[coyotos-dev] coyotos.sleep, sleepTill, the epoch

[Valerio Bellizzomi]

On 11/02/2007, at 9.15, Jonathan S. Shapiro wrote:

>On Sat, 2007-02-10 at 02:38 +0100, Valerio Bellizzomi wrote:
>> On 10/02/2007, at 0.44, Pierre THIERRY wrote:
>> 
>> >Scribit Valerio Bellizzomi dies 09/02/2007 hora 01:23:
>> >> So as you say, it is not necessary to record reboot time. I still
have
>> >> some confusion about how to record the "OS install time" for
>> >> administrative purposes, but the uptime question is resolved.
>> >
>> >Couldn't the process that records uptime also store it's first boot
>> >time?
>> 
>> But wait, the first boot time could not be the install time. Because
the
>> system image is installed (copied on hard disk) by hand via a tool,
then
>> you have to restart the machine to reboot from hard disk.
>> The OS install time is at the end of the system image copy operation,
the
>> system is not running at this time, the installation tool is running.
It
>> is not told that you reboot immediately. Or I am probably missing
>> something important ?
>> 
>> val
>
>You are missing the fact that we cannot possibly be interested in the
>install time if the system is never booted. Firstboot time is
>sufficient.

It may be sufficient but it is not precise. If for example, I install the
machine on Friday and boot it on Monday, well first boot time can be used
but it will not tell the truth. Because the machine was off during the
weekend, technically we don't care, but administratively it makes some
difference.
So, I propose that the installation tool simply writes down on disk the
time/date of copy completion in some defined disk area, this would be the
precise "image write" time.

val