[tz] tz leap-seconds.list

Paul Eggert eggert at cs.ucla.edu
Sun Oct 25 21:51:55 UTC 2020

On 10/25/20 10:45 AM, Kristin Minchev wrote:

> The [leap-seconds.list] file itself doesn't have any license information 
> (nor NIST's).

Under current US law a file that contains no license information normally 
defaults to being copyrighted with no license granted. The NIST is an arm of the 
US government, though, so under US law its publications are public domain so we 
don't have to worry about redistributing the NIST file even though it contains 
no license info.

I just now searched for the top ten "leap-seconds.list" hits on Google. The #1, 
#3, #4, #6, #7 and #9 hits were about the NIST file, and of these all but #9 
were about the tzdb copy of the NIST file (#9 was unclear). The #5 and #10 hits 
were about the IERS file. (#2 and #8 were false matches.) So even though the 
IERS file is published by the primary source for leap seconds, licensing issues 
have prevented the IERS file from being the most popular in practice. We've 
attempted to contact the IERS multiple times to get its file's license 
clarified, but haven't made much progress.

> And most of the data comes from IERS anyway.

That's fine: under US law, data cannot be copyrighted. There was a court case 
about this; see 

I installed the attached patch to the tzdb development repository to try to 
explain this all a bit better, without getting too long-winded about it.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-leapseconds-now-says-why-NIST-not-IERS.patch
Type: text/x-patch
Size: 1150 bytes
Desc: not available
URL: <http://mm.icann.org/pipermail/tz/attachments/20201025/ef2eb4d4/0001-leapseconds-now-says-why-NIST-not-IERS.patch>

More information about the tz mailing list