"Simon (Vsevolod) Ilyushchenko" wrote: > 1. Bob: The reason I'd like to have the option of timezones is that modern > timestrings like ISO8601 often have them, so using Z to indicate UTC (which > is indeed the majority of our timestamps) would both handle the common case > and allow the possibility of using other timezones.
When I last read bits of 8601, I can't recall it allowing textual names of TZ, I hope not. https://en.wikipedia.org/wiki/ISO_8601#Usage Best avoid any TZ scheme that allows textual names for local timezones, Insist on numeric only. I once wrote C to make sense of dates in mail headers, http://www.berklix.com/~jhs/src/bsd/jhs/bin/public/mailname/mailname.c It was a pain, needing extensions as I discovered more acronyms. Then I wondered about the world of politicians who muck around changing whether & when they will change dates of summer / winter shifts this year, or go double summer time, or suppress summer time & stick to winter / solar, or lock permanently on a 1 hour summer offset, & that for every nation on earth, & regions within some contries, & eg I recall Australia used to have 3 x 40 min shifts across the Nullabor. All the ever changing timezone complexity is dealt with eg in FreeBSD in special libraries, & most likely different libraries on other BSD & Linuses, so a porting hastle. Best avoid porting or maintenace work by insisting on numeric only date strings (+ 'Z' & '-' & ':' ). Simon (Vsevolod) Ilyushchenko wrote: > Unix type systems (and those using NTP) have a system clock which is > on UTC time. Long ago I had a BSD 4.2 binary kernel whose native time was PDT=GMT-7 http://www.berklix.com/~jhs/symmetric/ I think numerous really old Unix binary kernels did. It would seem risky assuming all modern commercial binary kernels use UTC ? even if the modern PD src/ BSD & Linux may. Cheers, -- Julian Stacey http://berklix.com/jhs/ http://stolenVotes.uk Arm Ukraine, Zap killer Putin, grain & fuel loss hits poorest. _______________________________________________ Tiff mailing list [email protected] https://lists.osgeo.org/mailman/listinfo/tiff
