On 2021-06-01 20:42, Guy Harris via tz wrote:
On Jun 1, 2021, at 5:57 PM, David Patte via tz <tz@iana.org> wrote:
I write both astronomy and family history apps, and I write my own parsers to extract the data I want from tz tables on the fly. I believe many using the historical data portion of tz would likely do the same thing. Type (2) people simply want to continuously improve the accessibility, locality and accuracy of historical tz data, for apps that want PAST tzdata, not future tz data. It's not a difference between wanting past data vs. wanting future data. OSes definitely use past tz data: $ ls -lT /usr/local/bin/yacc -rwxr-xr-x 1 root wheel 123916 May 6 14:18:05 2018 /usr/local/bin/yacc The difference is how far back in time is useful, not whether the past is useful at all. For OS purposes, times before the OS's time stamp epoch are probably of less interest than times at or after the epoch, and for many OSes, even times after the epoch but before the first release of the first OS to use any file system supported by the OS are of less interest. The UNIX epoch is probably a good dividing line between the two groups.
It can be useful to tag (touch) (scans of) historical or official documents, pictures, media with real timestamps corresponding to the date issued, taken, or expired, as well as location, and time zone info. The current range covered by GNU date: @-67768040609740800 -2147481748 Jan 01 Mon 00:00:00 @67767976233532799 2147483647 Dec 31 Tue 23:59:59 is stupendously more than historical, supported on BtrFS, BSD UFS2; the MS Windows NTFS date range from 1601-01-01 (to 60056-05-28) is more useful than the Linux ext2-4 range from 1901-12-13 (to ext4 2446-05-10), and Apple FS from 1970-01-01. J/MPEG EXIF/IPTC/XMP metadata (exiftool) uses ISO-8601 strings, allowing for uncertain values specified as 00 or spaces <https://exiftool.org/forum/index.php?topic=5007.0;all> but not all file content formats support user specified metadata. -- Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada This email may be disturbing to some readers as it contains too much technical detail. Reader discretion is advised. [Data in binary units and prefixes, physical quantities in SI.]