Finally, a bit more news!


Maybe consider another round of crowdfunding? (not for more pre-orders since you mentioned stopping those; just for us to fund finishing our orders). If you can give us a reasonable goal to meet I think many of us waiting would be willing to pitch in (especially if one of the rewards is moving up in line AND getting a solid shipping date :oops:)
 
Last edited:
My personal perspective to raise capital here is to crowdsource a new CPU to avoid the whole software snafu altogether and drum up excitement for a modern SoC platform and more open OS compatibility.
The Ti OMAP is dead as dead can be, it has issues, and even once we work out the kinks it's so far behind now that even doing remedial tasks is asking for a lot.

New Backers pay for the new CPU development + Margin + Price of BOM. But get a new CPU, while bankrolling the Pyra shipments.
Those of us with Pyras get to experiment and develop on this new SoC to get it in a good state. Meanwhile Pyras keep shipping to build userbase. Win-Win.

Another idea is limited edition things. Pyra is a novelty item (quite honestly for niche collectors with money to spend in many aspects), I think $1,000 for a 1:100 early prototype v2 CPU would be popular amongst the existing owners.
I'd personally pay $2,000 for a Modern SoC ARMv8 Pyra that can run legit out of the box Debian (like most ARM boards can) with an OLED LCD panel (can remove touch if needed) and actual USBC.


Hate to say it, but we're going to have to think out of the box, otherwise we're looking at a slow and agonizing death spiral.
 
Last edited:
Yeah, there have been other wannabes that have come and gone.
New Backers pay for the new CPU development + Margin + Price of BOM. But get a new CPU, while bankrolling the Pyra shipments
Unfortunately, on the flip side, they'd likely start demanding other costly upgrades for their money, like USB-C, camera, thinner case, etc.

That being said, a gofundme page, or similar, with well documented milestones, and clear and stark reminders of the implications of other upgrades and the reality of what is possible, might not be such a bad idea.
 
Well, good news hearing this crazy project is alive.
I'm not billionaire (yet ?), but in case of, I won't forget you, as you don't forget the community and your line is very honest. That's rare and valuable thing.
I would be ok for a (relative small) donation, but as far as I understand, this solid community is too tiny to bring enough money as the amount is really heavy in such case.
My idea was more in an open smartphone alternative, wich could interested more people who wants non-spy-open-linux phone, and the pyra form-factor could be perfect : a mix with phone ability, umpc and gaming. No idea how it's doable to bring Pyra to 5G with a good antenna, but it's maybe a fonctionality who could start a crowfounding.

My two sences....

Keep up the good work
 
I still think it's important that the Pyra exists. It is an interesting project with alot that can be learned from.
ACK
But the long and complicated Pyra development showed that it is nearly impossible to build such a device in Europe/USA, compared to China. No matter how much money you throw into companies here to make such an device. Whatever they do there in China, they seems to have resources we can't get our hands on. Even small, unknown companies throw out handheld devices like popcorn. With or without an Indiegogo or Kickstarter campaign. (Somehow funny to see that they usually only need these crowdfunding platforms as marketing device). I think one must be as big as Valve to be able to make such a Handheld outside of Asia.
Well, it's not quite the same business model. I'm sorry if I'm being demagogical. I'm sure there are honest and dishonest people here like there, and I'm not convinced fo the advice at the end of that article that going high end brand gives peace of mind. I just think in big brands the malware must be just more professionalised and hard to detect, or even just absent until some OTA is requested by the powers that be. But we insist to use incredibly complex stuff and blindly trust a huge chain of procurement just to be cool or keep addicted to screens, so it's not easy to solve. Projects like the Pyra may not be feasible, but are near the closest one can get to how it should be (and the Pyra still uses too many blobs with too little user base for up to date software).
This is not meant to be a rant, I'm just impressed that the Pyra went that far with that limited ressources ED has and without being abandoned. Keep on going. :)
I still looking forward for a new CPU Board, maybe its then fast enough so i dont need this Filming Setup anymore whit Camera aimed on Pyra Screen so i can run a Screen Capture Software, made allready a test but it was too slow then for Emulators..

The problem is that it's becoming increasingly hard to find components that can give any assurance that all software is 100% free and transparent (that can have vulnerabilities too, and bad actors can secretly exploit free software, but at least it gives a clear selling point of being as transparent as possible). Everything requires blobs nowadays. The only option left is not using any hardware, and that is socially/economically/administratively/legally "prosecuted" too. You hav to have a mobile phone even to join the local library, nowadays. So either old hardware and fewer features or surrender are kind of the only compromise
 
My personal perspective to raise capital here is to crowdsource a new CPU to avoid the whole software snafu altogether and drum up excitement for a modern SoC platform and more open OS compatibility.
The Ti OMAP is dead as dead can be, it has issues, and even once we work out the kinks it's so far behind now that even doing remedial tasks is asking for a lot.

