org.freedesktop.PowerManagement

[David Zeuthen]

On Thu, 2007-03-29 at 22:10 +0200, Danny Kukawka wrote:
> > Danny, I think if there are systems where standby is actually working and
> > suspend to ram is not, Suspend() should do S1 behind the back of all
> > applications. And if there are systems where both S1 and S3 are working,
> > S3 should be preferred anyway. ACPI standby is really dead these days.
> 
> This is maybe a possible solution, but the user want to know if he suspend to 
> ram (S3) or if he go to Standby (S1) if he call something in the 
> powermanagement application. At least there are some differences between S1 
> and S3. Power consumption is only one.
> 
> You can't say in KPwersave or g-p-m "go to suspend2ram" and call standby in 
> the background. The user has to know what happen. At least the user which 
> until now used standby and not s2ram (because he know it would fail) would 
> never use s2ram or you lead the user to believe that his machine work now 
> with s2ram but this is not true.
> 
> IMO we should allow Standby as an optional method and we should have a way to 
> differ between them. This would also only effect older machines, actual 
> laptops don't support S1 anymore.

Keep in mind that org.fd.PM is an interface for _applications_ not for
PM daemons; applications just don't care what the mechanism is. They
just want the machine to go to sleep, they really really don't care
about whether it's S1 or S3 or how much power is saved. 

(Applications _might_ care about difference between Hibernate() and
Suspend() given that the former introduces a higher latency. In fact, I
can argue that the org.fd.PM should only offer a Sleep() method to
prevent applications authors having headaches about whether they should
choose Suspend() or Hibernate().)

Again, the golden rule here is to figure out the needs of the
_application_ - the user is already served well by the UI present in the
PM daemon implementing this interface. As such, the PM daemon can care
as much as it wants about S1 vs. S3; I don't see a problem with the PM
daemon offering 

 [*] Use ACPI S3 for Suspend
 [ ] Use ACPI S1 for Suspend

as an option in it's configuration interface. You should be free to do
that in your implementation. The org.fd.PM specification does not
prevent you from such an UI and it shouldn't. This specification should
care about the needs of applications.

(I don't, however, such UI is good UI nor do I think it belongs in
gnome-power-manager. But that's up to Richard and the GNOME project I
suppose.)

     David