… and might require blobs?
1 Like
It has been possible for a number of years to install Ubuntu Touch, Sailfish OS and postmarketOS on select Android phone models. If they use a Linux kernel, they usually don’t fully support the hardware, as is the case with the Shift6mq. If they use an Android kernel and drivers with libhybris, there is a better chance that all the hardware can be supported, but the makers of mobile processors usually only release one Android kernel for each chip and never upgrade it.
In the case of the SHIFT6mq, its Snapdragon 845 processor was announced by Qualcomm in December 2017 and the first phones shipped with the processor in Feb. 2018 with the Linux/Android 4.9 kernel. Qualcomm has never upgraded the kernel, and will stop supplying security updates in Jan. 2023, so the SHIFT6mq will have an effective lifespan of 2.5 years (Jun 2020 - Jan 2023).
Initial support for the Snapdragon 845 was added to mainline Linux 4.18 in June 2018, but 4 years later there is still no support for the GPS and suspend and the cellular modem is only partially supported. If Shiftphone had decided to release the phone with Linux, it would have used the Android drivers+libhybris, because figuring out how to get the GPS, suspend and cellular modem to work with standard Linux would probably take years of work. Therefore, Shiftphone would have been stuck with Android drivers which would have only been supported for 2.5 years.
We need the chip makers like Qualcomm to support Linux, because phone makers aren’t going to ship Linux phones without support from the chip makers. Yes, we can install postmarketOS on 11 different phones with the Snapdragon 845 processor, but none of them are functional phones, so it is meaningless unless you want a portable WiFi device without suspend and GPS.
My hope is that someone will build a Linux phone that has a Snapdragon processor with mainline Linux support, and add separate cellular modem/GNSS and WiFi/Bluetooth chips.
4 Likes
There are three key issues at hand here. One is performance, one is thermals, the third is battery life, in use but also in idle (without suspend).
Performance can be approximated by the number and kind of CPU and GPU cores and the memory bandwidth the SoC can work with. Regarding the other two, node size is often seen as the holy grail – which it may not be:
While the AllWinner A64 (slow cores, poor GPU, poor memory bandwith), is made in 40 nm, the TI OMAP 4430 (in the mainline supported Droid4) is made in an at first glimpse worse 45 nm node, and has comparatively amazing battery life—it was designed for phones. It also runs cool, which matters a lot for phone designs.
With NXP iMX8M, AllWInner A64, Amlogic S922X, Rockchip RK3399 and also the RK3566 and RK3588 we have chips that weren’t designed for phones. I’ve seen reports that the Rockchip RK3566 is supposed to run cool without a fan or heatsink, the same applies to the StarFive JH7110 BTW, which both might land in future PINE64 phone hardware (I just hope that PINE64 will not name their RISC-V powered devices PineBook/PinePhone, because otherwise support threads will be hell).
The best chip with decent mainline Linux support we currently have is the Snapdragon 845; the best supported devices right now are the Xiaomi PocoPhone F1 and the OnePlus 6 (with postmarketOS; the latter seems much nicer, but is harmed by OnePluses proprietary charging crap, which requires you to put the device into EDL mode to charge at somewhat decent speeds). I expect the SHIFT6mq to overtake these two soon, but at least according to the postmarketOS wiki it’s not the best supported device yet.
All that said, I don’t really see massive issues on the hardware side. The PinePhone can work as a daily driver, if only the software is light enough—a surprisingly large number of “daily driving” users use Sxmo because it is light enough, other light candidates are Maemo Leste and Nemo Mobile, which likely mainly need additional contributors to make their software fully viable sooner. For me, speeds are okay with Phosh too—at least on DanctNIX and postmarketOS. You just need to breathe when in a hurry while waiting for Firefox to open, then it’s workable.
With the Librem 5, performance with Phosh is fine—the main issues are idle battery life and heat, which both are somewhat workable when you turn off that Redpine WiFi whenever you can.
We also should not focus too hard on the hardware, as it’s a war won by scale, something we’ll likely never have. It’s no surprise Apple is ahead (and with it TSMC): Tim Cook is a genius at scaling businesses, and they did so with their own silicon by also massively investing into the best foundry they could work with (TSMC). Industry giants like Intel and or (Android manufacturer) Samsung can’t keep up; and Linux Phones never be able to keep up either in terms of performance, battery life or the best camera.
