@deedend I’d just wait for the final announcement or the mail that asks to confirm shipping details. In case they can’t offer a suitable modem day one, I’d then ask for details, and possibly to delay shipment until they worked it out. I think it’s awesome that they didn’t just settle with the modems they have so far.
Probably there is also a commercial things, i mean there is also a competitor, with a lower cost, the pinephone, probably saying something more now could be a bad move, and i also think big actor do the same with their product, it’s named hype.
I don’t really care to know now about hardware, i’m curius ofc, but my main goal is to receive a phone without hardware problems, because ship it from eu to us os not really funny, and to have a phone whit a long life regarding hardware and software, last but not least a blob free device i can use as i’m using my android.
That’s exactly how I feel.
Modem will decide whether or not I’m cancelling the order.
Even if they leave the option to add another one on our own, I would need to know exactly which models are available and for what price. Then, I would need them to be tested on major networks here in US. I can’t afford to shop around, spending another hundred or so and eventually still not have an adequate network connectivity.
Another important element is VoLTE. If that can’t be enabled, I won’t be able to make or receive calls in most areas I either live or commute to.
Since this is US based company, they should do all these testings and present them to buyers.
So we all know what exactly we are getting and what we can expect from upgrades that are available.
Otherwise (as I said in other forum), for some people this can end up as super expensive iPod.
I’d be much more optimistic about the final hardware being ready if they had given us a solid release date by now, but we’re two weeks into Q3 and every source on the release date just says “sometime in Q3.”
I’m anxious for a release date as well, but I suspect we will get one (plus hardware details) at the end of the month, when the price goes up. I’d actually be (pleasantly) surprised if the got those details before then.
Well, I already paid for my phone and I only hope the team is making a good Linux phone. I do not care if it takes a little more time - the end result is much more important. I do not care very much about update announcements either because they do not make the phone better. On the contrary they might take time from the important work with the phone.
We can discuss this feature and that feature but the main thing for me personally is that I get a Linux phone and can get rid of my Android phones (I have one Samsung and two Motorolas) because I am completely fed up with Snoople and all kind of other snooping applications. Thus I am ready to pay Purism a heap of money (I am not rich) to this end. I have only Linux machines at home (running Debian) but there has been no possibility to have a mobile phone other than Android or Iphone (and Apple is out of question for me). Sure, there has been attempts to construct a non-Android phone among others by the former Nokia people but the Librem 5 project is the first (known to me) which has focus on important privacy and security matters. So I won’t disturb you but wish you luck. What I have read it seems to be on track so keep up the good work !
Is it possible to duplicate the IMEI number from one of the old devices and use it? Just wondering if I could use VoLTE with that workaround. Major US carriers allow that important feature on certified devices and I doubt that Purism will have resources to go through that process.
Though, I was surprised to see essential phone get that certification done on Verizon, so can’t say it’s impossible
They slipped something, although digestible only for developers:
the first cut of the phone device tree is available here.
From this i conclude, that all the hardware pieces are chosen, and the main board is ready. There might be business reasons to still decide this or that camera, (with USB interface, you really can decide this on the last minute) but they tested all the candidates already.
What will the Librem 5 camera specs be?
I tried to gathr some insight from the device tree. I found.
The USB-C PD Controller tps65983, which is enable for doing both roles, sink and source, for power and data and suporting alternate roles for the like of USB 3 and DP . Nothing unexpected here but nice to have it confirmed.
The Touch screen (controller) GT5688: I couldn’t gather any thing interesting from this. Maybe someone else can?
The power management unit BD71837MWV: I also have no expertise in this an can’t make anything out of this.
A controller for a dual led camera flash and torch lm3560: So yeah dual flash, maybe two tone? For all the photographers here.
There will be an RGB front LED.
The battery charger modul bq25890 support up to 14v 3A for quick charge, which meats the 9v of USB-C PD specs and the 9V setting of QuickChage 4 and 4+. OTG devices can be supplied with 5v 2.4A, which should be enough for most USB devices but you can’t charge an other phone with QC . So seams good.
The battery monitoring chip is max17050. There are more spec about the battery like min/max voltage and charge/draw amps but i couldn’t find the most interesting the capacity.
For the SD card i think i found the most interesting thing hidden in the device tree, but i’m not 100% sure. The wifi sdio interface is configured with a bus width of 4 but the sd card interface with 8. So i believe this means the sd card read will support USH-II or maybe even USH-III as the additional pins are need for those only. This would mean at least 3x the bandwidth 104 MB/s vs 312 MB/s with probably lower latency in the same order.
