There is likely missing information, so feel free to correct any mistake I may have made. I wanted a centralized place to find this information for other people’s convenience. It’s difficult and unreasonable to ask people to read through over 300 replies to this thread just to find the relevant links so I’ve taken that time for everyone else.
I’m not a gatekeeper for this community wiki so anyone is free to add who they wish to add. I’ve taken the liberty to add them at your recommendation though.
So, A) if you are from these parts, send messages to appropriate entities (see end of this page), and, B) I hope one/some of these linux phone HW and SW projects could use something like this to get financial support.
There is a new blog: PureOS Crimson Development Report: July 2024. It starts with “thanks to our subscribers” and gives then a great overview of recent Crimson related development.
Ever since I switched to Mobian (and not once thought of going back to Byzantium) on my L5 I was wondering why it needs both PureOS and Mobian. Can someone explain to me (or refer me to the forum posts if already discussed) why Purism doesn’t merge with Mobian to use synergies?
Put differently, will efforts put into developing Crimson/PureOS also benefit Mobian? Why doesn’t Purism focus on hardware and perhaps contributing some L5 hardware specific software development to Mobian so that there can be one bleeding edge system for the L5 instead of two half-baked ones?
Perhaps this is already basically the case. Perhaps it’s not as simple as it sounds. Please someone explain to me.
Not in any way ever speaking on behalf of Purism but my 2¢ …
Purism intends to upstream what they can into Debian. That could benefit Mobian. That could benefit other distros as well. That could lead to the question (in the long term) … why does either PureOS or Mobian need to exist for the Librem 5?
There could be some merit in having a single consistent distro across all the Purism hardware. This is doubly so with convergence where you would have to justify any inconsistencies between, say, the experience on the Librem 14 and the experience on the Librem 5 when in convergence mode.
It is also not clear to me whether Mobian specifically meets the goals of purity, for example, if PureOS sought FSF “certification”.
It’s an interesting question, especially now when Purism is lagging behind with PureOS Crimson.
Part of the background I think is that when the Librem 5 was first developed, Mobian did not even exist. Purism needed to have an operating system that would support the Librem 5, and PureOS was a way to solve that. Waiting for something like Mobian to exist would probably not have worked; I think the Librem 5 itself played some part in Mobian coming into existence, and the Librem 5 would not have existed without PureOS.
Then there is the matter of different choices made when maintaining a GNU/Linux distribution. With PureOS, Purism can make their own decisions. One example where this matters is in the choice of defaults, for example on PureOS we get Duckduckgo as default search engine, while in Mobian the default is Google (at least it was last time I tested it). And as @irvinewade said there is the FSF endorsement, where PureOS is on the FSF list of “Free GNU/Linux distributions” while Mobian is not.
In my opinion, it is good that Purism maintains PureOS and I really hope they manage to continue with that. For me as a Librem 5 user it means that I have someone to turn to with software-related issues, which sets Purism apart from other companies selling hardware who don’t care much about software.
And that is one point I don’t quite understand. I am wondering why, for example, energy consumption apparently is so much better on Byzantium. Is it that Purism maintains their tweaks for an older kernel but not for a newer one used in Trixie? So as PureOS gets lifted to lets say Crimson (or probably would need an even more up to date version?), Purism’s contributions would then benefit Mobian/Trixie? I simply don’t know how this works. But if going back to square one explanations is too much an effort, don’t bother.
Makes total sense.
Definitely makes total sense and might be one key reason.
I agree, and I am happy they did that and pushed Gnome on mobile!
Generally, I share your opinion on that (even though I have great faith in the Debian community!). I just hope that this won’t create too many inefficiencies and double work for both projects and ultimately disadvantages for all. That is (still) my concern. But anyway, there are smart people at work and I’m positive they’ll manage the best feasible way - sooner or later.
Mobian’s stated goal is to merge all patches to Debian. So eventually (hopefully) there will only Debian → PureOS. Currently Mobian uses the same kernel patches as PureOS. But getting all patches accepted into mainline kernel is going to take some time.
One difference is about policy of updating software - Mobian will only provide bug fixes to software in bookworm (some subset of updates may be available via debian bookworm backports), but PureOS will provide new versions of mobile core apps (phosh, calls, chatty etc). So when PureOS crimson releases, it will be very close to Mobian Bookworm, but crimson could get newer phosh, calls, chatty etc where as on Mobian we will have to wait till trixie release for these updates.
I have no way to gauge the development velocity, but I’m glad that software updates are trickling in each month. It solved a Bluetooth problem I had with the old wireless card
This plan is now no longer secured due to recently acquired information on the Qubes OS Forum:
In the case that both the Librem 16 and NovaCustom V54/V56 laptops with hardware kill switches are available before Crimson is officially released, here are the reasons why I would consider switching to the latter:
Intel® Core™ Ultra 7 processor 155H with Dasharo firmware (Coreboot + Heads with Nitrokey 3A Mini workflow integration pending compatibility research) and Intel Management Engine disabled via HAP bit.
Supports at least 96 GB of DDR5 RAM @ 5600 MHz.
PCIe 4.0 support.
Privacy screen can be preinstalled.
3 years warranty, firmware updates at least until July 10th, 2030, replacement parts at least until July 10th, 2031.[1]
Here is information that has no impact/weight on my decision, but may be of relevance to Purism:
The NovaCustom V56 can be configured with a 2,560x1,600 display (LG LP160WQ2-SPB1).
The NovaCustom V54 and V56 can be configured with an Intel Wi-Fi 7 BE200:
The NovaCustom V54 and V56 keyboard layout can be completely customized, including backlight illumination and ANSI/ISO options.
Here are the reasons why I would consider the Librem 16 instead:
Feature parity against the V56 with at least the first three reasons (switching Dasharo with PureBoot).
Actually implement a write-protect dip switch on the motherboard for the BIOS and EC:
Dasharo firmware (Coreboot + Heads) support for the V54 and V56 are planned for March 2025 and earlier:
In the chance that Crimson is stable before both the Librem 16 and V56 are released, treat my potential purchase of the Librem 16 towards development of Dawn instead.
Thanks for sharing. One question, does the use of Clevo as an ODM vs. custom or in-house design/developed like Purism does, have significance for you?
I have used qubes on my desktop and have been searching for a laptop that can handle qubes reasonably well. I distilled my search down to Purism, NovaCustom, NitroKey and StarLabs. I realized novacustom and nitrokey both use clevo. where as Purism and StarLabs design and build in-house (or at least don’t rely on clevo). In my mind, the fact that they are designed and built in-house seems better. But, I can’t honestly say why I think that. I just tend to have more concern with mass-produced, or companies that focus on mass-production as it could be a potential vector for introducing vulnerabilities.
Along these lines I thought of going with starlabs, but they seem to be in the process of redesigning their Star Fighter laptop, so I’m waiting to see what they come up with.
No, because my main concern for hardware is strictly focused on boot firmware root of trust. This means Coreboot with Heads is the bare minimum criteria for any potential x86 product consideration, therefore Star Labs is excluded due to only using Coreboot with TianoCore.
While Purism (likely) doesn’t use clevo, they still have/use a Chinese ODM. Purism has not revealed who that ODM is. Clearly that ODM allows (and/or helps with?) more customization (e.g. the Hardware Kill Switches).
