Librem 5 phone by Purism


Such a Pyra phone would be heavy and bulky wouldn't it ? If there are ports and a replaceable cpu board.
 
Why is purism not using mer/fremantle as a base? Reinventing the wheel. Redoing the work...
 
Considering the development time, the price and the avability, of the OpenMoko, Neo900, and Pyra,
I think that I have far more trust into the Purism Team.

Their Librem line of laptops are REAL, USABLE and equiped with the CURRENT GEN hardware,
And as far as I can see they are pretty successful.
 
What about a Pyra-Phone?
Strong maybe from me, depending on what happens with call and text functionality on the Pyra. Personally, I am more interested in a tablet, or toss in an e-ink screen and call it an ebook reader.
 
I would back this project if there was no such thing as a Pyra.

As it is I am hoping the Pyra will fill my need for an open, private linux device that can make phone calls and such. A bluetooth earpiece or dialer and some scripts should do the trick quite nicely.

I'm actually considering brushing up on my meager coding abilities and trying to write a proper answering machine software for the Pyra. Paid voicemail be damned. Shouldn't be TOO hard to have the Pyra auto-answer an incoming call after 4 or 5 rings, play a specific audio file to the caller, then record any response, convert it to FLAC, name the file after the caller's phone number, and stick it in ~/.messages/.

Then I can listen at my leisure and store as many as I want. I can use any greeting I like of any length, record any length response given by the caller, and see the date & time & number of each message with a simple ls -l. Maybe over time I can even add some advanced features, like what YouMail offers. At first though I'd aim for a dead simple answering machine that works reliably.

That said, the Librem 5 looks awesome. If I had more money I'd back it anyway, just because I think it needs to exist.
 
So it seems KDE plasma support brings in money.
Also, the presentation is stellar. Easy to decipher, clearly laid out, pleasant looking.
Between that, what the pyra has on offer, and fairphone's environmental efforts, i think the variables of a good, well marketed product, with an audience, is to be found.
 
Such a Pyra phone would be heavy and bulky wouldn't it ? If there are ports and a replaceable cpu board.
Yes, it will not be the slimmest one. But the slim ones have neither ports for external devices nor a replaceable CPU. So it is a trade-off what is more important for a GNU/Linux device. Is it more a portable computer and communication device or a life-style gadget.
 
Last edited:
You will want to look at multi-microphone noise suppression if you are attempting to make a phone. Otherwise the sound quality on the call will be bad.
 
Also, there are a billion slim ones on that are all aiming to be the same thing, with very little difference between them, readily available.

Librem makes the case for being safe in a way that gives the user freedom, without that it wouldn't make sense for it to be crowdfunded.
 
Also, there are a billion slim ones on that are all aiming to be the same thing, with very little difference between them, readily available.

Librem makes the case for being safe in a way that gives the user freedom, without that it wouldn't make sense for it to be crowdfunded.
I think crowdfunding and user-freedom don't have any relationship. They are orthogonal features of the whole offer.

Rationale:

Crowd-funding: is just selling things before they have been designed, tested, produced and became tangible. I.e. pre-order, upfront-payment etc. for an "Idea". The only reason that special crowd-funding platforms exist is that Amazon and Alibaba don't offer that. So it is a feature of the marketing and sales process and not of the product.

User-freedom: can exist for crowdfunded products but also for products where there is an investor behind, paying for development and the first production batch. User-freedom is a feature of the product and not how it is sold and development is paid for.

Well, there might be one relationship: people who are more open-minded for free&open software may be more open-minded for buying untested devices in idea stage.
 
Last edited:
Considering the development time, the price and the avability, of the OpenMoko, Neo900, and Pyra,
I think that I have far more trust into the Purism Team.

Their Librem line of laptops are REAL, USABLE and equiped with the CURRENT GEN hardware,
And as far as I can see they are pretty successful.

Well, building a laptop is a LOT easier than such a small handheld device.
For laptops, you don't really have space constraints, can use a standard x86 board which has tons of reference designs, etc.

It's basically layouting an existing schematic onto your PCB - without the size constraints we have on the Pyra.

However, IF they actually fund the device, it'll probably be a smoother ride than with the Pandora.

We've developed the full system with 200k EUR, with a minimum order qty of 1000.

These guys need at least $1,500,000. The phone itself is $600.

They need at LEAST 2000 phones backed, which is twice the amount the Pyra has.

With the higher order qty and lower production costs they have (no keyboard, more simple case, standard LCD with capacitive touch, etc), it would probably be the same as if we had 600k - 700k EUR ready for development.

There are multiple factors that made Pyra development that slow:

  • Use as little money as possible to keep within the 200k range (otherwise I could've easily produced tons of different test PCBs to see which memory works best, etc.)
  • Almost everything is custom made
  • One hardware developer doing it all. That's also a payment issue. If I had a team available, it would be a lot faster and a lot more expensive.