For now this is all i could see in the device tree. Apart from trivial things like there will be kill switches headphone jack, gnss, magnetometer, two cameras.
The TPS65983 USB-C PD chip only supports DisplayPort alt mode over USB 3, so I it looks increasingly unlikely that the Librem 5 with support HDMI. We knew that it was unlikely considering that the i.MX 8M requires a binary blob of HDMI. Kind of a bummer since most people don’t have DisplayPort monitors, but it is a freer standard than HDMI.
Looking at arch/arm64/boot/dts/freescale/imx8mq-librem5-phone.dts, I found that the Vishay VCNL4040 GNSS will use I2C, and not UART (not that anyone cares, but I find it interesting how the Librem 5 is communicating with the chips without binary blobs in the kernel).
camera 1 and 2 are in the I2C bus are: “/* just a placeholder for nuw */”
I suspect that Purism is having trouble trying to implement a camera with i.MX 8M because NXP has not documented its camera interface very well and people on the NXP forum can’t figure it out:
I see that someone has figured out how to get a 13MP camera to work with the i.MX 8M, so it isn’t impossible, but the 5MP camera used in the Librem 5 Dev Kit, is the same one that NXP provides as a reference module and people are having to study that one to try to figure out how the registers work, because NXP hasn’t documented them. This isn’t like a Snapdragon or MediaTek SoC where everyone attaches a camera sensor, so its well documented and tested.
Confirmed by you as the current solution meaning there will be no HDMI 1.4 out over USB-C (using the USB 3.0 protocol) or should we anyway wait for an official confirmation from Purism?
So at this low of a level I’m not sure how this works. DP passively converts to HDMI insofar as I can connect a devices display port out directly to another devices HDMI in with nothing else involved and it just works.
With that said, would using DP over USB passively convert to HDMI without issue or are there pieces at play here I don’t know about/understand?
Sorry for the sidetrack but depending how how this works may affect how much HDMI over USB even matters?
I haven‘t read the hole document, but what i read is that they named dp and thunderbolt as example. That wouldn‘t exclude HDMI. @amosbatto can give some more information where you get from that hdmi is not supported?
I allways understod thunderbolt as that i has to include the alternate modes. So that would mean they are named indirectly.
@Quarnero confirmed was mabye a little to strong. but the information isn‘t from me but form purism via a git commit. close to their announced release time Q3. So i think this can be seen as quite certain. But.you are right confirmed is the wrong word here an missleading
Actually i’m pretty confident that this chip can be used with hdmi. If i google for it many usb-c docks with hdmi seams to use it. i think this chip is mostly for the power delivery part the altmode muxing is done by another chip. And that is why it only names examples in the docs.
But still i have not read much of the docs and could be wrong.
Thank you mon ami. I might be able to read some of this but it looks like others have I believe they send this to the company building the hardware to flash after eh? Though, I think they plan to build it in house like the new librem key??
No, production will be in China and supervised by Purism staff (according to last official statement). I would expect some parts of the final assembly to take place in a Purism location, e.g. fitting the modem, battery, and putting the OS on it. That would be close to what they do with the laptops.
Well, that’s the devkit and it uses the camera for which NXP has example code.
Most certainly Purism wants to give us a better one, but as it has been pointed out it might be tricky to get it to work. And at least they want to be certain that it works before starting production.
That question is whether we get the same low-quality Omnivision OV5640 5MP reference camera that NXP provides or do we get a better camera. If NXP hasn’t provided good documentation, then it is more difficult for Purism to use a better camera sensor.
Looking at the pins of other USB-C PD chips that say that they support HDMI Alt Mode, it looks like the TI chip has the same pins and it has GPIO pins that could be used, so it might be technically possible. However, TI doesn’t provide an application note for how to implement HDMI Alt Mode, whereas it does provide an application note for implementing DisplayPort Alt Mode.
It seems like a lot of needless pain to select a chip which doesn’t have documentation for how to implement HDMI Alt Mode, when there are other chips that provide that documentation and advertise that they support HDMI Alt Mode.
It is probably best to await for official confirmation from Purism, since this just a draft device tree listing, but I wouldn’t bet on HDMI Alt Mode.
@ramnasko, I am thanking you for being clear and polite to me and this community. I appreciate your contribution and work as well.
DisplayPort and HDMI use different protocols. Many DisplayPort ports support both, enabling the use of passive adaptors. If the port really does only support DisplayPort, then a passive adaptor won’t work and you need an active adaptor instead.
I hope they ditch the selfie cam and use that extra millimeter of space for a bigger battery lol Linux nerds that are going to buy this probably arent going to be sitting there taking thousands of selfies to show the world that could care less lol