Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
The Courts Open Source

What3Words Sends Legal Threat To a Security Researcher For Sharing an Open-Source Alternative (techcrunch.com) 141

A U.K. company behind digital addressing system What3Words has sent a legal threat to a security researcher for offering to share an open-source software project with other researchers, which What3Words claims violate its copyright. From a report: Aaron Toponce, a systems administrator at XMission, received a letter on Thursday from London-based law firm JA Kemp representing What3Words, requesting that he delete tweets related to the open-source alternative, WhatFreeWords. The letter also demands that he disclose to the law firm the identity of the person or people with whom he had shared a copy of the software, agree that he would not make any further copies of the software and to delete any copies of the software he had in his possession. The letter gave him until May 7 to agree, after which What3Words would "waive any entitlement it may have to pursue related claims against you," a thinly-veiled threat of legal action. "This is not a battle worth fighting," he said in a tweet.

Toponce told TechCrunch that he has complied with the demands, fearing legal repercussions if he didn't. He has also asked the law firm twice for links to the tweets they want deleting but has not heard back. "Depending on the tweet, I may or may not comply. Depends on its content," he said. U.K.-based What3Words divides the entire world into three-meter squares and labels each with a unique three-word phrase. The idea is that sharing three words is easier to share on the phone in an emergency than having to find and read out their precise geographic coordinates. But security researcher Andrew Tierney recently discovered that What3Words would sometimes have two similarly-named squares less than a mile apart, potentially causing confusion about a person's true whereabouts. In a later write-up, Tierney said What3Words was not adequate for use in safety-critical cases.

This discussion has been archived. No new comments can be posted.

What3Words Sends Legal Threat To a Security Researcher For Sharing an Open-Source Alternative

Comments Filter:
  • by shaitand ( 626655 ) on Monday May 03, 2021 @10:43AM (#61341370) Journal
    Sorry to hear that he caved to these thugs.
    • Re:A shame (Score:5, Insightful)

      by hey! ( 33014 ) on Monday May 03, 2021 @11:00AM (#61341494) Homepage Journal

      This really sounds like a SLAPP suit. You can't copyright an idea, so unless this guy actually copied their software or database the idea of suing him for copyright violations is ludicrous. Anyone is free to implement the same idea as long as they build their implementation from scratch.

      The cost of defending yourself in a copyright suit can be millions, but even getting to summary judgment will cost a defendant tens of thousands of dollars and months of his life.

      • by EvilSS ( 557649 )

        You can't copyright an idea

        No, but you can certainly patent one: https://www.cbinsights.com/com... [cbinsights.com]

        • Re:A shame (Score:5, Informative)

          by hey! ( 33014 ) on Monday May 03, 2021 @11:25AM (#61341630) Homepage Journal

          Except they didn't claim patent infringement.

          Looking into this a bit more, it looks like what the researchers did was create a tool that enables researchers to poke around in the what 3 words database to find problems. It's possible that this research use of the patented idea doesn't have all the elements needed for a patent infringement claim, so they went with copyright -- possibly a EULA violation claim.

          • by EvilSS ( 557649 )

            Except they didn't claim patent infringement.

            Looking into this a bit more, it looks like what the researchers did was create a tool that enables researchers to poke around in the what 3 words database to find problems. It's possible that this research use of the patented idea doesn't have all the elements needed for a patent infringement claim, so they went with copyright -- possibly a EULA violation claim.

            I think they lawyers are abusing the DMCA using copyright claims to expedite getting it pulled down, knowing he probably can't afford to fight back and any site getting the DMCA will have to honor it unless it's disputed, and that there are zero fucking repercussions for doing so. I'd say he should get help and try to fight it, but those patents are pretty damn big guns they could pull out in a fight and unfortunately, they would probably win.

            • Yes and no. On paper you are right but patent trolls like this haven't been well received in the past couple years in the court systems. Software patents are fairly successful for licensing and defensive war chests but active battle in open court purely to stifle innovation hasn't gone over well. That all assumes a patent which nobody has shown exists.

