--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

On Mon, May 29, 2017 at 10:12 PM, David Niklas doark@mail.com wrote:

On Mon, 8 May 2017 05:45:36 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:

...

---

crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

On Sun, May 7, 2017 at 10:10 PM, zap zapper@openmailbox.org wrote:

...

On 05/07/2017 04:29 PM, ronwirring@Safe-mail.net wrote:

...

All software for the mali-t860 is open source?

none. MALI is proprietary.

I'm confused. Luke, if you plan on making an RK3399 into an eoma project how can you get RYF status if the mali GPU is closed source?

this was discussed a year ago or so. same process as for the EOMA68-A20

For that matter, how can you get RYF cert. for your current eoma68 project?

by leaving out the proprietary crap, simple as that. see below.

Unless I'm mistaken and it uses a different GPU?

it is actually a different GPU but that does not change the assessment process carried out by the FSF.

Or you just leave the HW crippled?

if the FSF considered the device to be "crippled" by it not having the 3D engine running, such that there was a genuine risk that people would actively seek out the installation of proprietary software.

in the case of e.g. a proprietary on-board WIFI device that *would* constitute a genuine risk of people *actively* seeking out proprietary firmware, and consequently the FSF quite naturally refuses to certify devices that contain non-removable proprietary on-board WIFI chips.

however in this case it actually turns out that if you use the proprietary 3D GPU for the tasks that i suspect you *believe* will quotes accelerate quotes certain operations (such as X11), the MALI embedded GPU (or its associated proprietary software - we can't actually tell which because we DON'T HAVE THE DAMN SOURCE) is so piss-poor at its job that it actually SLOWS DOWN CERTAIN OPERATIONS of X11.

given that 2D acceleration is already covered by fbturbo, and works really well *and is entirely libre software*, the *need* for the 3D engine just for basic Small-Office / Home-Office and day-to-day usage is NOT A CONCERN.

so does that make it clear that the evaluation process (which was described a year ago) is not just a hard-and-fixed process?

now, if on the other hand this was a dedicated Games Console product, *that would be an entirely different matter*. applying for RYF Certification on a 3D Games Console product which has a 3D GPU which *only works with proprietary software* would probably constitute too much of a risk that buyers *WOULD* in fact go out of their way to download the proprietary drivers.

but this design *isn't being sold as a 3D Games Console*, is it?

does that help clarify?

l.