Why are Libreboot PCs and Replicant Androids so aging?
Under my observation the newest Libreboot laptops, routers, servers are i2-amd64 or equivalent, also the latest Androids that are supported for Replicant are already Galaxy Nexus, Note 2, S3 few years ago. These systems aren’t that modern to the laypeople.
Will there be workshops manufacturing modern processors requiring Libreboot microcode or not requiring any?
Will there also be Android OEMs making devices brought with Replicant?
hardware that lends itself well to reuse and tinkering is aging. a few years ago, for the first time and without realising it, i purchased a laptop that didnt have a way to change the battery without opening the case.
i was pretty irritated about that. luckily, as opening the case to get at the battery goes, it was probably more painless than most models. worth noting, i would have not purchased it had i realised it had this anti-feature.
oems are not gaining sympathy towards our type of use. they keep making devices that are increasingly disposable. i believe there will be a turning point, but not necessarily in the next 5-10 years. i used to have a lemote. heck, i used to have a compaq proliant and i once owned a pcjr. ive used a trs-80 and a c64 plus one of these: https://en.wikipedia.org/wiki/Minivac_601
Read the libreboot FAQ and you'll understand why: https://libreboot.org/faq.html
The TL;DR version is that newer x86 systems have fatal freedom problems.
The long-term solution is to leave x86. Please put your life jacket on, as the Titanic has hit the iceberg: https://jxself.org/titanic.shtml
Hopefully the Carpathia makes it to us before we sink. We know what happened to the real Titanic.
this has actually helped shape my thinking about this. sure, its not a complete change-- i am confident that things will get to where you say they already are.
but i am more ready to help support this argument, in my own way-- but its more ready than i was before i read this.
again, something ive talked to alex oliva about in the past. but ive mostly focused on x86 and thats definitely not going to work in the long run. the last time i had a machine with an arm processor though, the only distro that supported it wasnt fully free.
I watched Stallman's comments in the video embedded in that "titanic" page. I understand the concerns here, and I have long held a vision of a future where computers are manufactured by democratic cooperatives owned and governed by end users, instead of by parasitic corporations. But in the short term, it would make me very sad if we retreated from the strategy of trying to help people liberate whatever computers they already have, effectively limiting software freedom to only those who can afford to buy a new computer from a boutique supplier.
One thing that might help is to develop tools that can detect malicious uses of the potential vulnerabilities created by proprietary software and hardware. So for people with hardware that cannot run in full freedom, the strategy would be to help them remove and replace as many pieces of non-free code as humanly possible, and monitor the behaviour of any non-free components that remain. That way, we could move from a fear-based discourse that can come across as somewhat tin-foil-hat ("non-free code on your computer might be spying on you!"), to an empowerment-based discourse ("here's how you can find out more about what your computer is doing with your data").
I am a translator!
Simple answer: the masters that rule you find it really very delicious to backdoor your hardware so that if need be they can access your computer in no time, no matter the OS or the general security state of it.
The same delicious and very gentle gentlemen are working very hard, every day to backdoor encryption, they want weak encryption for you, the one they can break, again, if need be, in no time.
Now that you know how you might ask why? Well, because they love you too much to leave you on your own, and for your own sake, for your security. So that Ibraheem can stay far away and they can get closer and closer to you (and your pension).
They are coming to get your sweet sweet lemote_yeelong_2.0 and your sweet sweet free software and they will succeed eventually, no doubt.
one might think they are better off going amish at this point. however, youd lose access to the largest library in the in world, youd lose access to a global telecommunication network (for better or worse, but since the amish practice shunning as a form of emotional abuse, denying them access to the outside world is harder to defend as a practice) and as microchips and nanotechnology continue their very slow invasion of the material world-- its probably better to understand computers than to try to avoid them.
assuming that autonomy is still a goal. which means we should have computers, but we should probably rip out the mic and cover the camera at least. (and ultimately the speakers.)
people bring cameraphones to stallman speeches, he asks them (whether rhetorically or seriously) to turn them off. snowden recommended taking the battery out or putting them in the refrigerator. im pretty sure my phone can receive calls in the fridge-- im a lot less sure i can receive calls near the x-ray room in the hospital.
"they will succeed eventually"
yes, but they will fail eventually, too.
Boot Guard and similar anti-features render it impossible to modify the firmware.
And according to M$'s policy, in order to preload Losedows OS, manufacturers must implement UEFI with Secure Boot enabled.