              Other people mentioned trade secrets and trademark both of which are much more tricky beast to overcome but he can establish he didn't have access to trade se
              • by EvilSS ( 557649 )
                Yeah but the problem is the patents are legitimate currently. Fighting a patent lawsuit, which the company could bring and not get immediately tossed, cost a hell of a lot of money. And if they filed it in the US (those are US patents), under the American rule, he probably wouldn’t get his attorneys fees back even if he won.
          • Re:A shame (Score:5, Informative)

            by AmiMoJo ( 196126 ) on Monday May 03, 2021 @12:32PM (#61342260) Homepage Journal

            The code and demo are still up here:

            https://pballett.github.io/wha... [github.io]

            It's reverse engineered from their system. They don't have a database, they use an algo to compute the three words. A pretty simple algo at that.

            He then noticed that it doesn't work very well, sometimes generating very similar words for locations some distance apart. Not great for something billed as "life saving" for use in emergency situations.

            I'd add that the choice if words is poor when considering non-native speakers.

            • Thanks for the link! In case anyone was wondering what the word list was it is on line 16 of the html file:

              const wlist = { "00000":"lungs", "00001":"grind" , "00002":"pandora" , "00003":"biodegradable", ...

              Spaced manually added after the comma to avoid the shitty /. "Lameness filter encountered. Post aborted! Filter error: Looks like ascii art."

              There are other "gems" like this that have hard-coded passwords:

              // This is just a block cipher
              function f(L,R){
              x = parseInt(L);
              y = p

            • I also seems incapable of generating words for the dallas fort-worth area, the pin oddly sticks.
            • Re:A shame (Score:4, Informative)

              by hey! ( 33014 ) on Monday May 03, 2021 @02:22PM (#61342922) Homepage Journal

              Looking at the code, I wouldn't call this "reverse engineered". Rather, it "does something vaguely similar".

              What "What Three Words" claims to do is to divide the world up into "3 meter squares". If you know anything about GIS or cartography, you're probably aware that the world isn't flat. That means you can't divide it up into three meter squares by simply treating the latitude and longitude as if they were Cartesian coordinates.

              But that is precisely what the code in "WhatFreeWords" does: it simply takes slices of the *geographic coordinates* to identify its bounding lines. Doing it that way means the areas defined by "WhatFreeWords" are only square (ish) at the equator. The further away you get from the equator, the narrower they will be east to west. That may be just fine for many applications. Too much precision is not as big of a problem as too little for many uses.

              Now is is mathematically impossible to do what "What Three Words" claims to do: tile the world with three meter squares. That's because the Earth is a sphere and you can't *perfectly* tile it with squares without overlap or gaps. However, you can get an approximation that's good enough for practical purposes by dividing the world up into zones within you can treat coordinates like they were on a flat plane. It's not rocket science, but it's not as simple as taking slices of latitude and longitude text.

            • > He then noticed that it doesn't work very well, sometimes generating very similar words for locations some distance apart.

              Not quite ("some distance" means "a large distance", right?) - the problem is it generates similar words for locations that are close to each other.

              It's fine if a mis-heard address sends you to the other side of the planet - it will be immediately obvious that the address is wrong and the error can be fixed. But if a mis-heard address might send the emergency services to the other s

            • I've also wondered what the relevance of this thing is. To use it, I'd have to pull out my phone, skip past any mapping app that has "Share my location" built into it, go to some What3Words-enabled mapping app, get the three words, call someone, try and remember what the three words were... it's a totally stupid way to do something that you can, given that a smart phone is required, do much more simply using almost any other mechanism.
        • The key aspect of a patent is the disclosure. You can discuss and share information on patents pretty openly (for now). Filing with a patent office makes the information a matter of public record, as opposed to keeping it as a trade secret. The record is part of how you defend a patent.

          The limits and application of copyright law are pretty well understood by now, and it seems far fetched there could be any copyright violation by pointing people to a competing third-party implementation. DMCA chicanery aside

        • Not if the patent office is doing their job - ideas are specifically excluded from patentability. Only specific implementations of those ideas are eligible. Which is one of the reason software patents are so bad - generally speaking only the idea was ever documented, making it almost impossible for a competitor to come up with an alternative implementation that avoids the patent protection.

