8 ms·

If you ask a lawyer, you will get a "You should update them". If you ask them to justify it or pushback, you'll normally get a "Yeah, no need". If something ha

If you ask a lawyer, you will get a "You should update them". If you ask them to justify it or pushback, you'll normally get a "Yeah, no need". If something has no perceived cost, and has marginal utility, then you'll get asked to do it.

colejohnson66 4y ago
It's one of those things that you don't need to do, but it doesn't hurt to do. So managers (and lawyers) will recommend you do it anyway, possibly as a CYA measure.
- ipqk 4y ago
  It adds noise to the "last updated" timestamp. Yes, you can always find the last substantial update to the file, but now it's not as up front. It's not a huge cost, but it's a cost nonetheless.
- xoa 4y ago
  Yes, though "CYA" isn't necessarily entirely fair either. The legal system isn't exactly entirely deterministic, particularly if there is a jury involved, so I think the tendency towards conservatism and seeking every edge isn't entirely unreasonable. Cost does certainly have to be weighed against any measures, and particularly for technical issues costs may often not be obvious to lawyers themselves or managers. But after actually experiencing all the stress and costs of a legal case just a few times, and only as a peripheral player, I can understand worries around all of it. Also to the root's line here: >I gave multiple examples of other code bases at Microsoft and Google and other big companies that dont include the year at all in some cases and definitely dont update every year. This is one of those things where it seems reasonable at first blush but then completely not on second look. There are lots of cases in tech where "look these huge players do this" makes sense, but when it comes to either scaling or law things are different. Here I'd be looking for examples of small players doing that, because the big players have such powerful law teams and deep pockets that risks they'd dismiss as unimportant might be existence-threatening to a small company.
  - lapcat 4y ago
    Is there a single case in all of legal history where someone has lost a copyright because of failure to update the copyright year? If not, then you don't need to do it.
    - amanaplanacanal 4y ago
      As far at I know, you don't even need a copyright statement anymore. Everything is copyright by default.
      - monkpit 4y ago
        This was my understanding as well, I was surprised that the article didn’t end with the advice to remove it altogether.
        lapcat 4y ago
        You don't need a copyright notice for legal reasons, but it can deter boneheads who mistakenly believe that anything without a copyright notice can be copied freely.
        ghaff 4y ago
        I doubt it deters people who are just going to grab it anyway.
      - wongarsu 4y ago
        Which is how it has been for most of the Western world for well over a century. Automatic copyright one of the core principles of the 1886 Berne Convention, but the US only joined in 1989.
        ghaff 4y ago
        It surprises people that the US had, relatively speaking, more copyright hoops for creators to jump through than most of the rest of the world prior to harmonizing with the Berne Convention.
- maccard 4y ago
  > but it doesn't hurt to do Except it does. It requires an engineer to maintain the dates, and the company to audit that the dates are correct. It directly interferes with source control ("updated copyright" was the bane of my existence when trying to git-bisect a few years ago).
  - rocketbop 4y ago
    Surely it’s straightforward to use templating and always pull the year from a date library. No changes to source control needed. Edit: as the article mentions open source documentation it’s possible the repo itself is what needs the year, and there’s no way around hardcoding it in the included licence.
    - JohnFen 4y ago
      That's how I've always done it. I'll probably continue to include copyright notices and years regardless of legal necessity, but I've never lost sleep over it or put a lot of effort into it. It just seems like a "does no harm, and may clarify things" sort of deal. This is pretty much the advice of my attorney. "It's not legally required, but I can imagine situations where it could be useful anyway. If it doesn't cost you anything, why not include it?"
- thih9 4y ago
  > CYA measure Apparently CYA stands for: "cover your ass".
- arrrg 4y ago
  I think legally meaningless CYA statements are actually dangerous. They can suck the air out of actual understanding of the legal situation at hand and, in the worst case, lead to an warped understanding of copyright that interacts with the world with supposed magical incantations that are legally totally meaningless or even the opposite of the truth. (I’m thinking of those YouTube videos e.g. showing some snippet of a film uploaded by someone who doesn’t have the copyright with the totally bonkers “copyright Universal 2020, no infringement intended” note. I know, that‘s mostly about someone who is not a business and who usually doesn’t have anything to do with copyright being confused about the difference between attribution and copyright and thinking giving attribution gives you copyright – which is bonkers. Fair use does require attribution, but simply copying something verbatim is certainly not fair use.)
