Philipp Blum <philipp-b...@jakiku.de> wrote:
> That is actually a quite interesting option to evaluate. There are
already a
> couple of projects for that:
> https://github.com/takahirox/riscv-rust
> https://github.com/d0iasm/rvemu
> Of course this will be very inefficient, but this wouldn't be a productive
> dev environment. Just for people playing with RIOT OS. For that purpose it
> should be good enough.
>>
>> btw.: I am currently working on running WASM inside RIOT.
> I don't get why you should want to run wasm inside RIOT, when you can just
> compile the source code for the desired architecture. What is the use-case
> for this?
It means that we can have a sensor/device with all sorts of nice onboarding,
security, and sane stack, which is upgraded/patched on a long-cycle.
But the actual application code can be downloaded on a much shorter cycle,
and can be authored by far less capable (and far less trusted) people.
[See RUST talk at last summit]
Could also have a different license.
Whether this will really work given the battery optimization concerns is
something we'll have to learn.
--
] Never tell me the odds! | ipv6 mesh networks [
] Michael Richardson, Sandelman Software Works | IoT architect [
] m...@sandelman.ca http://www.sandelman.ca/ | ruby on rails [
signature.asc
Description: PGP signature
_______________________________________________ users mailing list users@riot-os.org https://lists.riot-os.org/mailman/listinfo/users
