Canada Post has problems… because it’s run by boneheads

So Canada Post is in the news cycle again – specifically, concerns that they’re phasing out door-to-door delivery of mail. That’s tragic, sure, but before you shed tears for Canada Post and blame the Internet for its current woes, let me give you a peek at the side of Canada Post that you don’t often hear about in the major news outlets. It may put the corporation’s current troubles in an entirely new light.

If you live in Canada, your mailing address has a 6-character code attached to the bottom of it: the “Postal Code”. It was introduced in Canada between 1972 and 1974 as a way to make the postal system more efficient, and – after some initial pushback by the postal worker unions – was a roaring success. Now the postal code is a near-required part of any address – any form you’ve ever filled out that requires your address has probably had a space for it. It is also used in other ways. For example, it is occasionally used by researchers doing studies to organize data – employment, poverty, health – by location. It is also used by multiple businesses to help you locate a store.

The code is made up of 3 capital letters and 3 digits in alternating order, starting with a capital letter, and is usually written with a space between the two groups of three. All ten digits are used, but the letters ‘D’, ‘F’, ‘I’, ‘O’, ‘Q’, and ‘U’ are not, because they are too difficult to distinguish from other characters with optical character recognition techniques.

  • The first character is the “postal district”. There are currently 18 postal districts (out of a possible 20 – ‘W’ and ‘Z’ are not assigned), assigned roughly east-to-west first across the provinces then the territories. Most provinces and territories get one each – except Nunavut and Northwest Territories share ‘X’, for historical reasons – but Ontario gets five and Québec gets three because they have such large populations.
  • The first three characters collectively (before the space) are the “forward sortation area” (FSA). The first character is the postal district described above. The second is zero if the area is a large rural area, and non-zero otherwise. And the final character identifies the rural area, or an entire city, or – for large cities – a part of a city. So, for example, the city of Milton is covered by L9T, while the nearby rural areas (including Bolton, Campbellville, Moffat, etc.) are all in L0P. Burlington is covered by L7L (Burlington northwest) through L7T (Burlington southwest).
  • The final three characters (after the space) are the “local delivery unit” (LDU). Some of these have special meaning (9Z9 means business reply mail), but most simply identify some subsection of a FSA. For example, in the L0P rural area, Bolton, Campbellville, and Moffat are L0P 1A0, L0P 1B0, and L0P 1J0 respectively. FSA L8S contains McMaster University in Hamilton, and L8S 4L7 is specifically the John Hodgins Engineering (JHE) building.

Of course, there are a few special or reserved postal codes, most famously the one reserved for letters to Santa Claus: H0H 0H0 (from “Ho, ho, ho”).

Functionally, the area covered by each postal code can be described as a polygon – large polygons for the postal districts made up of smaller polygons for the FSAs, which are in turn made up of smaller polygons for each LDU. If you can figure out the coordinates for the polygons, you can draw a map of the postal codes, or be able to figure out a postal code given an address or GPS location. (For example, here is the polygon for my postal code – it’s a little hard to see, but it’s that red line, and it’s not quite right. Here is another example with a rural postal code – zoom out to see the polygon.) Figuring out the polygons is simply a matter of trial and error – all you need to do is get enough people to tell you their address (or GPS location) and postal code, and you can spot the boundaries between the postal code areas. The more data you get, the better your estimates of the polygons will be.

And that is exactly what a small company named Geolytica started doing in 2004, albeit in a particularly clever way.

Basically, (I’ll show exactly how they did it in a moment), they used the addresses submitted by visitors to the site to submit data, or correct errors, to construct a geographical database of postal codes (among the other things the site did). In addition, their entire postal code geolocation database was released under the Open Database Licence (ODbL), which is effectively the same as CCBYSA. That means the data they collected could be used in other open projects, like OpenStreetMap.

What could be wrong with any of that, right?

Well, as we all know, Canada Post has had some tough times in the last few years, posting loss after loss. They have had to cut back on their operations significantly (culminating, now, in the phasing-out of door-to-door delivery completely). They did make some efforts to modernize and offer new services – for example, in 2010 they set up a comparison-shopping site that allowed Canadians to compare prices on items from over 500 stores. Didn’t work out, though – the comparison shopping program was shut down after two years. No doubt a lot of their woes have been due to being functionally ignored and under-funded. But I would like to suggest another hypothesis – that the reason Canada Post finds itself in such dire straits right now has less to do with the advance of technology, and a lot more to do with their ripping incompetence. And my evidence of that?

Canada Post sued Geolytica over their open-source postal code database.

In their original lawsuit, they contended that postal codes were the private, intellectual property of Canada Post. They were suing Geolytica for copyright violation.


Oh, it gets better!

What Canada Post actually accused Geolytica of is appropriating Canada Post’s postal code database. In other words, they claimed that Geolytica somehow got a copy of Canada Post’s postal code database, and then started sharing it with the world. Hence, copyright violation. But what actually happened – as explained in Geolytica’s response to Canada Post’s legal threat, is quite different… and far more interesting.

