:: [DNG] ..another viable basis idea f…
Top Page
Delete this message
Reply to this message
Author: Arnt Karlsen
Date:  
To: dng
Old-Topics: [DNG] Announce: FlyingTux project
New-Topics: Re: [DNG] ..another viable basis idea for Devuan as a hypervisor?, was: Announce: FlyingTux project
Subject: [DNG] ..another viable basis idea for Devuan as a hypervisor?, was: Announce: FlyingTux project
On Fri, 30 Jul 2021 14:49:32 +0200, metux IT consult wrote in message
<d0ccdb85-2d0a-27e8-dc66-11f7bc00cd9e@???>:

> Hello folks,
>
>
> maybe a bit offtopic,


..I disagree. ;o)

> but allow me to announce the FlyingTux project:


..welcome onboard DNG. :o)

> It's an build/runtime infrastructure for running desktop and mobile
> applications in containers and build an entirely container-based
> mobile OS based on it.


..I like the https://www.qubes-os.org/ approach, but not
their use of systemd nor of rpm.
Do we do something similar here?

> The primary motivation is my long frustration about the monstreaus and
> practically unmaintainable Android, which also still lacks lots of
> common management abilities we know from the GNU/Linux world.


.."wonder why Google designed it that way." ;oD

> In some ways, FT can be seen as an conceptional combination of
> containers (docker, k8s, etc) and apps (android, etc). One major
> difference is that also the app images are based on some defined
> distro base (for start, just alpine, others to follow later) and the
> images are created on the host, based on host specific settings like
> hw setups (eg. automatically deploys the right mesa drivers). In
> future steps some packages of the app distro base (called 'osbase#
> here) will be replaced or customized, in order to provide better
> integration with the ecosystem and strip unneeded stuff.


..which means an early step will be trim down the basic (net-install?)
images we have, as far down as possible, and build app, vm etc images
upon those stripped down base images.
I guess we also want a bare bare metal hypervisor of some sort. ;o)

> Another key difference is moving common functionality (eg. various
> data sources, communication protocols, ...) out of the individual
> apps into generic services - and the binding between individual apps
> and actual services instances can be customized by the user (e.g. one
> can bind some apps to fake gps instead of the real one, separate
> address books or user directories, etc, etc).
>
> Here's a more detailed description:
>
> https://github.com/metux/flyingtux/blob/master/README


..this means we can set up e.g. chromium in throw-away-after-use
containers? Or stick new stuff with lib conflicts in containers?
Etc?

> Note that for now its very experimental and fast changing. Don't
> expect anything field-ready yet. But it's already good enought to
> isolate some common desktop apps like gimp, chrome, etc.


..I see no .deb package? Pretty soon we'll need debug packages
to shake out bugs.

--
..med vennlig hilsen = with Kind Regards from Arnt Karlsen
...with a number of polar bear hunters in his ancestry...
Scenarios always come in sets of three:
best case, worst case, and just in case.