If we manage to have “good enough” devices (a cool running RK3566 PinePhone 2 may be just that, and I hope Purism will manage to reasonably iterate on the Librem 5, too), and we manage to tell a story around privacy, digital wellbeing (I think it’s a feature to not have native clients for TikTok or Instagram) and sustainability (being able to use devices way longer), then that’s IMHO good enough.
6 Likes
I think the project has some good ideas about how to implement parts like the GPU and VPU in software, which will dramatically lower the costs and avoid patent minefields. I was impressed by how they were trying to engineer the CPU using old ideas from the 1960s to avoid patents. However, once they announced that they were switching from the RISC-V ISA to the POWER ISA because it is a freer standard, I knew that the project wasn’t going to go anywhere, because they weren’t being practical. Given the lack of recent updates, I think the project is effectively dead.
I actually think that it could work, if someone took one of the existing free/open source 64bit RISC-V cores and improved it. Get a couple companies like PINE64 to promise to build boards for it if you can get the funding to build it, and then use those promises to go raise a couple million dollars in venture capital to do the first iteration at a cheap planar node like 65nm or 40nm. After you prove that it can work with the first iteration, keep raising more capital to do it at successively smaller nodes.
@linmob, You’re looking at it in terms of whether the processors are good enough for Linux enthusiasts. I’m asking the question whether Linux phones will be able to attract more mainstream users who expect to replace their existing Android/iOS phone. For people like me, the i.MX 8M Quad is good enough, and even the A64 can work, but I’m a pretty committed user, who is willing to give up a lot in order to use a Linux phone. Most people, however, aren’t going to buy a Linux phone, until it is able to replace their existing Android/iOS phone (or it has some new functionality like convergence that their existing phone can’t do).
For example, most people expect to use their cell phone as their camera/camcorder, and aren’t willing to carry around a second device just for that purpose. With the i.MX 8M Quad, there is no hardware video/image encoder. NXP says that the processor can encode up to 1080p@30fps video in software, but that will use a lot of energy and generate a lot of heat, so I’m not sure if it will work in practice. Currently the L5 limits the temperature in its processor to a max of 50C, and I doubt it will be able to encode 1080p@30 video at that temperature, and maybe not even 720p@30 video for 30 seconds. With the RK3566, the spec sheet says that it can encode H.264 and H.265 in hardware up to 1080p@60fps, but I can’t find any info whether there will be drivers in the PinePhone 2 to support it. If the RK3566 has to do software encoding, what kind of video encoding can it handle?
With a Snapdragon or MediaTek Dimensity, I am sure that it can encode 1080p video without overheating or running down the battery. Even if there aren’t drivers to do it with the hardware encoder, the processor has enough processing power to do it in software.
If Linux phones have a decent processor, it could be the difference between the current Linux phone market of roughly 50,000 units per year versus 500,000 units per year. When more Linux phones are sold, the high software development costs are spread over more devices, so it costs less per device, so more users are attracted to the platform, which in turn attracts more developers, so more apps are created, which in turn attracts more users, and we get a rolling snowball effect.
If we guesstimate that mobile Linux needs about $5 million in paid software development per year to become a viable alternative to Android/iOS, the current market of about 50,000 phones/year means $100 in software dev costs per phone, whereas a market of 500,000 phones/year means $10 per phone. The reality is that probably only 10% of Linux phones are going to help pay for the software dev costs, and 90% are going to freeload. With the current Linux market of 50,000 phones per year and 10% paying for software development, that 10% would have to spend $1000 per phone. In contrast, with a Linux phone market of 500,000 units per year, that 10% would only have to pay $100, which is a more viable proposition.
I’m making up numbers here just to illustrate a point, but hopefully you can see why I want Linux phones to have better mobile processors.
Oh well, it isn’t the end of the world if the RK3566 is the best processor that we can get for a Linux phone. I actually think that it is pretty incredible how far Linux phones have come in the last couple years despite having subpar processors, so it isn’t a disaster if Linux phones continue having only a tenth of the processing power found in flagship Android/iOS devices. There will still be people like me who want to use Linux phones with any kind of hardware, but I do think that it will be much easier to attract new people to the platform if Linux phones have better hardware.
