Could the Pyra be cheaper and available for *BSD ?

[openbsd98324]

Hello,

I consider to order a mini arm machine as a notebook replacement. Very light and thin: pinebook64. It is very affordable and sufficiently powerful.
It would replace my notebook, getting older.

However Pyra is interesting project, keep going !
Could the device Pyra be slightly cheaper?

Concerning NetBSD, there is into the kernel arm*/aarch64 so support yet. However, since Pyra has similar hardware as few ARMs, it would be possible.
Eventually the WIFI won't work, but we can use the second USB for a classic WIFI Dongle, urtwn0.

Kind regards,
James

--
FLDE Desktop for Pyra? [attached]

 

[levi]

Linux is free to use as well, and internal wifi is supported with it already. I wouldn't be against porting *BSD to the Pyra because it's interesting software, but I don't see how it could make the product any cheaper, especially because ED would either have to supply it with a compatible wifi stick, or pay someone to fix the internal wifi for him.

 

[openbsd98324]

Linux is free to use as well, and internal wifi is supported with it already. I wouldn't be against porting *BSD to the Pyra because it's interesting software, but I don't see how it could make the product any cheaper, especially because ED would either have to supply it with a compatible wifi stick, or pay someone to fix the internal wifi for him.

I think that this image is very very likely to work on it, however how to boot it???

Cheaper is kinda interesting. Rather big question, how can the pinebook64 be somehow significantly cheaper.

 

[levi]

The pinebook64 I'd expect is entirely made and manufactured in China or Taiwan.

 

[openbsd98324]

The pinebook64 I'd expect is entirely made and manufactured in China or Taiwan.

unfortunately, this is what we have expected. Low salaries and job conditions.

Well, EU was about 20 years ago, aware that the delocalization of industry to Asia incl. China was necessity. We see the result today.

 

[TeDaDeS]

The Pyra is made in Europe, it's has a low production volume and it's a complicated niche product. The initial investment cost are really high, I think creating the moulds for the cases already cost 50k (that's already 50 euro per unit if you produce 1000 units, and you didn't produce any case at that point). Then you have all the actual parts, which are purchases at lower volumes. The labor costs like: designing, shipping, assembling, testing the units. And covering the risk of all additional unknown costs, like replacing less-then-perfect cases or shifting suppliers etc. Which makes a product like the Pyra much more expensive compared to most mass produced products.
And some costs are compensated by ED already, by his shop earnings. So it wouldn't be fair to lower the price.

 

[cerno]

Linux is free to use as well, and internal wifi is supported with it already. I wouldn't be against porting *BSD to the Pyra because it's interesting software, but I don't see how it could make the product any cheaper, especially because ED would either have to supply it with a compatible wifi stick, or pay someone to fix the internal wifi for him.

Linux as a OS for game console emulator is perfectly fine for me. However, when I am going to use the Pyra as a PDA and small form factor portable computer, I have some issues with Linux in general, and systemd, pulseaudio, wayland, etc. in particular. To at least rudimentarilly support porting of NetBSD to the Pyra, I am prepared to pay for a developers machine (or, if things will go well on the financial front, two; one for development, the other for QA and production). Unfortunately, I can not pay thousands and thousands of $ for a commercial developer (I wish I could).

As a sidenote, I'm a Pinebook Pro owner. And I have yet to boot NetBSD on that machine.

 

[netcat]

Almost 20 years ago I had a friend who was obsessed with NetBSD and got it running on his Zaurus. To him it was a technical challenge, no practical element just more secure in irrelevant ways. Not sure if he ever booted X11.

Actually I ditched Linux for FreeBSD on my desktop around that time too. Booted up a bit quicker, had a cleaner package manager, kernel compiled in 45 seconds. Again, it was nice technical challenge but impractical in the end due to lack of binaries and I had to go with the most generic and old hardware.

Don't get me wrong I love BSD, Jimi Hendrix live at the Berkeley Greek Amphitheater and Bill Joy if my favorite computer scientist. But Linux is getting better after all these years with stuff like BPF, which I find really cool.

My point being I'd suggest using Linux, unless you want to be even more of an outcast. Linux is the most successful* software of all time.

Also: you underestimate development costs by a large factor.

 

[openbsd98324]

I am using a RPI3 and RPI4. Glad that there is not Autoboot.txt ! really.
u-boot on SD/MMC is very cool.

A possible solution to reduce the cost would be to suppress the NAND "Harddisk". Anyhow we dont need the nand, since we use the SD/MMC all the time.

+1: it would be lighter and thinner. Nice for pockets and carrying.

(20 years, desktop was possible, like today... have a look.).

 

[netcat]

we like retro games not retro browsers.

 

[TeDaDeS]

A possible solution to reduce the cost would be to suppress the NAND "Harddisk". Anyhow we dont need the nand, since we use the SD/MMC all the time.

Removing the NAND makes it cheaper, but not a lot. A 32GB eMMC is about 25 euro's. So it would be cheaper, but not by much.
You also don't want to many variants. There are already 'cheaper' ones that have less memory and no modem.

 

[levi]

I think the ones with 2GB of RAM have been nixed, leaving the only variants the ones with different or no modem.

 

[cerno]

Almost 20 years ago I had a friend who was obsessed with NetBSD and got it running on his Zaurus. To him it was a technical challenge, no practical element just more secure in irrelevant ways. Not sure if he ever booted X11.

Actually I ditched Linux for FreeBSD on my desktop around that time too. Booted up a bit quicker, had a cleaner package manager, kernel compiled in 45 seconds. Again, it was nice technical challenge but impractical in the end due to lack of binaries and I had to go with the most generic and old hardware.

Don't get me wrong I love BSD, Jimi Hendrix live at the Berkeley Greek Amphitheater and Bill Joy if my favorite computer scientist. But Linux is getting better after all these years with stuff like BPF, which I find really cool.

My point being I'd suggest using Linux, unless you want to be even more of an outcast. Linux is the most successful* software of all time.

Also: you underestimate development costs by a large factor.

... _getting better_ heavily depends on your needs and requirements. Mine are increasingly incompatible with (at least mainline) linux. I select my software based not on success rate, but fitness for my purposes.

If there will be no NetBSD port to Pyra, then Pyra will remain a toy for me. Extremely nice one (and worth buying, of course!), but for daily production machine, I would need to look at some competition. (If I will not find any suitable small form factor device, then I will carry a goddamned full-size laptop everywhere, I can live with it.)

...I know the rates, I did a bit of custom development work (some quarter of century ago)... :-/

 

[levi]

I think ED's shop sells a number of x86 handhelds without the windows tax. Might be simpler to get NetBSD on.

 

[Git1]

Could the device Pyra be slightly cheaper?

A possible solution to reduce the cost would be

There's no reducing the price; ED is already making a huge loss, he's using funds from his other ventures to fund development. The only possibility is reducing ED's loss, and I imagine he is already doing his darnedest to do that.