      • Yeah but for the reasons you already mentioned there never would have been a suit. It would be frivolous and they would end up having to pay the tab.

        You should be able to sue entities (including law firms) engaging in these kind of legal strategies. It really isn't any different than putting a gun to his head.
        • Re:A shame (Score:5, Informative)

          by Immerman ( 2627577 ) on Monday May 03, 2021 @11:25AM (#61341628)

          > It would be frivolous and they would end up having to pay the tab.

          In any sane loser-pays system, certainly. But Xmission is in Utah, and my understanding is that trying to reclaim legal fees in the US, even for frivolous lawsuits, is generally an ordeal. And you still have to be willing and able to pay up front to get your case to that point, with no guarantee sanity will actually prevail.

          I want to say Utah is (was?) a notoriously good place for copyright trolls as well - wasn't SCO vs IBM fought there because SCO went court-shopping?

          Basically, any interaction with the US court system is likely to be expensive and time consuming.

          • This British law firm was likely threatening to sue in the UK, not the USA.

            • Actually they weren't threatening to sue anywhere in particular. Just to "pursue claims" over unspecified "related issues" in an unspecified jurisdiction.

              Assuming so though, that's not necessarily an improvement, as it means international travel, dealing with another country's legal system, and having any existing legal counsel be utterly useless. All substantial difficulties and expenses before even setting foot anywhere near a courtroom.

          • "my understanding is that trying to reclaim legal fees in the US, even for frivolous lawsuits, is generally an ordeal"

            IANAL but my understanding it is a matter of the court ruling that way. In a case like this the court would be quite upset for having its time wasted.

            "And you still have to be willing and able to pay up front to get your case to that point, with no guarantee sanity will actually prevail."

            True enough.

            "Basically, any interaction with the US court system is likely to be expensive and time consu
            • Or any court system. No need to introduce anti-US bigotry especially since the US legal system is better than most.

              The US court system favors entities with money more than the court systems in many other countries do. Both in terms of costs of a court case and in the potential financial outcome when loosing.

              • "The US court system favors entities with money more than the court systems in many other countries do."

                Only in the sense that the defendant has to pay out-of-pocket. The court systems of most first world nations heavily favor 'industry practice' and 'government' and popular sentiment which is equally consistently in benefitting the wealthy and powerful over the individual. This is compounded by most of those other systems not even paying lip service to the avoiding the many trampling on the liberty of the
            • >IANAL but my understanding it is a matter of the court ruling that way.

              IANAL either, but my understanding is that while a court *may* rule that way, it's very unlikely to do so up front, even in cases of obvious abuse. Typically you have to pursue such restitution separately from the original case.

          • SCO was based in Utah, which is why it was here. I believe that East Texas is one of the notorious patent troll places that perhaps you were thinking of.

            • Oh yeah, East Texas is definitely bad, but neither patents nor Texas were relevant to the SCO-suit, so I don't think that's what I was thinking of.

              >SCO was based in Utah

              Was that not part of the shopping? Did Utah-based SCO do or create anything other than the Linux shakedown+lawsuit to say they were "based" anywhere for other-than-legal reasons? The original SCO obviously did, but that was when they were the Santa Cruz Operation. After selling the business to become new-SCO in Utah I thought they were

  • by hduff ( 570443 ) <hoytduff@nOSPAM.gmail.com> on Monday May 03, 2021 @10:44AM (#61341378) Homepage Journal

    Seems like the appropriate reply.

    • What? Just like that?

      • by applique ( 1526743 ) on Monday May 03, 2021 @11:02AM (#61341504)
        Wikipedia - In the case of Arkell v. Pressdram (1971), the plaintiff was the subject of an article.[68] Arkell's lawyers wrote a letter which concluded: "His attitude to damages will be governed by the nature of your reply." Private Eye responded: "We acknowledge your letter of 29th April referring to Mr J. Arkell. We note that Mr Arkell's attitude to damages will be governed by the nature of our reply and would therefore be grateful if you would inform us what his attitude to damages would be, were he to learn that the nature of our reply is as follows: fuck off."
  • Is the WhatFreeWords software available anywhere?

    • by nitehawk214 ( 222219 ) on Monday May 03, 2021 @10:53AM (#61341450)

      https://pballett.github.io/wha... [github.io]

      Come sue me, bitches.

      I have always found What3Words to be entirely worthless. The three words have nothing to do with the actual location, and alone give you no useful information. At least with longitude and latitude I can at least discern on what continent the location is on.

      • It's easier to remember three words than it is to remember a series of numbers. It's meant to be simple to use.

        • You mean something like TwoFiveSeven?

          • No. More like "Person, Woman, Man, Camera, TV"

          • Let's examine and compare.

            WhatThreeWords supposedly has a resolution of three square meters.

            A GPS coordinate would be something along the lines of Latitude 40.741895 Longitude -73.989308 (5th ave half way between 23rd and 24th street), if we shorten that down to 40 by -73 then the location is 70+ miles to the southeast in the New York/New Jersey bight.

            Or, did you mean having the three words be only numbers? Assuming only the land surface area of the earth (196.9 square meters), one would need 196,8
          • Think street addresses as the relevant comparison - especially for places that have no such formalized concept.

            You need 5 decimal places of GPS coordinates for one meter (individual doorway, aka chaotic "street address") resolution .

            73.92745 North by 132.53964 West
            is twenty words instead of three. Include a city/region as well, implicitly or explicitly, and you've also got extremely robust error detection.

        • by nagora ( 177841 ) on Monday May 03, 2021 @11:09AM (#61341544)

          It's easier to remember three words than it is to remember a series of numbers. It's meant to be simple to use.

          The problem is that it may be easier to remember three words than lat+long, but it's not easy to remember several sets of three unrelated words. So you end up having to look them up with your GPS, at which point you have the lat+long right there in front of you.

          Additionally, if comms is flaky to your would-be rescuers (say) then the three words can easily be misheard without the hearer having any real notion if the word they think they heard is unreasonable in the context because the words have no context. Digits have only a limited choice from zero to nine; if it sounded like "fish" then the listener knows to ask for a repeat.

          • This.

            It seems this is there to quickly and accurately convey a physical location from one person to another. If that communication is via a reliable method (like in writing or by POTS, then there's enough bandwidth to not have to use arbitrary three word phrases. If, on the other hand, the communications are marginal or almost non-existent, the lingo needs to be very unambiguous and clear. Lots of syllables make conveying those words difficult. The fact that it appears to be all in English is a serious fla

          • If comms are flaky, you end the voice call and send an SMS with your coordinates. This has a much higher chance of getting through when comms are bad.

          • As I understand it the primary purpose of WhatThreeWords (and competitors) is for human communication of local addresses in places where street addresses, etc. don't exist. The assumption is that you're using a GPS, or at least an annotated electronic map, to find the location either way.

            To get one meter resolution to identify individual doors (so it can be a viable alternative for street addresses even in densely populated areas) using latitude and longitude requires just over five decimal places for eac

            • Or I'm at horse, staple, battery?

              Twenty three words versus three. Which is more likely to introduce an unnoticed mistake?

              Or more likely:
              Operator: what is your location?
              Caller: I'm in the national park at the edge of the lake
              Operator: what 3 words?
              Caller: what?
              Operator: what 3 words?
              Caller: sorry what are you talking about?
              Operator: what is your 3 word location?
              Caller: I have absolutely no idea what you are talking about
              Operator: It's this new system that's makes things so much easier
              Caller: Well how is that working out so far?
              Operator: If you could just get your phone, download this app, register, and login, then g

        • It's easier to remember three words than it is to remember a series of numbers. It's meant to be simple to use.

          Or even 4 words. Correct Horse Battery Staple

      • by vlad30 ( 44644 )
        So I clicked your link and the random location was nielsen.sitcoms.raped you owe me a keyboard
        • by vlad30 ( 44644 )
          More importantly though tried those in what3words and got a very different result. So in my 5 second of research I can declare that any system that is not a standard and has difficulty in English let alone other languages should not be used. I'll stick to lat/long which my phone gives as standard location
      • Source code download link:

        https://github.com/pballett/wh... [github.com]

      • At least with longitude and latitude I can at least discern on what continent the location is on.

        Or if two places are close to each other

    • by Frobnicator ( 565869 ) on Monday May 03, 2021 @10:58AM (#61341482) Journal

      Looking for it, just like you.

      Their web site is suspended for DNS dispute, the Wayback Machine version says it has been excluded for legal reasons.

      With more digging, found an archival site version of their page [archive.is] as well as the JavaScript version [cryptome.org] and python version [cryptome.org].

    • Here's a re-implementation, written on my phone, in <30 minutes:


      import System.Environment
      import Data.Char
      import Data.List
      import Data.Maybe
      main = do
      d <- readFile "/usr/share/hunspell/en_US.dic"
      let ws = filter (all (`elem` ['a'..'z'])) $ takeWhile (/='/') <$> lines (map toLower d)
      wc = fromIntegral $ length ws :: Double
      args <- getArgs
      case length args of
      3 -> let [

  • You're fucked.

    (You're equals two words, btw. FYI.)

    • https://www.youtube.com/watch?... [youtube.com]

      Car Rental Agent : [cheerfully] Welcome to Marathon, may I help you?

      Neal : Yes.

      Car Rental Agent : How may I help you?

      Neal : You can start by wiping that fucking dumb-ass smile off your rosy fucking cheeks! And you can give me a fucking automobile: a fucking Datsun, a fucking Toyota, a fucking Mustang, a fucking Buick! Four fucking wheels and a seat!

      Car Rental Agent : I really don't care for the way you're speaking to me.

      Neal : And I really don't care for the way your compan

  • ⦠unless he actually (had access to and) copied their source code and/or data?

    • And another questionâ"when the hell is slashdot going to complete the 90â(TM)s adoption of Unicode?!

    • by Luthair ( 847766 )

      The article says they reverse engineered the algorithm then re-implemented (in a couple languages) a compatible version without any of the original code.

      To me this sounds like someone trying to use copyright law like patent law for an idea that was presumably not patentable otherwise they'd invoke their patent.

    • by EvilSS ( 557649 )
      Sounds like the lawyers are trying to manipulate DMCA, which you are right, they should be slapped for if true and he didn't copy their code. Unfortunately, it looks like they have several patents on the idea as well, and those would be an actual problem for him: https://www.cbinsights.com/com... [cbinsights.com]
  • UK laws (Score:5, Funny)

    by hdyoung ( 5182939 ) on Monday May 03, 2021 @10:48AM (#61341410)
    No surprise - this is the UK we're talking about. The legal bar for stuff like this stupidly low, just like the legal bar in the US is stupidly high.

    In the US, I can't win a libel suit unless they break into my house and murder my family while twirling their long mustache and live-streaming it on 3 different platforms simultaneously. In the UK, I can sue for libel, and win, if someone has the gall to politely criticize the font size in my product advertising.
    • I'm 90% sure those loopholes in the libel laws were slammed shut.

    • by AmiMoJo ( 196126 )

      It's not that bad in the UK. Private Eye has won many libel suits against it.

    • UK also has loser-pays legal system, so if you win your lawsuit you're looking at getting about 2/3 of your legal costs back, all of them if the losers lose badly enough. In the US you have to foot the bill for defending yourself against preposterous claims, and *maybe* you can get some of that back if you prove your opponent's case was so egregiously bad that it should never have seen the inside of a court room in the first place. Maybe. If you're lucky, and have a good enough lawyer.

      Also, XMission and

  • Seems they never heard of Barbara Streisand.

    Now, truly nobody will use it.
    I mean, before, nobody used it either.
    But now people will actively avoid it.
    The three people who have heard of it, and remembered.

  • by feedayeen ( 1322473 ) on Monday May 03, 2021 @10:55AM (#61341466)

    500 trillion square meters. Dividing that into 3 and taking cube root indicates about 50,000 unique addresses for the 3 coordinates. The Oxford English Dictionary has about 171,000 words. Apply limits on min and max length and compare both spelling and phonetic representation for edit distance to reject terms too similar.

    How is a lookup table with about ~50,000 words and a hashing function a company?

    • They probably need an algorithm that prevents generating three words that would offend somebody. I'm surprised they could come up with 50,000 unique phrases that doesn't offend anybody. That is some serious magic.
    • by Entrope ( 68843 )

      Why do you divide by three? The cube root alone reflects using three words. Then you need about 80,000 words for square-meter resolution of the ~510 trillion square meters on the surface of the globe.

      I always thought the company's magic was in the non-uniform mapping, which assigns a denser grid to places near major land masses.

      • I skipped a few words at the start, the 500 trillion value was square meters of surface area on the Earth. I'm interpreting a 3 meter area as a grid of squares approximately 1.7 meters on a side which is where the addital step occurs with that being the worst case.

        Variable density grids and it being 3 meters on a side instead would change the result some, but you're dealing with 20k to 50k words depending on what optimizations you make since thanks to the cube root, reducing the addresses by a factor of 10

        • https://support.what3words.com... [what3words.com]

          I take it back, I have no idea what they're doing. They seem to be claiming they have "4.8 billion combinations" so that would be 4.8 billion locations. That's close to the same number of buildings on the planet so I have no idea what they're actually doing but 3 meters square is clearly nonsense unless someone messed up billion and trillion.

        • by Entrope ( 68843 )

          Sure, but 20k versus 50k words is a big difference once one considers phonetic and morphological similarities, as you suggested in your earlier comment. The (original) Metaphone algorithm only yields 47k words with distinct encodings from the 100k-word /usr/share/dict/words on my system, and only 32k words with pre-encoding lengths between 4 and 10 letters (keeping the shortest original word for each encoding).

    • Same way "locate nearest idle participating driver and direct them to person wanting a ride" has made Uber a company. That's a far simpler algorithm.

      They came up with a useful addressing scheme, particularly for places where street addresses aren't a realistic option, wrote the software to implement it, and got busy with marketting. Might even have a few patents to keep the competition at bay for a bit.

    • Also, I think your math is wonky, why are you dividing by three?
      500 trillion possibilities expressed as a "3 digit number" = (500e12)^(1/3) = ~80 thousand possible values per "digit", so you need approximately 80,000 unique words to pick from.

      • Also, I think your math is wonky, why are you dividing by three?.

        My guess is that the world's surface is only 29.2% land, and there is little demand for 3 meter location precision in the ocean.

        • If it's for rescue, I would think the ocean matters too.

          When I listened to an interview with them they didn't lessen the resolution anywhere, but they did prioritize the easier words to the more likely locations in general.

      • They described it as 3 meter squares. If that means a square 1.7 meters on a side, division by 3 converts it from square meters to the squares w3w uses. It could also mean 3m x 3m which would be division by 9 instead. Further optimizations by unimportant regions can be applied on that but beyond the scope of my interest as due to the cube root reducing the cells by a factor of 10 reduces the count by only a factor of ~2.15.

        • Thanks, So they do - I somehow missed the "three-" part.

          Though, isn't a three-meter square 3x3 meters? So you should divide by 9.

          Yeah, Wikipedia say they have a resolution of roughly 3m.

    • How is a lookup table with about ~50,000 words and a hashing function a company?

      Because no one else has done it before. Therefore, it is a good idea, and worthy of SV venture capital.

      I just have to ask: What percentage of people have 50,000-word vocabularies? The average is about 30,000 in English.

    • by tlhIngan ( 30335 )

      Because it's extremely useful.

      But because it's a commercial entity and proprietary system, SAR (search and rescue) teams strongly discourage its use.

      This is one of those things where open-source would greatly save lives because it can then be built in to mapping apps and all that.

      Right now I believe they charge companies to integrate it into thei mapping apps and SARs don't want users to try to install apps while lost with a weak signal.

      So yeah, it's one of the odd cases where it can help, but chances are,

    • I'm asking that for almost every new company nowadays.

      How is WeWork a company?
      How is Twitter a business? Blogs, but wait! It's limited in charscter count! Because without them, that wouls be completely impossible!
      How are most apps a business? They are just a permutation of two functionalities that would be trivial to combine.
      How is thick water a business?