3 Likes
It’s hard not to agree with you points and arguments, as always, but they do paint this in a bit glum light. As I see it (and you too at the end of the last), this is a Linux phone and it will - for the foreseeable future - do different things than those other phones. It would be nice to have the latest and most powerful in any event, for other things or the same than what others are doing, HD video etc. Since that does not seem to be possible, I see this more as an opportunity (or a must) to take things to new directions.
Let’s list some:
- Convergence is one, possibly a big one. I don’t think we’ve seen even a fraction of what this can do and possibly become a driving force to get better processors. Who would try these first? Who have money to get these first? IT people. Who make decisions about hardware? See previous answer.
- Since energy is problematic, why not use smaller screens, make smaller phones (there is a market for those, even Apple has made theirs smaller) that use less energy via screens. Make most of what we have and lets stop trying to compete in inches. Especially if convergence.
- Security, privacy and networking can be taken to whole new levels - not just how much better they are but how much simpler, as well as modifiable, than what is offered via the big players. More internal political and/or nationstate competition turmoil coming, the more like networks and central services are going to be down.
- Linux is natively in a lot of devices and IoT. Can linux phones not have better access and controls to interact with those (special sensors, ports, IR, servers-at-your-hand etc.)?
- A few years ago we talked about how much ecological it will be to use a linux phone, which argument doesn’t seem to have disappeared. More like that just makes more sense now.
- Parts of the world (EU) looking at energy prices and others noticing that datacenters are taking up huge chunks of regional and national power production (to the level that some development has had to be halted), a device that doesn’t send much data to and doesn’t need centralized services seems like a solution that many may want to take part in… motivating to want to use “non-flagships”.
There is much to do still and plenty of room to grow, do business and create FOSS. Your point stands and it would be nice - and indeed needed in the long run - to have better mobile processors. But it’s not the whole thing. It’s not the whole linux phone.
1 Like
I think going forward Purism if they release a second iteration of the Librem mobile device should take a page out of Fairphone, not necessarily the eco-friendly obsession as the priority (they can if they choose though), but a device that has replaceable parts and take it to another level where perhaps it could even be modular. I mean Librem 5 is probably the thickest smartphone on the market right now, use that to your advantage, make it more modular. I mean even right now you can swap out batteries, SD cards, and modems, so it does let you open it up, but a modular device would be awesome. Purism could upgrade the parts every few years so they can still make money. Sort of like people that upgrade their PC’s every few years.
1 Like
But, why is Power ISA considered to be such a dead end? I remember the hype about Power 9 PC chips … seemed like momentum is building, yet the clamour seems to have waned.
i’m not sure what all the hardware L5 users are deriving their experience from…but it must be a different L5 than the one I received lol
I haven’t used my L5 a whole lot, but the little I have used it… it runs way slower for just about everything compared to even older android phones i have or have used.
not only does it run apps slower, but the battery makes it pretty much unusable for me
the whole argument of “why do you need a faster cpu?” is not good IMO… Yes, they are working on fixing some of the speed and battery through software, but come on… Even after fixing some issues through software, the phone is STILL going to be slow compared to older android phones. I mean just look at the processor specs!!
if the L5 is this slow NOW after just being released, i imagine within a few years time, its going to be even more laggy and slower than it is now. i used a Note 3 all the way from 2013 till today, and it just started getting slow on apps within the last 1-2 years or so. I DONT see an L5 making it even more than a few years.
i get that people want to be supportive of such a project/phone… i certainly do as well, otherwise i wouldn’t have backed it. but, the truth is the truth. how old are android phones that have comparable specs as the L5?? probably about as old as my Note 3 or maybe just a little newer.
it sucks we can’t get decent hardware just because its a small project. i want a phone with comparable specs, no matter if it can run a little better on linux (which the L5 does not at this point). i dont want a low spec laggy battery hog phone that is going to be obsolete in a few years because it started out with old low spec hardware, ESPECIALLY FOR THE PRICE IT COSTS!!
me personally, i buy a phone (especially an expensive one), i pretty much use it for as long as i can.
i dont see the L5 lasting too long in terms of processor speed, app lag, and battery.
none of the above ^^ even takes into account the fact that the phone is pretty much unusable at this point as a daily driver, and probably will be for a few more years
this has been my experience anyway. i will probably sell my phone, and come back to the project (and similar projects) in a few years to see if its more mature, and has better spec hardware.
