Hi Roger,
I use UTC(Universal Time, Coordinated) for standardised terminology to
represent date/time insted of GMT. UTC and GMT point almost same time.
http://en.wikipedia.org/wiki/Coordinated_Universal_Time
On Sat, 28 Nov 2009 15:38:08 +0100
Roger Erens <roger.erens at e-s-c.biz> wrote:
> In order to make things easier to discuss: suppose you, me, and Tim
> move an MS-Excel file around. It will contain rows with just two
> columns: factnumber and a GMT date-time (e.g. "fact1",
> "20100924_1659").
> I assume you will do the following in the simulation during the
> connect-a-thon:
>
> 0) Startup up an ntp-daemon that broadcasts the time of your
> workstation (I'm not knowledgeable about this?)
> 1) Adjust the time of your workstation to GMT 20100924 16:58
> 2) Now you enter in MS Excel the first record ("fact1",
> "20100924_1659") and save the file
> 3) You adjust the datetime of your workstation to GMT 20101016 11:15 and
> 4) Have Tim's and my workstation pick up that date-time
> 5) Send the file to Tim
> 6) Now Tim enters in OpenOffice the second record ("fact2",
> "20101016_1117") and saves the file
> 7) Tim tells you via irc/GoogleWave/Skype/phone that he changed the file
> 8) You adjust the date time of your workstation to GMT 20110331 08:45 and
> 9) Have Tim's and my workstation pick up that date-time
> 10) Tim sends the file to me
> 11) I use some python scripts to update the file with a third record
> ("fact3", "20110331_0849")
> 12) I send that file to you at the connect-a-thon
> 13) You open up the file and show it to the public (and collect the applause)
> 14) You clean up the file and start another journey with new future
> date-times for another round at the connect-a-thon
>
> I'd like to know how you propose to do step 0, 3, 4, 8, and 9.
To simplify this date/time problem, I assume this record is observation
entry because it has only one time to record and no workflow. And we
must distinguish two time in records, when it occured and when it was
recorded. Events in hospitals are not recorded just when they happened, but
in most cases, they recorded after they happened.(For example, doctor
writes patient record after they examed patient).
0) NTP daemon is available anywhere in the Internet
We can adjust RTC within 10 mili seconds over the Internet automatically.
http://en.wikipedia.org/wiki/Network_Time_Protocol
3-9) If the recorded time means the event occured or event will occured,
we do not have to adjust workstation time.
I understand what you and Tim have addressed only just now. Sorry to
have much time and bother you.
In this connectathon, I assumed virtual time flow. As your example,
0) Adjust all workstation using NTP server
1) Tim send me the file recorded 'event1', '2009-10-30T12:18:11BST'
2) I received file and record it after change BST to UTC 'event1,
'2009-10-30T09:18Z' and record with timestamp.
3) I send Tim with the file recorded 'event2', '2010-01-22T01:22:22JST'
4) Tim receive the file and record 'event2', the time may be changed to
UTC/BST with timestamp.
Each system generate sample records before connectathon. To simplify
more, all the record within system and messages should be record in UTC
not localized timezonen and presentation layer of each system translate
UTC to localized time zone.
--
KOBAYASHI, Shinji <skoba at moss.gr.jp>