      I wrote a W3W-equivalent as a script, in a comment above, by the way. From words to coordinates and back. With arbitrary dictionaries.

      • Only bit of useful functionality it's missing is a edit distance comparison for likely candidates during a miss. Yeah, that makes it a bit more complex, but it's also a CS homework problem I remember.

    • THe correct way to calculate this, to match W3W's claims (3 m resolution) is to take the world's land area (510 trillion square meters * 0.292), divide that by 9 (3 m x 3 m), and take the cube root to get 25482 words.

      Using this usage-ranked list [norvig.com] of the 100,000 most common English words we can explore what a location made of the lowest ranked words might look like. Using the three lowest out of the top 25482 we get "LIAS-LUMPUR-GULLY".

      EIther they will have to expand the word list after imposing special rule

  • Dumb Ass Idea

    Anybody who was worked with geospatial data and geocoding would understand why.

    • by ichthus ( 72442 )
      Pretty good idea.
      Anybody who types domain names into their browser, instead of IP addresses, would understand why.
    • It's a lousy solution for working with geospatial data. It is however much better than geospatial data for ordering pizza in a place with no street addresses. Much less having your 5yo child remember their home address in case they get lost.

      You need five decimal places per coordinate for 1m resolution with lat/long. That means around twenty words to read out the same level of detail. And an error in the any of the last several digits is likely to go undetected until you arrive at the wrong location, whi

      • With street level addressing a child would need to remember one word and finding their home would not be that difficult. Plus, would you want to trust finding your child to a company that exhibits the behavior described in the article? Aside from that, what is their business model? Who would want to pay this company their vig to get something that is not demonstrably better than the many other (and free) geocoding alternatives?
    • So, when somebody asks for an address, you give them degrees/minutes/decimal seconds?

      Or do you give them an address?

      All that this concept is, is decoupling 'address' from 'physical paths somebody built.' There's little conceptual difference between saying 'I live at 12 Wankerstrasse' and 'I live at Hidden Forbidden Holyground.'

  • Reminds me of something I did once, I had a site I used that kept throwing up a random banner asking me to download the mobile app. It had no way to continue or defeat it.

    So I wrote up a nasty-gram to the company and basically said. I am a developer. I will create a plugin for browsers that remove your banner and give people a way to keep using the site without using a mobile app. I will give it away for free and then other features will appear whenever you do something we don't like.

    Or you can just put a "

    • by piojo ( 995934 )

      Wow, that was you? Thank you! I just assumed their analytics told them they were losing users with their heavy handedness. (Both factors may have had influence.)

      • Yep, I actually did this to a few companies, some of them gone now... but this story was about Quora.

        I've also done similar to Grooveshark (before they got shut down), and complained hard to StackOverflow about similar practices but they aren't scared of developers there obviously.

        But yeah you probably guessed some other site which I don't know if I deserve credit for, but if it was Quora, that was me :)

      • Are we talking about Reddit here?

  • Then it's not a copy and the commercial version is no longer compatible with the free version. I consider that an improvement.
  • The letter gave him until May 7 to agree, after which What3Words would "waive any entitlement it may have to pursue related claims against you," a thinly-veiled threat of legal action.

    If What3Words "waives" their "entitlement...to pursue related claims", wouldn't that mean they would be giving up said "entitlement"? In other words, that they would be giving up any right to sue? Or have I had stroke? Does anyone else smell toast?

    • "Do what we say, and we'll promise not to take you to court."

      The unspoken corollary: 'don't do what we say, and we will take you to court. We may or may not win, we may or may not even expect to win, but you'll spend a lot of money you probably don't have in the process.'

  • Full post at https://wcodes.org/patents [wcodes.org]

    TL;DR The W3W team UK patent-filed a week after WCodes' author publicly shared their work (albeit with a marginally different 5-word output with limited 1k dictionary entry, as proof of concept). The UK court granted W3W the patent in 2018 despite the obvious prior-art. W3W is hereby a patent troll for doing this stuff.

To be is to program.

Working...