Unsure about cloning vs compiling

[Nard SDK mailing list]

Hi Greg
Yes, I am afraid there is misunderstanding. It is not possible to download the 
pre-built binary and continue where I left of. That one is for demo only. 
Instead, for your own project, you are required to build the image yourself. 
This is a deliberate feature. This is how one best build products when you 
have 100+ devices in the field you need to maintain long term. If you have 
only one or two devices in total, then Nard is likely overkill.

Thanks
/Ronny


---------------------------------------------
> Hi Ronny,
>
> I recently learned about Nard and it seems perfect to use for my keg washer
> project.  Just to give you a quick background, I am currently using a Pi
> Compute 3+ running Rasbian to control a keg washing machine.  The Pi runs
> the physical machine via GPIO (using RPi-GPIO) and also runs an HDMI
> touchscreen using python, tkinter GUI, sqlite, and some window manager that
> I am not sure of, but is obviously included with the desktop Rasbian.  This
> machine/app does not require a network/wifi (plus the Pi Compute does not
> have onboard wifi, but I have been using a USB wifi dongle to download my
> python app onto the Pi).
>
> My goal is to be able to install Nard with my app and all the dependencies
> on multiple Pi Computes.  The first thing that confuses me, is why couldn’t
> I just start with your pre-built binary version and then install all the
> needed dependencies and my python app and then just copy the eMMC card?  I
> read that device ID stuff is kept in the settings folder, so I assume I’d
> just have to change that for each Pi Compute after cloning.  It seems
> easier than compiling my own, would that work or is there something I am
> missing?  Forgive me if I have some gross misunderstanding.  I am
> definitely interested in learning to compile but I know it could be a steep
> learning curve.
>
> Thanks,
>
> Greg