Re: OT: Linix goes politics

On a sunny day (Sun, 27 Oct 2024 11:19:08 -0700) it happened john larkin
<JL@gct.com> wrote in <8r0thjl4rq4abu9g15g0j95vo10q8d9bgc@4ax.com>:

On Sun, 27 Oct 2024 12:08:29 -0000 (UTC), antispam@fricas.org (Waldek
Hebisch) wrote:
>

Don Y <blockedofcourse@foo.invalid> wrote:

On 10/26/2024 3:58 PM, Waldek Hebisch wrote:

Pi Pico is more powerful than the onboard computer.  But there were
also mainframes in ground support center.  Pico can perform more integer
instructions per second than those mainframes, but has less memory.
And mainframes had fast mass storage (drums and a disk farm).
Raspberry PI version 1 has more memory and SD-card has more bandwidth
than several mainframe disks.  OTOH I would avoid SD-card in mission
critical operations, so probably two Raspberries (two for reliablilty,
ground support mainframes also run in redundant configuration) with
external USB SSD discs...

 
How long do you think any of that would RELIABLY operate in a
supraterrestrial environment?

>
I do not know.  Standard commercial MCU-s are used in many small
experimental low-orbit statelites, and seem to work, so they may
be enough.  Certainly before lanching people into space one needs
to spend time to test and select stuff that will work well there.

>
Cubesats only last a few years, so commercial parts are often OK.
>
I have a friend in the small sat business. I'll ask him.
>
People last a lot shorter time in orbit than uPs.

 https://www.mobilityengineeringtech.com/component/content/article/33364-single-event-effects-in-high-altitude-aerospace-sensor-applications
One way in airplane electronics to counter radiation errors
is to use more than one system
and have it do a vote for the best results.
With just a few dollar each you can bring many raspberries...
and then look for the ones with reasonable output, do a majority vote.