So the Librem team has planned in enough money for a quick and proper development and production run, which is good.
On the other side though: They need a lot more backers than we do. Which could mean that it will never see the light of the day.

So both approaches are different, and both are okay. Ours is more risky, theirs ensures a smoother development if it ever happens.
[doublepost=1505743494,1505743283][/doublepost]
I'm hoping that the Pyra will have a faster CPU-board upgrade available by 2019 (Librem 5's expected release date).

Well, the i.MX8 is easily available, also for the Pyra.
But I'm not sure that's a good replacement... it certainly is powerful, but it's not meant for mobile usage as well:
https://www.nxp.com/products/microc...tualization-vision-3d-graphics-4k-video:i.MX8

It's for automotive market.
I've seen it this year at the Embedded world with one huge fan on top.
Like the OMAP5, it's build with 28nm, so it will also have heat issues with small consumer devices.

EDIT: Correction, there is one suitable for mobile usage:
https://www.nxp.com/products/microc...video-dsp-error-correcting-code-on-ddr:i.MX8X

But that one even seems to be a bit slower than the Cortex A9 though.

The i.MX6 is still Cortex A9, so probably slower than the Pyra.
[doublepost=1505743544][/doublepost]
https://puri.sm/shop/librem-5/
Looks good on the specs and features.

Well, as mentioned above: Not sure the i.MX8 is feasable and the i.MX6 is slower than the OMAP5.
 
pabs on debian-arm shared these links

https://www.qualcomm.com/news/releases/2016/10/27/qualcomm-acquire-nxp
https://www.fool.com/investing/2017/06/29/qualcomm-isnt-giving-up-on-the-38-billion-nxp-deal.aspx
https://www.fool.com/investing/2017/07/03/european-regulators-stop-the-clock-on-qualcomms-me.aspx

I think we will at-least see the i.MX8, but but.

At this point, i don't really see a whole lot of value in improving speed for a good while.
Reducing weight makes sense for handheld operation, but it is a game of a human eye-quality, for the size permitted. Games requiring top-end systems really either aren't for handhelds, or better enjoyed elsewhere.

Edit: I think the forever lucrative trade and enticing idea of assuming or assembling specs is contrary to actually putting something that exists together. It is there the journey truly begins, and only through that efforts can be made to improve the bigger picture.

Librem made a huge jump up to $456,077 (!) KDE really makes a difference.
 
Last edited:
Yes, the iMX6 is an old chip. I've bought a Hummingboard with a single core one of them on it for a permanently on machine, because I hope it's efficient, but not even as fast as my Pandora I'm expecting. Sadly, it seems I may need to return it because I don't seem to be able to get it to boot.

Regarding a pyra-phone, I'd expect more uptime for a phone than I do for a handheld, and less heating issues. If the latter all goes away, and the CPU turns out to be acceptably efficient if clocked down, the OMAP5 might be an option, but I'd be more confident with a lower rated CPU (although not quite to iMX6 levels).
 
Yes, the iMX6 is an old chip. I've bought a Hummingboard with a single core one of them on it for a permanently on machine, because I hope it's efficient, but not even as fast as my Pandora I'm expecting. Sadly, it seems I may need to return it because I don't seem to be able to get it to boot.

To follow up, I finally managed to get this to boot a couple of days ago. Sadly, it doesn't seem that efficient - running a memtest on it from uboot made the CPU too hot to touch, and registered as 70 degrees C on my thermometer. Booting an OS and leaving it idle, and it still rises to 20 degrees over ambient. You can't feel this outside the case I've got it in though, so I'm not sure how it really compares with the OMAP3-type chip in my Pandora for example.

I just checked using my smartmeter, and it reckons that usage is around 3W idle, but resolution on that meter is only +-1W, so take that with a pinch of salt. It does seem clearly higher than my Pandora, and it does underclock the CPU to around 350MHz when idle (close to what my Pandora does). It does have ethernet though, and I have a keyboard plugged in to the USB, and I'm not sure what those take (USB is 5V at up to 500mA times two I think, so up to 5W, but my USB 1.1 keyboard nowhere near takes that). All ports will take a little trickle current to keep active
 
$609,052 (40.60%)/ $1,500,000 with 31 days to go

Edit: What makes this so interesting is the prospect of KDE and Gnome going head to head on the same hardware.
 
Last edited:
Edit: What makes this so interesting is the prospect of KDE and Gnome going head to head on the same hardware.

How so? I know in my experience of being a graphical GTK user (not used full-fat Gnome since Gnome 2) is that I often end up with some Qt libs if I'm not paying attention, and I suspect it may be even harder for KDE users not to end up with some GTK libs on their systems after a while. You can run the backends of both systems at the same time, and I consider this to be pretty normal behaviour. Running the front ends at the same time would be something new though!
 
Back
Top