The way it worked was simple. You entered an address – say “426 Brant Street, Burlington, ON L7R 3Z6” (which is Burlington city hall) – and Geocoder would use that address to query several open databases, like OpenStreetMap, and give you the compiled result data in some way, including maps and other useful info. However! At the same time, it would strip the postal code out, and store it in Geocoder’s private database, along with the street address and other information it got back from its open data set queries. So now it knows that the postal code “L7R 3Z6” contains the street address “426 Brant St.” in Burlington, Ontario, and the GPS coordinate {43.326003,-79.798585}. With enough searches (it took years for Geocoder, and actually involved aggregating data from several sites), your data gets better and better. Eventually, for example, you can discover that the border between postal codes L7R 2G4 and L7R 2G7 lies between 502 Brant (43.327642,-79.800602) and 504 Brant (43.327676,-79.800652). But even partial data is helpful, if imperfect (for example, as I mentioned, the information for my address is wrong – it has me practically in the park, and I can’t even see the park from here).

And while the database itself could be covered under copyright – it would be a work – the data contained within (postal codes) is not. They are simply facts. It is a fact, not the result of a creative decision, that the postal code for 504 Brant St., Burlington is L7R 2G7, and a fact that the postal code for 508 Brant is also L7R 2G7. From those facts one can deduce – which, again, is not a creative decision – that 506 Brant has postal code L7R 2G7. I did not look that up. I did not consult Canada Post to get that information. They cannot say they have copyright on that information.

Now, up to this point everything is still somewhat sane. Canada Post would just be guilty of corporate stupidity, and of filing a lawsuit without bothering to verify the facts first. Stupid, and evil, but pretty run of the mill stupid and evil for a corporation. They got some negative publicity, they got laughed at, time to go back to the boardroom and lick your wounds. Maybe, you know, find a way to stop hemorrhaging money.

Instead, Canada Post decided to double down.

Oh, they still claim stole their database – only now they’ve named founder Ervin Ruci personally in the suit, and they’ve also added a bunch of his other sites (along with several that have similar names but that he has no connection with, and some that don’t even exist – great investigative work there, Canada Post assholes). They’re also digging in on the argument that postal codes are creative works, not facts. But they’ve added a new twist:

They’re claiming copyright on the words “postal code”. Seriously. You have to see this – just look at how they formatted one of their cease and desist letters: every single instance of the words “postal code” is marked up with “OM. (What the fuck does “OM” even mean anyway? Did Canada Post just invent something other than a trademark or service mark?)

Note, too, that they’re not going after the big name violators. Above I posted links to several major corporations who are shamelessly using Canada Post’s copyrighted words “postal code” on their sites! Think Walmart’s getting legal threats from Canada Post? Don’t bet on it. Ruci and Geolytica were soft targets – they weren’t supposed to fight back. This is copyright law abuse, pure and simple.

And let me put things further in perspective. All across the world right now, there are initiatives underway to open government databases like this, making access to public information easier for private citizens to use as they will. For example, the UK released their postcode database – complete with geolocation data – under their Open Government Licence (OGL) in 2010 (before the first Canada Post lawsuit!).

But it’s not actually that Canada is behind the curve here. Because the rest of the government is embracing open data. Indeed, Canada Post is the only crown corporation fighting the trend! I have personally downloaded complete data sets from the 2011 census from StatsCan for my own activism-related research – for free, as it should be; my taxes already paid for the census, so I deserve the data.

In fact – and if nothing else, this is the key piece of evidence of the stupidity of Canada Post, and the emptiness of their legal claims – part of the postal code database is already open data! Yeah, really! Brace yourself, because this is a double whammy – first to the claim by Canada Post that postal codes are their own intellectual property used by them only for their mail delivery services, and second to the claim that the data is theirs to sell and unavailable otherwise. You see, it turns out that Statistics Canada used postal codes for the census! And! Because StatsCan doesn’t have their head up their asses like Canada Post does, when they made the census open data, they also made relevant metadata open data… include the FSA part of the postal code database (the first three characters). You’re probably thinking, “no way!”. Yes way. Fourth option from the bottom. And note the sign of Canada Post’s stupidity: it’s the only data set with a copyright notice.

In other words, what Canada Post is wasting thousands and thousands of dollars – that we are giving it to support its failing business model – on… is harassing small businesses who have spent years building their own database just because it includes the LDU (the last three characters of the postal code).

And not only that… but to spite their own customers, and “protect” this data that can’t be protected anyway (because, as Ruci proved, it can be deduced independently)… Canada Post has changed the policies of their site’s postal code look-up utility.

This is what Canada Post has been doing for the last few years while their business went tits up. Not developing new technologies and business models to keep up with a changing world. Nope. They’ve been abusing copyright law to harass small business operators who use open data to independently generate and share stuff that looks somewhat like stuff Canada Post wants to sell.

So no, I, for one, am not particularly surprised at the mess Canada Post has ended up in. And while it pains me to see a necessary service fall to shit like this, I will not weep for this corporation. They need to be put out of their misery. Perhaps they need to be replaced by a new crown corporation that can provide the important services Canada Post provides, and more – especially stuff for the information age.

CC BY-SA 4.0
Canada Post has problems… because it’s run by boneheads by Indi in the Wired is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

One Response to Canada Post has problems… because it’s run by boneheads

  1. I have been looking everywhere for further information on this lawsuit. Latest news items date back to 2013 and even cippic, who represents Geolytica, has no current information on their website. Do you have any idea of the current status of these legal proceedings? I find it weird that nothing is mentioned after the initial notice was filed. Regards,

Leave a Reply

%d bloggers like this: