Re: A few questions: Libre SoC, website, Rust

[Joshua Branson]

Jan Wielkiewicz <tona_kosmicznego_smiecia@interia.pl> writes:

Hey Jan,

I'm an occasional Hurd web contribute (I haven't contributed anything
useful in a while).  I'm not really a Hurd developer, but if I can
motivate/encourage you to achieve your goal, please let me know!

> Hello,
>
> I'm not sure if you're aware of this, but there's a compeletely libre
> CPU+GPU OpenPOWER chip in development and I think supporting it in
> the Hurd will be crucial for free software.
> https://libre-soc.org/
> Full source code is available, drivers and firmware will be also under
> a free license.
> The OpenPOWER foundation released a libre-friendly EULA this February:
> https://openpowerfoundation.org/final-draft-of-the-power-isa-eula-released/
> The SoC is going to be 64-bit, mobile-class chip with 3D acceleration.
> The developers are open to contributors and I think it is could be
> possible to get a grant from NLNet foundation for porting the Hurd to
> PowerPC (or whatever the architecture actually is).
> Here's the progress of development:
> https://libre-soc.org/3d_gpu/

I wonder if you know about this open 3D GPU:
https://www.crowdsupply.com/libre-risc-v/m-class

I'll email the leader the libre-risc-v and let him know about the
libre-soc.org website.  Those guys should work together!

> My objective is: x86 processors are really hostile towards user freedom
> due to backdoors like Intel ME or AMD PSP, the ISA itself is
> proprietary, that's why porting the Hurd to x86_64 is even less
> important than porting it to PowerPC.
> If everything goes well, OpenPOWER processors will gain in popularity
> and the Hurd together with it.

I'm fairly certain that the Hurd developers would agree with you in
this.  I believe that they would love it for the Hurd to work on
numerous architectures.  I believe that the Hurd's glibc's port to
OpenPower, will need to be completed.  For the more technical porting
details, you'll have to ask the other Hurd devs.  I believe that this is
quite an involved task.  :)

>
> My second question is:
> Do you have anything against if I and my friend modernized the website?
> It looks like abandonware and is really harmful for the project.
> The main page would be dedicated to promoting the project and would be
> graphically appealing to convince the visitor the project is not dead.
> The current page could be moved to a wiki section or somewhere else.
> Also navigation is too complicated and messy, searching doesn't work at
> all, because https://darnassus.sceen.net/cgi-bin/hurd-web is dead half
> of the time.
> Any special wishes?

Please do!  We would love it if the Hurd looked "hipster" and fun!  The
easiest thing you could do would be to tweak the css files, but I
believe that Samuel would be open to talks to move from ikiwiki to some
other wiki you might prefer.  Please send any patches to the website, to
bug-hurd@gnu.org.  We haven't used web-hurd@gnu.org in a while.

> Third question:
> Do you have anything against Rust contributions into the project? My
> friend is interested in contributing, but unfortunately in Rust, not C.
> I wonder if rump drivers could be written in Rust, thanks
> to the Hurd's modular architecture.

Again, I'm speculating here, but Samuel might be ok with this.  I don't
know how well rust support is in the Hurd.  You might have to port rust
to the Hurd, which is non-trivial.   I think we are still working on
getting Go to work on the Hurd.  :)

> Jan Wielkiewicz
>

--
Joshua Branson
Sent from Emacs and Gnus