[Automated-testing] Farming together - areas of collobration

[Carlos Hernandez]

On 02/13/2018 10:20 AM, Matthew McClintock wrote:
> On Mon, Feb 12, 2018 at 3:51 PM,  <Tim.Bird at sony.com> wrote:
>> One thing I've observed is that there are substantial differences
>> in the level of abstraction and the board management architecture
>> for these different systems.
> I've always thought step one is creating an abstraction to describe a
> board farm setup. Then the various tools can link into that to run
> what's required. It's also easier to add support for standalone
> components when an external tool can just request generic support for
> something like a power relay, network switch, etc. Whichever tool(s)
> does useful things on top of that will win, and you can (more) easily
> move from one tool to another, and all components and small and spread
> out.
>
> Not sure if something like that exists exactly already... it would be
> nice to share a small test that runs on my infra at home (smallish) or
> the lab at work (larger) and everyone else's lab across the world that
> uses the library that describes how your board is setup, what devices
> are present, topology, etc.
>
> -M
>

We need to:
1) Understand different approaches people/projects have taken to 
abstract HW control
2) Determine what's constant, deltas, good ideas, etc.
3) Propose High Level API. The API needs to cover not only actions but 
also how to describe HW setups.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.yoctoproject.org/pipermail/automated-testing/attachments/20180213/fac17bad/attachment.html>