torios-dev team mailing list archive

[Israel <israeldahl@xxxxxxxxx>]

On 11/16/2016 04:57 AM, Nio Wiklund wrote:
> Hi Israel and the ToriOS developers,
>
> I re-discovered a bug, which is independent of guidus. It has been
> around for a long time also with mkusb (without fixing).
>
> The persistent live system of ToriOS 1.0 has a different directory
> structure for booting compared to the live system, and the tarball is
> not where the OBI-installer expects it to be. It can be found and
> selected, and used. I did that in order to create the installed ToriOS
> system, where I tested that guidus works.
>
> In a normal persistent live system:
>
> /lib/live/mount/persistence/sda4/live/tarballs/
>
> In a persistent live system with ramdrive, boot option 'toram':
>
> /lib/live/mount/medium/live/tarballs/
>
> -o-
>
> I remember that you asked me, Israel, a long time ago, why I located
> the tarball outside the root partition (and located it hardcoded into
> the drive). I said that we could avoid 'double passed through
> compression', which is good, but it does cause trouble in the
> persistent live system.
>
> a. Do you think we should move the tarball into the root partition
> (and the squashfs package)? It would remove this problem.
>
> b. Or do you think I should fix it with mkusb/dus?
>
> c. Or do you think it is a minor issue, and we should focus on other
> and more urgent problems? (In other words, do you think anybody except
> me has tried to make a persistent live drive with ToriOS, and if
> someone has, did that person intend to use that system to install
> ToriOS?)
>
> Best regards
> Nio

Nio,

We can fix this in OBI.

OBI looks for the tarball, and picks the most logical place if there is
not tarball in the 'selected' file that is legitimate.

I will add this path to OBI later on.  I do not think it will be a big
issue.  I'll let you know when it is updated for testing

-- 
Regards