For manufacturers, in order to satisfy the majority of users who are deeply locked in M$'s "ecosystem", they must ship their systems with preloaded Losedows.
So, just buy one used SandyBridge or IvyBridge system from the grey market and hope it lasts long enough.
Mobile phones and tablets have even more anti-features than desktop and laptop computers. Some manufacturers even don't allow users to unlock the bootloader at all (M$, Apple, Huawei, etc.).
"Mobile phones and tablets have even more anti-features than desktop and laptop computers."
i absolutely completely hate the things. sadly i need a phone, and i always get the cheapest flip-crud i can possibly find that doesnt use at&t (never, ever, ever) or verizon (never again.) tablets are overpriced and worthless-- if they were free they would still be overpriced.
unfortunately we are making laptops more like tablets these days, and we urgently need better hardware. its not impossible, but it is going to get worse before it gets better. but the status quo sucks a lot.
So, are the Purism laptops and tablets the proper temp replacement to wait for free and modern amd64 processors with Libreboot microcode?
So, are the Purism laptops and tablets the proper temp replacement to wait for free and modern amd64 processors with Libreboot microcode?"
Probably not. Newer x86 machine have fatal freedom problems. The long-term solution is probably to leave x86 behind entirely. See the link to Evacuating The Titanic elsewhere in this thread.
i686 or below MUST be left for sure that just don’t have futures. But for amd64 ones we need to OEM our own ones to be instructed under RYF and use Libreboot-ready microcode.
However, we also need to wait for current GNU free systems to be arm64-ready.
By future to practice our all theories we need to hurry up for Libreboot-amd64 processors OEM and arm64 ports for our current GNU free systems.
"i686 or below MUST be left for sure that just don’t have futures."
amd64 doesn't have a future either. Maybe it does from a purely technical perspective but not from a software freedom one. Check out the libreboot FAQ. It goes into more discussion of the problems. Making our own computers with those processors won't fix those problems. So the future really is leaving x86 behind because we don't have a way to fix the fatal freedom problems in newer amd64 machines, unless some miraculous breakthrough happens that lets us break the cryptography. Maybe we'll have quantum computers to do that in some decades. Haha. :)
So we need to think for OEM’ing our own amd64 machines, just based on the Technoethical amd64 mircocode as software base to establish our own RYF hardware workshop, but also start our own arm64 free systems (especially Parabola, GuixSD, LibertyBSD) hurrily, if the workhop OEM‘ing amd64-based RYF machines is totally failed then we focus on arm64.
> So, are the Purism laptops and tablets the proper temp replacement to wait for free and modern amd64 processors with Libreboot microcode?
Nope. x86 is a dead end freedom-wise. Don't be mislead by Purism.
So Purism is also mislead by Intel not to be honest for fatal freedom issues from their i7 processors.
I’m starting a project to OEM my own amd64 processors using the same libre mucode from Technoetchical.
But also please ask Purism to port their PureOS to arm64 before my project is mature and productional.
> So Purism is also mislead by Intel not to be honest for fatal freedom issues from their i7 processors.
I don't think that Intel has ever claimed that their hardware is freedom respecting, so Intel is not being dishonest. The freedom issues are well known. See here.[1] Purism is surely aware of them.
> I’m starting a project to OEM my own amd64 processors using the same libre mucode from Technoetchical.
Is this even possible? What is your plan?
> But also please ask Purism to port their PureOS to arm64
This would be good thing. All free distros should consider supporting non-x86 hardware. It's possible that Purism might not see it as worthwhile, since their priority is probably to support the hardware that they distribute, but it's worth seeing if they are open to the idea. However, you would be better off asking about this on the Purism forum rather than the Trisquel forum.
> I’m starting a project to OEM my own amd64 processors using the same libre mucode from Technoetchical.
> Is this even possible? What is your plan?
No. This is a misunderstanding. Technoethical doesn't even develop microcode. They sell machines with libreboot, which is itself not even a microcode. The format of Intel's microcode updates are secret although people are trying to figure it out for certain processors: https://media.ccc.de/v/34c3-9058-everything_you_want_to_know_about_x86_microcode_but_might_have_been_afraid_to_ask
But even if they do that in and of itself won't solve the problems with x86. One of the problems is cryptographic signatures being used. Even if someone can develop free code (somehow) the machine will refuse to run it since we don't have the signing key. Only Intel can provide that, although maybe at some future point quantum computers can be helpful to obtain it. But, by then Intel will probably have moved on to some other method that isn't vulnerable to Shor's Algorithm and we'll still be left with what would be (at that point) still old computers.
