From: Linus Nielsen Feltzing <linus_at_haxx.se>
Date: Wed, 03 Dec 2003 00:02:00 +0100

Björn Stenberg wrote:
> It's not a solution, it just replaces one problem with another.
> Solving empty-battery-boot by killing power-triggered-boot is not
> acceptable. We've had this discussion already:
> http://rockbox.haxx.se/mail/archive/rockbox-archive-2003-10/0514.shtml

First of all, this is only applicable when you have flashed Rockbox.
Let's say that you have.

How will Rockbox know that it shouldn't try to spin up the disk? If it
tries to, it may crash. I know this is a rare case, but it is a very
important one, since you won't be able to charge it if Rockbox keeps
crashing on you.

>> I repeat: when you insert the charger when the device is off, you
>> end up in the charging mode of the Archos firmware, like it or not.
>>
> Not on my unit, you don't. You end up in Rockbox. Currently we stop
> in a charging screen instead of doing a full bootup, but that will
> hopefully be changed soon.

I am not talking about flashed Rockbox here. I am talking about running
Rockbox with a .ajz file, like most regular Rockbox users.

> Those not running Rockbox in flash are not affected by Rockbox' boot
> behaviour, so they are not a party to that issue.

You were the one who brought the Rockbox boot issue up, I was talking
about the behaviour of a non-flashed device, where you will end up in
the ROM charging screen when you plug in the charger. From there you
press ON to start Rockbox. And then I think it may be natural to return
to that mode by holding OFF.

We can't expect everyone to put Rockbox in flash. Even if we did, we
would have to implement an intermediate state, where we can't spin up
the disk due to low power, even if the charger is connected. Call it a
Charging State or whatever.

And remember, we were talking about the OFF/STOP behaviour, not the boot
procedure.

Linus
Received on 2003-12-03