Interesting that the researchers didn’t analyze W3W directly because it would violate their ToS, but instead constructed their own word list. I would think that this sort of criticism of a technology used by the government (for emergency services) would be acceptable, similar to “fair use” in the US.
Edit: it looks like the list itself is not even available, they would have had to reverse-engineer the app, so using a synthetic list makes more sense.
There's a reverse engineering of What3Words from a few years ago floating around out there called "What Free Words". It includes what claims to be their word list. I imagine the researcher is familiar with this but has chosen not to include it in their research.
(W3W is very aggressive about pursuing possibly spurious IP claims to remove any trace of What Free Words. In the past, even mentioning it in a comment like this is enough to attract a DMCA complaint.)
It's interesting because I'm fairly sure that work has already been done. Even without criticising the geocoding itself, the fact that there are singlular/plural pairs in the word list along with easily-confusable homophones says a lot about the product.
Ah, you got me! I didn’t realize from reading the TechCrunch article that the repo contained the actual word list, I stand by my original opinion that the takedown was legally valid.
You can copyright lists of words if their selection or arrangement amounts to creative work. I mean, a poem is a list of words. A list of words that have been picked because they have a certain length or because, when combined with other lists, they produce a memorable, unique combination may (and I guess does in practice) amount to a copyrightable creation.
You can't copyright the GPS coordinates of cities, because those are facts. The data would be the same whether you compiled it or not, and it would be absurd to grant someone a monopoly on that data.
If you compile a word list yourself, you've created something that doesn't describe the factual state of the world; you've created something new.
FWIW, before the debacle, even the (apparent) author of it appears to have admitted that it flaunted copyright:
> FWIW, I have reverse-engineered the what3words address decoding algorithm. If anyone else with the luxury of flaunting patents and copyrights is curious, check out the demo at https://cardinalhood.github.io/what3words/. It’s written in JS, and runs in browsers and Node.js.
Whether or not a collection of facts can be subject to copyright varies across jurisdiction - in the US you can't copyright the phone book, while the EU has a specific concept of database rights. What's critical in the US case is whether there's any creative expression in the collection of facts or whether it's effectively mechanical, and whether the w3w wordlist would be protected under that would seem like an interesting case.
Well, threatened with legal action and wasn't up for fighting it. It didn't get as far as him being sued, and while I can't blame him in the slightest for that choice, I would have been absolutely fascinated if it did go to court. A compatible reimplementation of something that's being sold to emergency services being slapped down because it's compatible sounds like the sort of thing useful precedent could have come out of.
It's worth noting that some huge limitations/risks of W3W are there because of the flaws in their word list choice, so if you're analyzing a different wordlist, you explicitly aren't looking at a big source of problems.
Edit: it looks like the list itself is not even available, they would have had to reverse-engineer the app, so using a synthetic list makes more sense.