So it has shipped. Where are the reviews?

No it still works and keeps a phone for at least a day. I however am spoiled by its original capability to hold for 7 days. So having it for 1 day just makes me nervous. My new sony fortunately also keeps up for 5-6 days so I’m back on track. Battery pack defeats the purpose of a mobile phone (sitting in my pocket)

1 Like

7 days on a Jolla?? I bet you keep it with all the data/bluetooth/wifi off, I used a Jolla for 3 years and usually it ran for 2/3 days with light use (and the data on). Anyway I believe you would be disappointed by Librem 5 then, all the clues suggest that it will have a short life battery…

1 Like

No, i just never used aliendalvik which drains battery dry fast. Also I mostly run my own apps which I carefully power-profiled.
And yes, I’m trying to tune myself in, promising myself to powerprofile it as well :slight_smile:


So meanwhile we are at Birch.
I am still looking for that review where NORMAL EVERYDAY USE of this phone is discussed. How is calling going? Does it work on the networks? Do telco’s accept this phone on their network? Can i app and call people with iPhones and Android phones?
If Purism wants the Librem5 to gain market volume, please stop positioning it as a NERD phone. Don’t get me wrong, it’s great to see what is technically possible, but the desired market volume -imho- lies in what this phone could bring to privacy aware, REGULAR phone users, looking for a privacy-proof alternative. It is after all a communication device.
All i see on the internet are unboxing videos and other techno-boring reviews and stories of people running their own apps. I care a lot about privacy, please show me how this phone can free me, the average user, from the current duopoly…


This is my experience after the first few days:


You seem to have missed

please stop positioning it as a NERD phone

That’s weird. Most people assume that it is, despite all the marketing seeming to target ordinary users that are fed up with the duopoly. However, at this alpha stage, why the heck should they target ordinary users even more? They can think about that when Evergreen is out.


BTW according to your review you can discover and even pair BT? did you try to connect it via CLI using pactl/bluetoothctl to get audio?

No, I didn’t try.
I grew frustrated enough trying to get my notebook running Ubuntu 16.04 to play with my Beoplay H9 headphones. Nothing related to Purism products, just Bluetooth audio. Other headphones work with the notebook, and these headphones work with newer versions of Ubuntu. Just the combination I need doesn’t work.

…and ordinary users bought two years ago. Shyyyt, I’m a security guard (renowned for spelling our own names right…on a good day!). I don’t know where this person is getting their market info, but it’s not based in my reality (and I’m guessing mine is the one out of the two with an actual stake in this?)


Certainly a good suggestion. My experience with Bluetooth under Linux implementations has been very mixed and clearly audio support needs to be improved more than non-audio, but both suffer. Setting up Bluetooth headphones using PulseAudio is quite horrid in most any Debian distribution and pairing is often unproductive, even if it can be achieved. Using some Bluetooth connection to print can be a decidedly inconsistent practice. I have no doubt that much of the software needs to be corrected at the PureOS end, while an even larger set of refinements needs to be made to the various distributions by whoever has the facility in the community. I would plan to use some 2.4 GHz wireless unit such as Artiste for home use with audio rather than be bothered trying to make Bluetooth workable in its current functionality. I would hope technology could come up with a more portable version of this as Bluetooth, even at level 5.0 is still rather poor by comparison in both reliability and quality of transmission. Over time, Bluetooth may work better for my PC to headsets with my current MX 19 or Fedora 31 implementations.

well, bt audio is crap by definition, regardless of the version or OS :slight_smile: maybe because it uses same specturm as wifi/microwave but It never works reliably at my home. Unless you put a phone or laptop right next to the bt receiver of the speaker - it will always be disrupted, and since I turn on music to enjoy it just spoils it and makes me very frustrated. same for headphones. put a hand in the pocket - and audio start dropping. So I’m mostly giving up on it. the wearable/carkit though is a valid point as it allows phone to be pocketed and mute while still being aware of what’s going on.

Thanks for posting the output of those commands for me.

OK, so that tells us that the small Cortex-M4 core in the i.MX8MQ is disabled when the device is running. I wonder if it turns on when the Librem 5 is in low power, or Purism hasn’t yet programmed that energy optimization. Purism posted that extra core is used by U-Boot to execute the binary blob to train the DDR PHY during bootup. I hope that it gets used for something else.

The extra Cortex-M4 core to run the OpenPGP card reader is also disabled. I bet that Purism hasn’t yet got that working, but I guess that we have to stick in a card to know for sure.

I guess that the Standard Microsystems Corp. USB Hub and media card controller (IDs 0424:4041, 0424:2640) is the Microchip USB2642 listed in the schematics.

If lspci returns nothing, I wonder if that means that Purism disabled PCI over M.2? I guess its safer with no Direct Memory Access, but it would have been cool if enabled.

It is nice to see it confirmed that the Broadmobi BM818 is indeed using USB 2.0 and it is based on a Qualcomm cellular baseband.

I wonder what “network:0” is for, since it says it is USB, but is labeled as “Ethernet”. I assume that “network:1” is for the Redpine Signals RS9116 WiFi/BT, because its driver is RSI-SDIO, but it is strange that it is labeled as “Ethernet”
Can you do me a favor and execute this command:
dmesg | grep 'SDIO'

I don’t see the GNSS. Could you do me one more favor and install the i2c-tools package and then use this command:
i2cdetect -r 2

OK, and let’s see if we can get the Linux device tree:
dtc -I fs -O dts /sys/firmware/devicetree/base > device_tree.txt

Can you please post the device_tree.txt file?

And thanks again for doing this.


Here are the outputs:

purism@pureos:~$ dmesg | grep 'SDIO'
[    2.519548] mmc1: new high speed SDIO card at address fffd
[    7.082387] bus: 'sdio': add driver RSI-SDIO WLAN
[    7.082419] bus: 'sdio': driver_probe_device: matched device mmc1:fffd:1 with driver RSI-SDIO WLAN
[    7.082435] bus: 'sdio': really_probe: probing driver RSI-SDIO WLAN with device mmc1:fffd:1
[    7.082504] RSI-SDIO WLAN mmc1:fffd:1: no default pinctrl state
[   10.886480] driver: 'RSI-SDIO WLAN': driver_bound: bound to device 'mmc1:fffd:1'
[   10.886583] bus: 'sdio': really_probe: bound device mmc1:fffd:1 to driver RSI-SDIO WLAN


purism@pureos:~$ sudo i2cdetect -r 2
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          -- -- -- -- -- -- -- -- -- -- -- -- -- 
10: -- -- -- -- -- -- -- -- -- -- UU -- -- -- -- -- 
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
30: -- -- -- -- -- -- UU -- UU -- -- -- -- -- -- -- 
40: -- -- -- -- -- -- -- -- -- -- 4a -- -- -- -- -- 
50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
70: -- -- -- -- -- -- -- --


purism@pureos:~$ sensors
Adapter: 30a50000.i2c
in0:          +3.54 V  

Adapter: Virtual device
temp1:        +45.8 C  

Adapter: 30a50000.i2c
in0:           +3.58 V  (min =  +3.32 V, max =  +3.74 V)
                        (avg =  +3.58 V)
ERROR: Can't get value of subfeature temp1_min: Can't read
temp ambient:  +45.8 C  (low  =  +0.0 C, high = +70.0 C)  ALARM (LOW)
curr1:         -0.55 A  (avg =  -0.55 A)

dtc unfortunately segfaults


Hopefully they get the issues worked out before Evergreen.

Thanks Richi,
Does this show you anything?
i2cdetect -l