RE: yet another time error... ...ooops, mustn't call it an error, mustwe

["Creager, Robert S" <CreagRS@LOUISVILLE.STORTEK.COM>]

Hey Tom,

What version of Windows are these? XP has NTP built in.  Anyway, one of your
linux systems can be setup to broadcast, and then you can use K9
(http://www.kaska.demon.co.uk/) to listen to the broadcasts and set the
time.  Works well.  There is probably other ways to do this.

If you want to use K9, I can setup your NTP server to broadcast.  If someone
else has other ideas for the Windows side, we can make it look at the linux
NTP server.

Cheers,
Rob

> -----Original Message-----
> From: owner-tass@listserv.wwa.com 
> [mailto:owner-tass@listserv.wwa.com] On Behalf Of droege@snapmail.us
> Sent: November 05, 2004 1:56 PM
> To: Michael Koppelman
> Cc: tass
> Subject: Re: yet another time error... ...ooops, mustn't call 
> it an error, mustwe
> 
> 
> Michael,
> 
> I do what I can.  I now have a nice wrist watch that is 
> automatically set from NBS.  It keeps time as good as can be 
> detected.  Before each night's run I set the computer clocks 
> from the watch. Normally to +/- 2 seconds.
> 
> Not ideal, but I am pretty careful.  I know you could set up 
> what is required, but I can't.  Lots of computers on my 
> system have their time 
> set automatically, but not the Windows machines that are 
> taking data.  For example, mike, on which I am writing, is 
> set by NTP.  If anyone wants to write a program that sits on 
> a Windows 98 machine and sets it's clock by reading one of 
> the linux machines, then go for it.  However, at the moment I 
> can read the Windows machines from the linux machines, but 
> not the other way around.  I also do not want the Windows 
> machines connected to the Internet, because then Bill sends 
> updates in the middle of a run and hangs the data collection 
> program.  When Rob was here he set up the network scheme and 
> I have been following his set up ever since. The set up did 
> not include reading linux from Windows.
> 
> OK, the complainers can find another system for data.  
> Believe me it is a lot of hard work to do what I do.  I try 
> to put my effort into the most important things.  People who 
> complain about the time stamp just don't understand the data. 
>  I would put an improved time scheme at about 27 on the list 
> of things that would be good to do. All are welcome to step 
> in and improve things where they can.  Now that I am 
> connected on line some of you could step in and fix things 
> like this.  You will no doubt need to send me messages like 
> "go to your windows machine and in the foo window select goo 
> and set permission x to y"
> 
> Hmmmm!  Here is a nice winter project for someone out there.  
> Once the clouds roll in (soon though I ran last nigt) I could 
> set up a telescope with a linux computer.  Then somoeone out 
> there could get it running under linux.  A nice project for 
> some of you that have always said they wanted to help.  Since 
> Rob is running under linux, it should be just a question of 
> moving his code to one of my machines.  Then we will be done 
> with Windoz.
> 
> I would put such a project far ahead of my manual time set in 
> priority. 
> This would then also solve the (non) time set problem.
> 
> Tom Droege
> 
> 
> > Are you setting the time automatically now? I am uneasy 
> with a wrist 
> > watch situation. I know we have discussed a myriad of solutions for 
> > networking and time synchronization in the past. Just 
> curious what the 
> > status is. My PC clock is synchronized before each night's 
> run so my 
> > data is probably time stamped correctly to within a second or less.
> >
> > Cheers,
> > Michael Koppelman
> > http://www.lolife.com/astronomy/
> >
> >
> >
> 
>