  - cxr 4y ago
    Another thing to be cognizant of is the Gell–Mann amnesia effect and adjacent phenomena. Especially when people talk about professions like lawyers, etc. it's almost as if they conceive of it as one unified workforce of hypercompetent experts who all know The Truth and can either tell you what it is or, failing that, then at least tell you what you should do in accordance with what The Truth is. But no industry works like that. Think of how many programmers you know or know of and the range of competencies involved, including absurd practices and just outright weird opinions/positions held by a surprisingly large proportion of folks. Half these people are B- players at best and who are only there because they figured out how to make a living following rituals—which they may not necessarily understand (and some of which may actually be counterproductive but are inoffensive in the sense that they aren't so counterproductive that it will result in being let go). Or forget that for a moment and just think of how many cases have been tried in court with lawyers on both sides. Do the courts regularly arrive at decisions that recognize counsel for both sides as being right (and it's just, say, the facts that are responsible for their clients losing the cases)? Or can you regularly find lawyers putting forth legal theories that simply don't hold up under scrutiny?
    - Georgelemental 4y ago
      There is a lot of credentialism and vetting involved in becoming a lawyer, which should ideally ensure a minimum floor of competence.
- chrismorgan 4y ago
  > but it doesn't hurt to do I’m not entirely convinced about this. It constitutes a false statement about the copyright status of the work, which may be illegal in some jurisdictions. There are a few unlikely scenarios in Australia where it might technically become a crime, though by and large Australia ignores anything but authorship in matters of such notices; but mostly it would require that you try to threaten someone unjustly before it would become a crime. In the United States of America, I think it becomes a crime to even claim copyright incorrectly (meaning “this is mine and is in copyright” when it’s public domain, nothing to do with year numbers), without even needing to threaten someone. Mitigating this, both require some degree of knowledge or reasonable expectation thereof. So, supposing a work-for-hire copyright term of fifty years (yeah, AU and US are both more than that, but I’m simplifying), if a work was actually produced and published in 2000 and you write “copyright 2010–2020” on it, this could actually be technically criminal between 2050 and 2070 in some situations, during which period the work is actually in the public domain but the claim implies it’s not, constituting copyfraud <https://en.wikipedia.org/wiki/Copyfraud https://en.wikipedia.org/wiki/Copyfraud>. (Fun fact, while thinking of this stuff: Australia has mandatory legal deposit of all publications, since 2016 explicitly including things like websites. I don’t do this for my website, and I’ve never heard of anyone else doing it—in fact it was only a few years ago that I heard about this law at all, when my dad started publishing books; it’d be frightfully inconvenient to do for a website, and it’s all rather ill-specified.) So, it might theoretically be illegal in some areas in some situations. And I think it should be illegal; it should hurt to go bumping copyright years meaninglessly. I also just find it funny to think about this practice of bumping copyright years coming back to bite your successors 70 or 95 years down the track, when what seemed harmless silliness at the time suddenly becomes a criminal misrepresentation. ’Course, I don’t think there’s actually much value speculating about specifics of copyright law that far in advance. The last century has shown progress in some areas of copyright (doing it automatically) and massive folly in other areas (obscene and ever-growing durations).
lamontcg 4y ago
Explain to the lawyer that these things are hand managed and that dates will slip in which have been copied from other headers in other files, making the years in the headers unreliable, and that if it ever came up in court that you'd have to testify to their unreliability -- while also explaining that git was considerably more reliable and that any one individual commit which had incorrect dates could generally be identified by surrounding dates. That will generally tickle their aversion to putting data that could fail under cross examination in front of a court and they'll be more likely to let you delete all the dates entirely. There's literally no reason for them anyway. You don't even need a copyright statement, but you probably do have a copyright file with years in it, and you've got git for supplimentary info on exactly when any given line of code went into the codebase. Also if you're updating every file every year you're losing the information on when a piece of code when into the codebase, which is the only important thing anyway in the case of patent lawsuits, expiration of copyright really isn't an issue worth worrying about (and I think bumping the copyrights every year is wrong policy there as well). While you're at it, remove all the Author lines as well with the people who no longer have anything to do with the project and their old e-mails which don't work.

