RPi3 are reasonably capable for the price.  For me, they make sense because:

* RPis make it easy to try non-windows OS (including Plan 9).
* Provide a usable, yet inexpensive ARM platform for Plan9.
* (almost) all RPI hardware components are supported in Plan 9.
* There is an enthusiastic community building everything imaginable for,
and with, RPI's.

RPi's aren't "the" answer, but neither is Intel-inside everything.  The
speculative execution debacle proves that the entire industry has too much
reliance on one architecture. Diversity of architectures is good for Plan 9
and the industry as a whole.

In a perfect world there would be equivalent popular platforms for MIPS,
Power, RISC-V and other architectures.

-Skip

On Sun, Feb 4, 2018 at 1:46 AM Ethan Grammatikidis <eekee57@fastmail.fm>
wrote:

> On Sat, Feb 3, 2018, at 11:46 PM, Bakul Shah wrote:
> >
> > Not to mention The RasPis are poor at
> > reliability.  Even a xenon flash or near a RasPi could power a
> > RasPi2 down! And since they do no onboard power regulation,
> > people had lots of problems early on -- add one more USB
> > device and the thing can become unreliable.
>
> This is probably an impossible question, but I've got to ask: Why do
> people even buy RasPis? Like, for anything? Even when the first RPi was
> new, a second hand laptop could offer far more processing power and
> reliability for the same price, sometimes excepting the disk of course. Add
> a base station with the old printer port and there's some GPIO; not as much
> as a RPi, it's true, but there are ways around that. One alternative for
> GPIO is the actually cheap boards from Ti or whoever which exist to
> interface Ethernet, WiFi, Bluetooth, or USB on one side (depending on the
> board) to GPIO and serial on the other. I think they're programmed in
> Forth, but I wouldn't be surprised if you can just download programs for
> them to do anything you'd want with remote control.
>
> --
> The lyf so short, the craft so long to lerne. -- Chaucer
>
>