In Tampa bay AT&T is about 17 seconds off -- Mike
On Jul 16, 2010, at 1:00, Peter Monta <[email protected]> wrote: > Oz-in-DFW writes: > >>> ... There is no way AT&T would be 12.4 seconds off ... >>> >> >> I used to work in the cell infra business. While it's less true today, >> there are still a number of operators that do not sync system clocks. >> The time supplied to users can be **minutes** off. >> >> Most newer operational standards can't tolerate this and "accurate" time >> (better than a ms) is important. WiMAX requires TDD base stations to >> base station alignment to be better than 1 microsecond. Most telecom >> operators want to avoid GPS at every site. It's a logistical PITA. > > Here in the Bay Area, AT&T/iPhone time has gotten noticeably worse > recently. The error used to be around 4 seconds; now it's 49 seconds (!). > > Emerald Time is fine for interactive use, but what I find very impolite > is that AT&T's bad timestamps are written into the EXIF headers on photos. > Sometimes I take pictures of sundials, for example, and a 49-second > error is not negligible for a carefully made dial. > > It would be amusing to arrange for a long-term record of the offset of > one's phone (which can of course change across multiple providers during > travel), say by using a background process to take a sample every few hours > against NTP sources or against GPS if the phone has it (or both). > Then any photos can be batch-corrected later if desired. Apple, give > me control over the time on my own phone, and please don't force me to > resort to these schemes :-). > > Cheers, > Peter Monta > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
