Draft laptop testing spec

[Matthew Garrett]

On Sat, 2005-06-04 at 14:42 -0500, Carl Karsten wrote:

> 1. ability to detect events (can we tell when X has happend? button, switch, AC, 
> docked)

Any hardware which doesn't provide these events is so indescribably
broken that we honestly don't care too much at that stage...

> 2. ability to transision between various states: (sleep, hibernate, on) 
> (docked/undoced)  (removable hardware in out)

Indeed.

> I think there should be a status app that shows when the #1 events happen.  That 
> way we can have a better idea why the box didn't hibernate when we pushed the button.

You can get that at the moment by stopping acpid and then doing

tail /proc/acpi/event

. There's no real reason to write a graphical app for debug purposes.

> What should lan and wifi do?  espicaly when the box wakes up in a different state 
> (different wifi AP, network cable unplugged)

See http://udu.wiki.ubuntu.com/NetworkMagic

> Also, box put to sleep, pcmica cards changed, box wakes up.

That /should/ already be handled.

-- 
Matthew Garrett | mjg59 at srcf.ucam.org