As I said in IRC, I'm firmly of the mind that we should make it easy to ensure that all time data is consistant. I think the easiest way to do this is to standardize on UTC. Whatever time zone your server hardware/OS is in, you can always get UTC from it.
I am not averse to having a 'super advanced OMG what the heck are you doing' option to set a UTC offset above and beyond.. but I really think it should be UTC by default across the board. Best Regards Teravus On 1/18/09, Gerhard Dünnebeil <[email protected]> wrote: > When dealing with time you have to distinguish between the time the > machines work on and the time the users sees. > > Introducing time zones at the wrong level also introduces a lot of > confusion as you (the developer) always have to know with which time > zone you currently deal. > > MySQL isn't really an issue as you can set the time zone you want to > work in with the session. > > So my thoughts: > Use UTC internally; that includes storage of data, machine to machine > communication, .... > Use the users time zone when (and only when) communicating with the > user, which is mainly when times are displayed. > > This is easier as one might think, as using UTC as the time base is > available in all standard OS (windows, linux, ...) these days. There is > no need to set a system clock to UTC as long as the system knows the > offset to UTC. > > Gerhard > > > Dahlia Trimble wrote: > > With the err... um... inevitable future event when the LL grid opens > > the doors to full interoperability, and given that their large > > customer base is accustomed to using "SLT" (California time) and all > > the scripts that may assume SLT, shouldn't we weigh that option over > > UTC or GMT or CUT or whatever it's called these days? > > > > On a side note, regions running in virtual machines may have less > > control over the system clocks than regions running in a regular > > machine. I'd like to suggest that region times could be configured in > > OpenSim.ini and/or set by a central server using ntp or a similar > > protocol. > > > > On Sun, Jan 18, 2009 at 10:32 AM, Charles Krinke <[email protected] > > <mailto:[email protected]>> wrote: > > > > There are some issues coming up about timezones and grids. It > > seems that events, scripts, web interfaces and other things are > > affected by timezones. It gets a bit more complicated when one > > considers that our mysql logic uses local time with calls that > > involve NOW(). > > > > Some folks feel that all the time on all the sims on all the grids > > should be set to UTC, and that is a reasonable approach. > > > > Others feel that a UGAIM should set the timezone for a grid and > > tell the sims connected to that UGAIM what timezone they are in > > and what time it is at the time of grid registration with a UGAIM, > > and that is also a reasonable approach. > > > > There is more. But, rather then just declaring what I might feel > > to be the answer, I would like to seek a consensus that is best > > for OpenSim and any comments or suggestions are greatly > > appreciated to this "timely" subject. > > > > Charles > > > > _______________________________________________ > > Opensim-dev mailing list > > [email protected] <mailto:[email protected]> > > https://lists.berlios.de/mailman/listinfo/opensim-dev > > > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Opensim-dev mailing list > > [email protected] > > https://lists.berlios.de/mailman/listinfo/opensim-dev > > > > _______________________________________________ > Opensim-dev mailing list > [email protected] > https://lists.berlios.de/mailman/listinfo/opensim-dev > _______________________________________________ Opensim-dev mailing list [email protected] https://lists.berlios.de/mailman/listinfo/opensim-dev
