Although defunct for years, the subway still has fans (here is a might-have-been map as if it had never been shut down).

Here are some pictures of the aqueduct as it looks today, next to a library built on the foundations of an old mill race. Currently, there are proposals to restore the aqueduct back into a canal park!

Pardon the links... I am not employed by the local government...

This is hard for Europeans to understand, but in a North American city three lefts make a right -- you can literally go around a block, any block.

You've never been to Pittsburgh, have you? We have way too many hills for that, and are victims of a pre-car street layout.

Amusing myself with web toys!

The cloverleaf trip I mentioned at #45 is now plotted on Google Maps: http://maps.google.com/maps/ms?ie=UTF8&hl=en&msa=0&msid=101129065083838639269.000474a7a931ff8796468&z=16

The one Stefan started at #51 is: http://maps.google.com/maps/ms?ie=UTF8&hl=en&msa=0&msid=101129065083838639269.000474a4b422a3f57a3b6&z=15