New Backers pay for the new CPU development + Margin + Price of BOM. But get a new CPU, while bankrolling the Pyra shipments.
Those of us with Pyras get to experiment and develop on this new SoC to get it in a good state. Meanwhile Pyras keep shipping to build userbase. Win-Win.

Another idea is limited edition things. Pyra is a novelty item (quite honestly for niche collectors with money to spend in many aspects), I think $1,000 for a 1:100 early prototype v2 CPU would be popular amongst the existing owners.
I'd personally pay $2,000 for a Modern SoC ARMv8 Pyra that can run legit out of the box Debian (like most ARM boards can) with an OLED LCD panel (can remove touch if needed) and actual USBC.


Hate to say it, but we're going to have to think out of the box, otherwise we're looking at a slow and agonizing death spiral.

while i would absolutely hate to abandon the OMAP5 userbase and the work that's now gone into it the past years - personally, i do rather agree with this. maybe excitement for a new hw platform is needed.

i was one of the many screwed over for the original Pandora preorder, and in the years since Pyra was announced i have put off buying any pocket-sized PC alternative (thus going without a device that would have been useful all this time). i intend to keep backing the Pyra, but i will probably go for a Steam Deck mk.2 when it's released...
 
Of course these comparisons between Pyra and "commercial" Handhelds are more or less Apples and Oranges. Different devices for different purposes and different audiences. I just wanted to point out what happened handheld-wise since the Pyra is in development, and it is quite alot within the last couple of years alone. And I'm actually glad that we have such a wide variety of (Gaming) Handhelds out there for every purpose and every price-range.
Yes, physical keyboards on these kind of devices are rare, but some have at least keyboards. For instance the well known GPD Win Max 2 (yes, it's big!) or the AYA Neo Slide or the GPD Win 4. Of course the prices for these devices are abysmal high, compared to the Steam Deck for instance.


All of them are handheld, but Pyra is real POCKET SIZE. No other mentioned device is pocket size.

Even worst: the few having keyboard (not being Pyra) have useless keyboards as device is too big to type with thumbs while you hold it with both hands. So they are handheld for gamming, but not handheld for typing.
 
while i would absolutely hate to abandon the OMAP5 userbase and the work that's now gone into it the past years - personally, i do rather agree with this. maybe excitement for a new hw platform is needed.

i was one of the many screwed over for the original Pandora preorder, and in the years since Pyra was announced i have put off buying any pocket-sized PC alternative (thus going without a device that would have been useful all this time). i intend to keep backing the Pyra, but i will probably go for a Steam Deck mk.2 when it's released...

I know we all want to constantly enable and be in an echo chamber, but there's basically no point in continuing the Pyra with the OMAP. Everything's EOL and deprecated, patching broken issues for almost 15 year old hardware is insanity.
Been in and out of Pandora/Pyra for over a decade (maybe closer to two at this point), my intentions are for the platform to succeed but basically what ED's update is paraphrasing is that the project is essentially dead (no money + operating at a loss = dead)...

There's no other option here other than to make a change, or let it go. Again, not a defeatist, but we have to get real.
Reuse the cases, and do a completely new crowdfunding for a Pyra Rebirth. Modern and fully supported SoC (graphics, audio, etc support), no crazy firmware snafu's, usability of modern arm ecosystem, etc.
Do a proper crowdsource on a major platform, raise $1-2M, perhaps hire a lean team to handle the work so ED stays as a project manager function to not get bogged down and tired of the project.
 
My personal perspective to raise capital here is to crowdsource a new CPU to avoid the whole software snafu altogether and drum up excitement for a modern SoC platform and more open OS compatibility.
The Ti OMAP is dead as dead can be, it has issues, and even once we work out the kinks it's so far behind now that even doing remedial tasks is asking for a lot.

Well, the OMAP5 is still in production, only the development itself by TI is dead. That said, it still receives updates in mainline Linux kernel - whereas with other vendors (like Rockchip, for example), there's no real support at all. Once they release the SoC, they release a kernel and that's basically it.
While the community tries to support modern kernels as well, not everything does work yet, so you would be stuck with an older kernel here as well, no difference.

I'm against crowdfunding anything until I have it ready, to be honest. It was different with the Pyra, as all the money was needed for the development, but the shop profits should be able to fund the development of a new CPU board and until this is ready and working, I wouldn't do a new crowdfunding.

