From: Alex Bennee <kernel-hacker_at_bennee.com>
Date: Mon, 10 Nov 2008 12:32:39 +0000

On Mon, Nov 10, 2008 at 12:21 PM, Jonathan Gordon <jdgordy_at_gmail.com> wrote:
> 2008/11/10 Alex Bennee <kernel-hacker_at_bennee.com>:
>> On Mon, Nov 10, 2008 at 11:58 AM, Daniel Stenberg <daniel_at_haxx.se> wrote:
>>> On Mon, 10 Nov 2008, Alex Bennee wrote:
>>>
>>>> Basically I'm looking for the easiest way to test new firmwares without
>>>> potentially killing my working install.
>>>
>>> Then backup/rename your existing .rockbox first, then unpack your
>>> experimental one on the target.
>>
>> I'm curious now as to the point of being able to run/boot arbitrary
>> .ipod files from the existing instance if the resultant firmware will
>> never work properly because it's missing the rest of the .rockbox file
>> hierarchy? Is there a tight coupling between the firmware and the
>> contents of .rockbox?
>
> .rockbox is hardcoded in both the bootloader and the .ipod... so yes..
> its pretty tight

Ahh I see. In theory if I change the .ipod firmware to see
".rockbox.devel" as it's root dir (via configure) I could happily
build firmwares that run in a different space as long as I boot my
stable firmware first to ROLO load the devel firmware.

I'll have a dig through the build system and see how feasible such a
change would be. I think it would make testing new firmwares easier.

-- 
Alex, homepage: http://www.bennee.com/~alex/

Received on 2008-11-10