None of the devices you mentioned were polished physical products competitive with their contemporaries, while also costing significantly more.
What the community needs is a compelling physical piece of hardware with first-class upstream driver support for everything and relatively modern specs.
We can iterate on the userspace as a community. But there won't be a growing community if you don't sell us a compelling open piece of hardware to run it on. We can't easily iterate on the hardware as a community, this is the part puri.sm should focus on delivering.
Sell us a modern device that respects our privacy and ability to replace boot loaders and run mainline kernels. Give use first-class driver support for everything included from the GPU to the cellular modem and cameras and all the APIs necessary to speak with them.
It just seems like a total disconnect from the reality of mainline kernel support for smartphone hardware, to be off in la-la land working on a GNOME phone OS before delivery. Pavel Machek's talk from 10/22/2017 [1] comes to mind as representative of the state things are in, in mainline.
I'm far more confident in the GNOME/free software community's ability to write dialers, messengers, email and photo apps than its ability to quickly bring up low-level support for such hardware and develop the new kernel APIs where things like V4L2 fall short.
"None of the devices you mentioned were polished physical products competitive with their contemporaries, while also costing significantly more."
I would actually argue the Open Pandora was a polished physical product. It worked incredibly well when it just came out.
"What the community needs is a compelling physical piece of hardware with first-class upstream driver support for everything and relatively modern specs."
Correct me if I am wrong, but aren't the Librem 5 folks trying to do just that? They chose an imx8 because it has really good upstream support, and then they can add upstream support for the peripherals.
What the community needs is a compelling physical piece of hardware with first-class upstream driver support for everything and relatively modern specs.
We can iterate on the userspace as a community. But there won't be a growing community if you don't sell us a compelling open piece of hardware to run it on. We can't easily iterate on the hardware as a community, this is the part puri.sm should focus on delivering.
Sell us a modern device that respects our privacy and ability to replace boot loaders and run mainline kernels. Give use first-class driver support for everything included from the GPU to the cellular modem and cameras and all the APIs necessary to speak with them.
It just seems like a total disconnect from the reality of mainline kernel support for smartphone hardware, to be off in la-la land working on a GNOME phone OS before delivery. Pavel Machek's talk from 10/22/2017 [1] comes to mind as representative of the state things are in, in mainline.
I'm far more confident in the GNOME/free software community's ability to write dialers, messengers, email and photo apps than its ability to quickly bring up low-level support for such hardware and develop the new kernel APIs where things like V4L2 fall short.
[1] https://lkml.org/lkml/2017/10/21/181