As an example, NIST already has a competition going on in search of post-quantum cryptographic methods with a goal to have draft standards available in the 2022/2024 timeframe: https://csrc.nist.gov/projects/post-quantum-cryptography Surely Intel would love to adopt something so as to frustrate such efforts of someone using a sufficiently large quantum computer to break the public-key cryptographic system that they currently use. Such an attack is, currently, only theoretical though.
The best that gd_scania's efforts can provide is something equivalent to what Purism does, which is to say machines that still have freedom problems.
gd_scania, I'm sorry that you don't seem to see this but x86 really is a sinking ship for software freedom in the long term. The problems are nearly impossible to solve, even if we make our own x86 computers, and no solutions are likely to be forthcoming within the next number of decades. It's time to evacuate the Titanic instead of trying to keep it afloat: https://jxself.org/titanic.shtml Please get your life jacket on and meet me on the boat deck. I hope you come to this realization before the Titanic sinks out from under you.
But I think amd64 and i686 are two ships, which i686 was already sunk and unrecoverable and over, but amd64 still needs our rescue, that's why I need to start an RYF project to OEM my own free amd64 processors with whose mucode sources are forked from Technoethical Core 2 ones. But where are hosting these mucode sources? I need to find them to start my project.
But also there are too few OSes having started arm64 ports from their amd64 distro. Only HardenedBSD has early arm64 projects, but yet to wait for GNU free systems to do so.
I'm here needing to ask Parabola, PureOS, GuixSD, LibertyBSD to be the first ones to establish our arm64 project. Especially if amd64 was over.
"But I think amd64 and i686 are two ships, which i686 was already sunk and unrecoverable and over, but amd64 still needs our rescue"
And there is the source of the misunderstanding. Modern x86 systems are not freeable, for all the reasons that have been explained. If trying this is what's needed for you to realize that x86 can't be saved, then all I can say is good luck, and I hope that it doesn't take you too long to come to the same conclusions.
As fully leaving amd64 we need to OEM our own arm64 laptops and NASes to start porting Parabola, PureOS, GuixSD, Trisquel, LibertyBSD to arm64 platform, where arm64 is quite recently mature but is still lacked of GNU free systems. So my IceOS project stated before for freeing HBSD will also start at arm64.
Trisquel will no longer be Ubuntu-based after ported to arm64 but will be independent GNU free system. :)
I have an HBSD-arm64 installer but not yet productional this stage.
I need to order few arm64 machines including few laptops and few NASes then install HBSD, ArchLinuxARM64, Gentoo for observations whether we should switch to arm64 for more active GNU free systems development at our future. :)
The final and most important thing is, to ask Purism for migrating to arm64 laptops and discontinue amd64 ones.
It's premature to stop supporting x86. Machines are still available, even if old, that can meet the FSF's RYF criteria and there are people using them. Since the goal is software freedom (not how new they are) it's still desirable to support those machines. In fact dropping x86 now would be more harmful than good because we don't have a lot of FSF-endorsed distros on other architectures like ARM and IBM POWER. And if we drop x86 what are we telling people on x86 to use instead while the hardware transition is happening? Non-FSF endorsed distros? I don't think so. I understand that Trisquel 9 will be available for ARM and IBM POWER but it will be a while (read as in: years) before dropping x86 is a good idea. And the good part is that these don't have to be mutually exclusive: During a transition multiple architectures can be supported at once: The old and the new. So there is absolutely no benefit to dropping x86 support at this time, and plenty of reasons not to.
I’m not proposing to drop amd64 ‘‘NOW’’, but I also have started a project to firstly start arm64 developments for the well known GNU free systems.
My another thread in ‘‘General Free Software Talk’’ also talks about starting these arm64 ports.
Currently proposing amd64 drop is always a prelude for this stage, and I won’t do things premature.
> [1] https://trisquel.info/en/comment/reply/24988/137353
I copy/pasted the wrong URL.
Even if they’re F-droid apps they still have anti-features, even if your devices are Replicant ones.
I also found that even if Replicant devices they still don’t have Replicant firmware (against nonfree ones) available, but just ROMs and recoveries for stock. :(