colejohnson66 4y ago

It's one of those things that you don't need to do, but it doesn't hurt to do. So managers (and lawyers) will recommend you do it anyway, possibly as a CYA measure.

ipqk 4y ago
It adds noise to the "last updated" timestamp. Yes, you can always find the last substantial update to the file, but now it's not as up front. It's not a huge cost, but it's a cost nonetheless.
xoa 4y ago
Yes, though "CYA" isn't necessarily entirely fair either. The legal system isn't exactly entirely deterministic, particularly if there is a jury involved, so I think the tendency towards conservatism and seeking every edge isn't entirely unreasonable. Cost does certainly have to be weighed against any measures, and particularly for technical issues costs may often not be obvious to lawyers themselves or managers. But after actually experiencing all the stress and costs of a legal case just a few times, and only as a peripheral player, I can understand worries around all of it. Also to the root's line here: >I gave multiple examples of other code bases at Microsoft and Google and other big companies that dont include the year at all in some cases and definitely dont update every year. This is one of those things where it seems reasonable at first blush but then completely not on second look. There are lots of cases in tech where "look these huge players do this" makes sense, but when it comes to either scaling or law things are different. Here I'd be looking for examples of small players doing that, because the big players have such powerful law teams and deep pockets that risks they'd dismiss as unimportant might be existence-threatening to a small company.
- lapcat 4y ago
  Is there a single case in all of legal history where someone has lost a copyright because of failure to update the copyright year? If not, then you don't need to do it.
  - amanaplanacanal 4y ago
    As far at I know, you don't even need a copyright statement anymore. Everything is copyright by default.
    - monkpit 4y ago
      This was my understanding as well, I was surprised that the article didn’t end with the advice to remove it altogether.
      - lapcat 4y ago
        You don't need a copyright notice for legal reasons, but it can deter boneheads who mistakenly believe that anything without a copyright notice can be copied freely.
        ghaff 4y ago
        I doubt it deters people who are just going to grab it anyway.
    - wongarsu 4y ago
      Which is how it has been for most of the Western world for well over a century. Automatic copyright one of the core principles of the 1886 Berne Convention, but the US only joined in 1989.
      - ghaff 4y ago
        It surprises people that the US had, relatively speaking, more copyright hoops for creators to jump through than most of the rest of the world prior to harmonizing with the Berne Convention.
maccard 4y ago
> but it doesn't hurt to do Except it does. It requires an engineer to maintain the dates, and the company to audit that the dates are correct. It directly interferes with source control ("updated copyright" was the bane of my existence when trying to git-bisect a few years ago).
- rocketbop 4y ago
  Surely it’s straightforward to use templating and always pull the year from a date library. No changes to source control needed. Edit: as the article mentions open source documentation it’s possible the repo itself is what needs the year, and there’s no way around hardcoding it in the included licence.
  - JohnFen 4y ago
    That's how I've always done it. I'll probably continue to include copyright notices and years regardless of legal necessity, but I've never lost sleep over it or put a lot of effort into it. It just seems like a "does no harm, and may clarify things" sort of deal. This is pretty much the advice of my attorney. "It's not legally required, but I can imagine situations where it could be useful anyway. If it doesn't cost you anything, why not include it?"
thih9 4y ago
> CYA measure Apparently CYA stands for: "cover your ass".
arrrg 4y ago
I think legally meaningless CYA statements are actually dangerous. They can suck the air out of actual understanding of the legal situation at hand and, in the worst case, lead to an warped understanding of copyright that interacts with the world with supposed magical incantations that are legally totally meaningless or even the opposite of the truth. (I’m thinking of those YouTube videos e.g. showing some snippet of a film uploaded by someone who doesn’t have the copyright with the totally bonkers “copyright Universal 2020, no infringement intended” note. I know, that‘s mostly about someone who is not a business and who usually doesn’t have anything to do with copyright being confused about the difference between attribution and copyright and thinking giving attribution gives you copyright – which is bonkers. Fair use does require attribution, but simply copying something verbatim is certainly not fair use.)
- cxr 4y ago