Based on the history of the project, I doubt we would receive many more new backers, and I'm not a fan of getting more and more liabilities as well.
Post automatically merged:

I know we all want to constantly enable and be in an echo chamber, but there's basically no point in continuing the Pyra with the OMAP. Everything's EOL and deprecated, patching broken issues for almost 15 year old hardware is insanity.
Been in and out of Pandora/Pyra for over a decade (maybe closer to two at this point), my intentions are for the platform to succeed but basically what ED's update is paraphrasing is that the project is essentially dead (no money + operating at a loss = dead)...
Don't underestimate the OMAP5. The raw CPU power is almost on par with the current Rockchips!
But yes, the Rockchip needs less power for the same tasks, has a better GPU and supports 64bits.
But the OMAP5 is not as weak as most think.
Reuse the cases, and do a completely new crowdfunding for a Pyra Rebirth. Modern and fully supported SoC (graphics, audio, etc support), no crazy firmware snafu's, usability of modern arm ecosystem, etc.
Yes, that is basically the plan. However, finding a proper SoC for that isn't easy.
There is almost no SoC available (if there even is one!) that doesn't have firmware snafu's and are fully supported by mainline.

hns did some tests, and the Pyra can happily idle for 14 hours on battery with his setup. That's idling, without any low-power-stuff setup.
Our OS can't do that, so apparently, we have some stuff enabled that drains the battery even though it shouldn't.
The OMAP5 can do a lot more once these have been figured out, and with AESS almost working, we're on a good way.

A new SoC would probably cause the same issues once again, unless we use an outdated kernel by the manufacturer that probably uses other binary blobs as well.
 
Last edited:
Well, the OMAP5 is still in production, only the development itself by TI is dead. That said, it still receives updates in mainline Linux kernel - whereas with other vendors (like Rockchip, for example), there's no real support at all. Once they release the SoC, they release a kernel and that's basically it.
While the community tries to support modern kernels as well, not everything does work yet, so you would be stuck with an older kernel here as well, no difference.
Well, yes. Rockchip had some support in rk3399, but I don't see it so fine lately, with rk3568 or rk3588.
I wouldn't say so much as no real support, but I think they drag their feet quite a lot, yes.
I'm not sure if the current blobs can ever be circumvented by free code, or maybe there's some signature verification somewhere (not that I've heard), so other SOCs may be worse, but that's still not good.
And In any case, if the SOCs ever becomes well supported, even by the community, then the new CPU board could be designed at that time. The risk is that then the SOC may be obsolete, but the risk of never having good support is bigger, so avoiding effort now makes sense.

I'm against crowdfunding anything until I have it ready, to be honest. It was different with the Pyra, as all the money was needed for the development, but the shop profits should be able to fund the development of a new CPU board and until this is ready and working, I wouldn't do a new crowdfunding.

Based on the history of the project, I doubt we would receive many more new backers, and I'm not a fan of getting more and more liabilities as well.
Not crowdfunding makes sense. Funding from shop profits, I don't know. If that makes sense to you, then I guess it's OK for everybody.

Post automatically merged:


Don't underestimate the OMAP5. The raw CPU power is almost on par with the current Rockchips!
But yes, the Rockchip needs less power for the same tasks, has a better GPU and supports 64bits.
But the OMAP5 is not as weak as most think.
Possibly.
Yes, that is basically the plan. However, finding a proper SoC for that isn't easy.
There is almost no SoC available (if there even is one!) that doesn't have firmware snafu's and are fully supported by mainline.
this. This is so painfully true. And I don't see a big enough concern or reaction, so blobs keep increasing everywhere: linux, u-boot, coreboot, even libreboot.
There shouldn't be need for linux-libre, libreboot.at, u-boot-libre, FSDG distros and all that jazz if more people just rejected blobs.
I recently watched GNUtoo in LibrePlanet and got depressed.
Update 2024: Rockchip seems to have droped some source and Collabora seems to have free TF-A sources built for RK3388. DDR init is still a blob and I think the GPU needs proprietary firmware too. Just one less blob.
hns did some tests, and the Pyra can happily idle for 14 hours on battery with his setup. That's idling, without any low-power-stuff setup.
Our OS can't do that, so apparently, we have some stuff enabled that drains the battery even though it shouldn't.
The OMAP5 can do a lot more once these have been figured out, and with AESS almost working, we're on a good way.
Yes, but the user base, at least the free software community around it, seems to be small. Just some hundreds of units, so at most few dozens of developers.
With rockchip, or NXP or Allwinner one might have more people doing stuff in more devices and might take advantage of something.
A new SoC would probably cause the same issues once again, unless we use an outdated kernel by the manufacturer that probably uses other binary blobs as well.
I agree, and the problem is that the community around other SOCs might be bigger but it's too tolerant of blobs (by choice or by necessity).
So it's not really a solution.
If I ever find a suitable SOC I'll tell you, but I know none now. And LibreSOC seems not to be around the corner, unless the ED's shop profits can spare some dozen million euros or so...
 
