[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: On persistency in newer live-boot

04/08/2012 06:26 PM, Daniel Baumann:
> On 04/08/2012 01:19 PM, Daniel Baumann wrote:
>>   overlay-full, for full persistent overlays
>>   overlay-custom, for custom overlays
>>   snapshot-full, for full system snapshots
>>   snapshot-custom, for custom snapshots (to be implemented)
> actually, i've though a bit more about it (sorry for the mess)..
> ..why care about full and custom persistency at all? full persistency is
> just a special case of custom persistency, where '/ bind' in
> live.persist is used (which currently is invalid in order to not
> interfere with full-ov). wouldn't that be simpler/nicer?

Implemented in commit f92f379, pushed to debian-next.

I actually thought about this early on during my work but decided not to
because there was so many special cases to handle. As time passed it
seems I have had to deal with most of them for other reasons, so now it
was pretty straight-forward to implement this.

> the only drawbacks qi can see are:
>   * it's kind of ugly to always have a /live.persist in the root of the
>     overlay partition. that could be workarounded by allowing/requireing
>     to use /.live.persist instead (which works on fat/ntfs too, not just
>     on the unix/unix-like fs'es).

An alternative solution (which currently works) is for the user to use
an option like: source=some_subdir

>   * in the full persistency case, a simple mkfs -L full-ov would not
>     be enough, it would be a two step process to also create the config
>     file. personally, that wouldn't bother me.

Well, the same already applies for /home.

> if we'd do that, we could, fs label wise, get away with 'overlay' and
> 'snapshot', both being below 11 characters, so no fallbacks for legacy
> fs'es/os'es would be needed.

Perhaps 'live' is good enough as a new label? I leave the naming to you.


Attachment: signature.asc
Description: OpenPGP digital signature

Reply to: