It’s a good point. We were going to wait until we knew what the final price would be (because that would be the first question someone asks support) but you make a good argument for at least updating it partially.
5 Likes
Just write “we do not know the price yet and will update this article once we know” problem solved.
But one question is still nagging me most. When you pitched for the final pre-order before the price increase. Were you really still believing those pre-orders would be delivered in Q3 as advertised?
I agree. I mean seriously? Trust China? Trust USA? There is little difference in who you trust. Trust yourself. That’s the best you can do. Read Snowden’s book, or the Levine book to get perspective on trust.
1 Like
Regarding the trust issue: I guess the main selling argument here is actually modem isolation - so, in theory, you don’t have to trust any of them. At least as long as the isolation works and the modem is cut from DMA access as well e.g. properly making use of the SoC’s IOMMU. My understanding is that this needs to be done since a PCIe device normally has DMA access.
However, while with proper isolation the modem could not be able to leak actual data from your device (e.g. encryption keys) or allow remote access, it may still leak your location as it has access to it.
Regarding the “made in Germany” topic: As I know, in German law, you may use this label if the last “substantial” step in manufacturing is done in Germany, no matter where all the components come from. Packaging does not account here, but e.g. if you would solder one resistor to the board it would be sufficient to put this label on your product. Therefore, at least for me, the label alone does not increase my trust in a product.
2 Likes
The modem doesn’t have location data. It has an antenna for GPS location data, but that gets left not hooked up. It has to ask the phone for GPS data. It does know what towers are nearby, so it can still give a pretty good idea of where it is, but then again the towers can see it anyway.
2 Likes
Yes, a major part of the value added. The corresponding (German) Wikipedia article is interesting. Because of the wage differences, a final assembly of foreign components in Germany will often fulfill that criterion. Mind, assembly, not packaging.
1 Like
It doesn’t have coordinates, but it has lai/rai and broadcasting beacons around. When you switch to low precision location mode - gps is turned off and your location is determined by google using geoip + 3g location.
The modem’s M.2 slot only has USB connections. It can’t do any kind of DMA.
3 Likes
Well this “rough” location data is already quite concerning to me as it is sufficient for making movement profiles of people - even if it is 50 meters off. I fear as long as the modem firmwares remain proprietary, we won’t be able to address this challenge. Therefore, having a “more” trustworthy modem (however you want to assess this) remains favorable.
Certainly. I wasn’t trying to belittle the difference in trustworthiness. Unfortunately, even if the modem were completely under the owner’s control, that location data would still be available to the network operators, as your phone must transmit in order to connect.
2 Likes
Help us Obbie, Juan, and Kenobi…
You’re our only hope 
In my mind, there is a major difference between when someone tracks you because you choose to transmit a signal and they tri-angulated your position while you transmitted, and the case where someone uses your own hardware against your will, to track you. For all practical purposes, maybe the tracking result is the same. But then again, maybe not. But the morals of these two different cases say everything about the difference between having a free country and living under an oppressive regime.
It’s all about who owns the phone and about consent. But when you show up in a public place, yeah, people are going to see you there. That doesn’t mean they have the right to search you simply because you showed up.
4 Likes
I think you’re missing the point that to receive a call the provider has to know where you are to send you the call. If you don’t want to receive calls and don’t want to be tracked turn the hardware kill switch off.
I’m not saying I don’t want the baseband fully freed, I do, just that the particular concern you’re presenting is already addressed as best as is feasible and even with everything fully freed the kill switch is the most user-friendly solution that I’ve heard presented.
4 Likes
Exactly. The issue with the untrustable baseband is it is a ‘beachhead’ from which further attacks against the phone can be launched. We’re seeing lots of sidechannel timing attacks against multithreaded CPUs. We now are putting untrustworthy devices on a USB bus. Sure, that keeps it from having DMA to the system memory, but it can still potentially exploit security problems with the kernel USB driver or similar. I’d far rather have an attacker have to get past a firewall on the modem first.
1 Like
Yup I do agree that baseband on usb it’s just a modem. Yes, modem could be compromised, and there are some attack vectors via usb (eg to emulate input device and attempt to control by sending keystrokes) but that’s a bit different security plane.
3 Likes
@kieran, you could have self found information from manufacturer that scheduled date of introduction to production of Revision 4 was by PCN on CW10/2019. I have no further or current comment and don’t know if this helps at all, but might help if nothing changed in scheduled production, even though I don’t think that such information would be available to the end user. There is nothing that I can guess about linked info, other than that I might have some fear as for every product exists something that is called EOL. Furthermore, I believe that Purism searches continuously for best solution or already have adequate modules (PLS8-E and PLS8-US) that they are able to control (USB serial drivers, firmware, etc.) on Linux OS. But if another adequate module manufacturer comes from China and have “better” product like, just speculating, N75-X (FCC ID PJ7-N75-NA) from Shenzhen Neoway Technology Co., Ltd why should I care where it is manufactured. Purism knows much better than I what they do and I trust them (like yourself). I think that my bottom line is like this: who makes final decisions for Gemalto nowadays is not my concern, but, IMO, it is not making life easier (or cheaper) to Purism either if Thales Group invests rather somewhere else / produces something else. And, if small companies are indeed making their decisions based on papers like the one I linked here it is just coincidence because I’m just customer (not subject matter expert).
While I almost completely agree with what you are saying, this part
is flawed and led to situation we have today with big techs who knew better what average user needs, So we do need as peer-reviewers so watchdogs who can flag issues before it’s too late (even in form of Jay). I trust purism way more than others but in general I can say - I don’t trust purism (as myself). Which is why I want to understand rationale behind choosing this or that component, because obviously all of the hw/sw choices will be highly criticized the more success this phone will be spinning. And I want to know (not believe) that those are still right. I can make my own opinion in sw world but in hw I never went further than cortex m4 isa/architecture (if going from discrete logic via mcu/fpga to cpu).
2 Likes
You make a good point about how trust should work, I like it. IMO this type of technical decision making stuff is where transparency matters. As opposed to the internal details of schedule slippage or unexpected issues, where I wonder if transparency hurts more than it helps, but is still expected by backers.
the way this usually works is neither belief is completely devoid of knowledge nor knowledge is completely devoid of belief … you have to FIRST believe you will find something somewhere based on certain past experiences of you or somebody else or something/someone drawing you there BEFORE you will KNOW what exactly it is or IF it is (at all - like REAL - based on how our current reality works)
naturally this example is completely VOID for closed hardware/software because there is no way to establish a public precedence with current state-of-the-art hardware/software unless it’s reverse-engineered instantly as it is released to the market (and even then it might already be too late for that) > results in no conscious trust or trust that is not trust because it isn’t firmly anchored in anything TRUST-ABLE … 
2 Likes
I don’t want to go to the depths of absolute knowledge and 42 but I want to know reasoning.
2 Likes