On 2025-02-06 13:12, Peter Wienemann wrote:
As part of this change copyright years were removed from the copyright notice (see [0]). According to [1] this happened because the upstream authors "got tired of updating those years and it matches the LF recommendation" (where "LF recommendation" is a link to [2]).

I maintain a package which did away with the years upstream.

I personally like the removal of the years. I've since followed that myself in many cases.

In theory, the year should be bumped on the first change of the year--that is sufficiently expressive to obtain copyright--to a given file. A close approximation of that would be to bump the year on the first change of the year to a given file.

Some (many?) projects will update the year on every copyright statement every year. IMHO, this is pointless busywork that's also wrong (because the passage of time does not grant a new copyright).

Other times, the years only get updated sporadically. In that case, they aren't accurate either. So what purpose do these years serve in practice?

Copyright notices aren't required anyway. Even if there was no notice, the code would still be under copyright. So even if the lack of the year makes the notice legally defective, it's moot.

Also, if one thinks years should be updated, do you also think that the list of authors needs to be correct on a file-by-file basis? If so, the list of copyright notices could easily get absurdly long for a long-running popular project. And even then, many of those notices will be out-of-date as someone's code is removed or rewritten.

--
Richard

Attachment: OpenPGP_signature.asc
Description: OpenPGP digital signature

Reply via email to