  Another thing to be cognizant of is the Gell–Mann amnesia effect and adjacent phenomena. Especially when people talk about professions like lawyers, etc. it's almost as if they conceive of it as one unified workforce of hypercompetent experts who all know The Truth and can either tell you what it is or, failing that, then at least tell you what you should do in accordance with what The Truth is. But no industry works like that. Think of how many programmers you know or know of and the range of competencies involved, including absurd practices and just outright weird opinions/positions held by a surprisingly large proportion of folks. Half these people are B- players at best and who are only there because they figured out how to make a living following rituals—which they may not necessarily understand (and some of which may actually be counterproductive but are inoffensive in the sense that they aren't so counterproductive that it will result in being let go). Or forget that for a moment and just think of how many cases have been tried in court with lawyers on both sides. Do the courts regularly arrive at decisions that recognize counsel for both sides as being right (and it's just, say, the facts that are responsible for their clients losing the cases)? Or can you regularly find lawyers putting forth legal theories that simply don't hold up under scrutiny?
  - Georgelemental 4y ago
    There is a lot of credentialism and vetting involved in becoming a lawyer, which should ideally ensure a minimum floor of competence.
chrismorgan 4y ago
> but it doesn't hurt to do I’m not entirely convinced about this. It constitutes a false statement about the copyright status of the work, which may be illegal in some jurisdictions. There are a few unlikely scenarios in Australia where it might technically become a crime, though by and large Australia ignores anything but authorship in matters of such notices; but mostly it would require that you try to threaten someone unjustly before it would become a crime. In the United States of America, I think it becomes a crime to even claim copyright incorrectly (meaning “this is mine and is in copyright” when it’s public domain, nothing to do with year numbers), without even needing to threaten someone. Mitigating this, both require some degree of knowledge or reasonable expectation thereof. So, supposing a work-for-hire copyright term of fifty years (yeah, AU and US are both more than that, but I’m simplifying), if a work was actually produced and published in 2000 and you write “copyright 2010–2020” on it, this could actually be technically criminal between 2050 and 2070 in some situations, during which period the work is actually in the public domain but the claim implies it’s not, constituting copyfraud <https://en.wikipedia.org/wiki/Copyfraud https://en.wikipedia.org/wiki/Copyfraud>. (Fun fact, while thinking of this stuff: Australia has mandatory legal deposit of all publications, since 2016 explicitly including things like websites. I don’t do this for my website, and I’ve never heard of anyone else doing it—in fact it was only a few years ago that I heard about this law at all, when my dad started publishing books; it’d be frightfully inconvenient to do for a website, and it’s all rather ill-specified.) So, it might theoretically be illegal in some areas in some situations. And I think it should be illegal; it should hurt to go bumping copyright years meaninglessly. I also just find it funny to think about this practice of bumping copyright years coming back to bite your successors 70 or 95 years down the track, when what seemed harmless silliness at the time suddenly becomes a criminal misrepresentation. ’Course, I don’t think there’s actually much value speculating about specifics of copyright law that far in advance. The last century has shown progress in some areas of copyright (doing it automatically) and massive folly in other areas (obscene and ever-growing durations).

lamontcg 4y ago

Explain to the lawyer that these things are hand managed and that dates will slip in which have been copied from other headers in other files, making the years in the headers unreliable, and that if it ever came up in court that you'd have to testify to their unreliability -- while also explaining that git was considerably more reliable and that any one individual commit which had incorrect dates could generally be identified by surrounding dates. That will generally tickle their aversion to putting data that could fail under cross examination in front of a court and they'll be more likely to let you delete all the dates entirely. There's literally no reason for them anyway. You don't even need a copyright statement, but you probably do have a copyright file with years in it, and you've got git for supplimentary info on exactly when any given line of code went into the codebase. Also if you're updating every file every year you're losing the information on when a piece of code when into the codebase, which is the only important thing anyway in the case of patent lawsuits, expiration of copyright really isn't an issue worth worrying about (and I think bumping the copyrights every year is wrong policy there as well). While you're at it, remove all the Author lines as well with the people who no longer have anything to do with the project and their old e-mails which don't work.