Last edited:
  • Like
Reactions: rSl
Yes, that is basically the plan. However, finding a proper SoC for that isn't easy.
There is almost no SoC available (if there even is one!) that doesn't have firmware snafu's and are fully supported by mainline.

MNT seems to do pretty ok in terms of software support with the NXP i.MX8M Plus that they plan to use in the Pocket Reform. But their form factor gives them more space, and they also have a larger battery (no idea how that SoC compares with the OMAP for power consumption).
 
MNT seems to do pretty ok in terms of software support with the NXP i.MX8M Plus that they plan to use in the Pocket Reform. But their form factor gives them more space, and they also have a larger battery (no idea how that SoC compares with the OMAP for power consumption).
i.MX8M needs blobs
 
  • Sad
Reactions: rSl
Sure, the MNT docs explicitly mention
  • Synopsys DDR4 calibration firmware lpddr4_pmu_train_*.bin (no source available)
  • Cadence HDMI controller firmware signed_hdmi_imx8m.bin (no source available, optional)
For me, those are less of a problem than an unsupported SoC that no one does any hacking on anymore. And any SD card you could plug into a Pyra probably runs on more closed-source firmware than that.
 
Well, the OMAP5 is still in production, only the development itself by TI is dead. That said, it still receives updates in mainline Linux kernel - whereas with other vendors (like Rockchip, for example), there's no real support at all. Once they release the SoC, they release a kernel and that's basically it.
Wow, I thought the production was stopped long ago. So there are still companies buying OMAP5 chips? I always wondered how this chip was ever used industry-wide. I heared the chip also was intended for the automotive sector.
Does anyone know where the OMAP5 SoC was or still is used?
 
Not crowdfunding makes sense. Funding from shop profits, I don't know. If that makes sense to you, then I guess it's OK for everybody.

Well, unless there's some rich guy throwing money to fund that without wanting anything back, I don't see any other way to fund the development :)

Yes, but the user base, at least the free software community around it, seems to be small. Just some hundreds of units, so at most few dozens of developers.
With rockchip, or NXP or Allwinner one might have more people doing stuff in more devices and might take advantage of something.

Yes and no.
Most of the Rockchip stuff is Odroid or Anbernic or similar. The users mostly care about emulation. There's ArkOS, EmuElec and so on - and they simply reuse the (outdated) kernel provided by Rockchip and try to optimize the rootfs.
This wouldn't really help us that much.

MNT seems to do pretty ok in terms of software support with the NXP i.MX8M Plus that they plan to use in the Pocket Reform. But their form factor gives them more space, and they also have a larger battery (no idea how that SoC compares with the OMAP for power consumption).

I've closely followed the NXP and was hoping for a proper sucessor of the iMX.6.
However, all their latest SoCs are either suited for automotive use (so they run as hot as the OMAP5 and have a horrible perf / W ratio. For example, the perf / W of the i.MX8M Plus compared to the RK3568 is four times worse)) - or they're suited for IoT, so for small applications.
None of the i.MX8 is suited for a small mobile use. I've neve seen an i.MX8M Plus without a huge heatsink on top, most often they use fans as well).

So that wouldn't give us a huge advantage compared to the OMAP5 as well.
 
@EvilDragon @hns About the battery, how is "made" the idle test ? I mean is the screen on at minimum brightness ?
With or without keyboard backlight ? Wi-Fi and BT off ?
 
Last edited:
I've closely followed the NXP and was hoping for a proper sucessor of the iMX.6.
However, all their latest SoCs are either suited for automotive use (so they run as hot as the OMAP5 and have a horrible perf / W ratio. For example, the perf / W of the i.MX8M Plus compared to the RK3568 is four times worse)) - or they're suited for IoT, so for small applications.
None of the i.MX8 is suited for a small mobile use. I've neve seen an i.MX8M Plus without a huge heatsink on top, most often they use fans as well).

So that wouldn't give us a huge advantage compared to the OMAP5 as well.

NXP i.MX 8M is used in Librem 5 smartphone and there is no big heatsink nor fan inside it: it is a smartphone much thinner than Pyra.
They choose NXP i.MX 8M because of support, long term support and availability. If it goes good on a smartphone it can go good on Pyra.
 
Back
Top