[Libre-soc-dev] Libre-soc-dev Digest, Vol 39, Issue 1

[Sable]

On 08/17/2023 10:08 PM, Luke Kenneth Casson Leighton wrote:

>
>
> On Thursday, August 17, 2023, Sable <Sable at disroot.org
> <mailto:Sable at disroot.org>> wrote:
> >
> > Interesting... wonder if they have a repo that shows their progress.
>
> https://libre-soc.org/task_db/report/
> https://bugs.libre-soc.org
>
> if you mean "RED Semiconductor Ltd"? no absolutely not,
> that is a commercially confidential venture that has to
> respect Foundry NDAs, which Libre-SOC *by definition* of
> being a Libre Project can under no circumstances sign.
Makes sense, I wonder what made them choose to go this path though. So
many other companies usually wouldn't side with libre developers like
yourself due to the gpl3 license if nothing else. My ignorance might be
showing though on this.
>
>
> > My bad, yeah libre. Sorry about that. My memory must be hazy. xD
> >
> > Sometimes admittedly, I talk to certain people about this and
> unfortunately, there is no other way to talk to them about distros,
> etc... for them to understand what I mean, etc...
>
> education. microsoft has an "Open" License for Windows NT Source
> code. it reads, "pay us USD 1 million a year and we will OPEN
> the source code to you. do anything with it though and we will
> sue the f*** out of you".
>
> ... but it's "Open", right?
>
My definition of open means that it could be considered libre although
it can be used for any purpose, even proprietary. But that being said,
not everyone follows that thinking, so good point.
> > The second one especially has my interest. Would love to have a
> netbook with a 24 hour battery life especially  under high load.
>
> about USD 15 million will do it.  USD 10m for the SoC,
> USB 0.5m for the laptop hardware (and component sourcing
> which is usually 12 months on its own), and about USD 4m
> for the software team including compilers, SFFS distros, and
> upstreaming which will take appx 2-4 years.
>
That is quite a bit of money, hope you have enough funding for your goals.
>
> > Also if possible being able to run qemu with other architectures
> like x86_64 or x86 being included in
> > that and be semi reasonable speed.
>
> needs JIT binary emulation/translation just like
> China ICT did with Loongson around 2018 (MIPS64 to x86)
> and DEC did back in 1991 (Alpha to x86), both got appx
> 70% of native hardware speed.
>
> otherwise the penalty is appx 1000x slowdown.
Yeah, I have seen this slowdown, before even on x86 64 bit hardware.
Without KVM anyhow, its a mess.
>
> > A lot of this probably is dream-like for a while even in the future.
>
> just money to make it happen, sufficient to get the right
> competent people at Industry-standard rates.
>
> there is a bootstrap process if that does not happen but
> it is a much slower route.
>
> basically if you want to see your dream come true faster,
> you need to take responsibility to help find the people to
> make it happen.
>
> l.
>
>
Not to mention time and energy as well in addition to money. So it is
actually possible, just depends on the effort that is put into it. Hm...
point taken.

Well, in any case, I sent you a separate email that you might have
missed, I think you will like the link on it. It shows a good example of
the problems of top posting. :D

In a funny way I might add.
>
>
> -- 
> ---
> crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
>