1 Like
That doesn’t have to be the case. I totally understand that software can become more bloated and software can become slower due to a bug’s getting fixed but, equally, software can become faster if someone spends the time to optimise a few things that were previously fairly modest and generic in their initial implementation. (In the most extreme case, it is even possible that debug / trace code was left in because it is early days but that can be disabled later on.)
As a hypothetical separate example, imagine that hardware acceleration of graphics is currently disabled because it doesn’t work (reliably) yet and then suddenly someone spends the time to fix enough crashes and it becomes viable to enable it.
Yep, the specs are not state-of-the-art … but what matters is actual performance, and actual performance doing the things that you actually want to do.
For me, the web browser is not pleasantly fast and that is an area that I hope improves in the future. Most other things are responsive enough for me. So the question has to be asked: why is the web browser not very responsive? It may well be that you can kill the problem with hardware i.e. a future faster processor, but that may not actually be the right solution, unless someone has looked into it.
3 Likes
i like your optimism, but i’m not.
a phone that has a tenth of the hardware specs, but cost 5-10x more?!
i love linux, and i love the idea of this phone, but not that much…even if i were rich and didn’t have to worry about money, i still wouldn’t use the phone in its current state. its just not a good phone currently with its current software/hardware. its slow laggy compared to older android phones i’ve used, the battery is absolutely horrible, its buggy still, and basic functionality is missing.
i can see using the L5 as a little “hobby project” computer to play and hack around with, but not as a daily driver phone. its just not in the same category currently, as least not for me.
as a daily driver phone its not very good, but as a hobby linux computer its a little better… but for just using as a play computer, i would still use something else personally, like a small laptop loaded with linux or something like that, or even a faster android phone loaded with linux
honestly and sadly, i dont think we’re likely to EVER get a competing comparable linux phone phone. big hardware/software companies dont want to lose market share to a linux phone…they’re not going to allow such a project to get traction i don’t htink.
just look at what microsoft is currently doing! they’re trying to get microsoft crap embedded into cpu’s so that linux and anything non-microsoft approved can even be installed on the computer…at a hardware level!
things aren’t getting better and with more options in technology and repair ability…in fact things are going quite the opposite way, more monopolized, less big companies with more control, more expensive, less freedom to repair, fix, customize pretty much every kind of tech now… <- in this regard, its good to back projects likt eh L5 if you can afford to and don’t care about the usability
maybe what purism (and other linux / freedom type movements) needs is some rich donors that are willing to donate big money to actually change things for the betterment of such tech. instead of having users that mostly dont have a lot of money…pay big money for lower tech… we need some big companies or rich individuals contribute to get the ball rolling for faster development AND better, more competing, higher spec tech.
but then, i start to wonder…IF purism received a large grant or donation for free that suddenly gave them access to better/faster tech and a lot more developers… would they even lower the prices? i don’t see that happening either to be honest. oh well, all we can do is wait and see if anything comes about with this tech/movement
2 Likes
the problem i have with this line of thinking is…
even if the sofware gets optimized…realistically how long can a person use the L5 at current specs before it gets even more laggy and slow??
software can only do so much. the saying goes “you can run linux and any old computer, its not windows”…this is not true for me at all. i have older computers (not that old though), that no matter what distros or drivers i installed on them, they still ran like shit.
fixing software is only going to go so far for so long IMO
Yep, the specs are not state-of-the-art … but what matters is actual performance, and actual performance doing the things that you actually want to do.
yes, current performance certainly does matter…and current performance and battery is a joke…lets be real here…we’re (or i) have been waiting for the L5 since the kickstarter. after a few years, the phone development has A LONG WAY TO GO. its probably going to be another few years until the L5 IS EVEN USABLE. the software is buggy, lacks A TON of features/customization, and pretty much doesn’t work for daily driver tasks
after a few MORE years of developement, will the L5 even be hardware relevant at that point?
its already losing a following IMO also. lets look at the resell value for a moment. when the phone just arrived to peoples hands, the resell value was $1000 - 1200+ or so? NOW, the phone is selling on ebay for not much over retail price…EVEN THOUGH PEOPLE STILL HAVE A LONG WAITING LIST TO WAIT…what does that say? to me it says, theres hardly any demand for the L5.
look at the raspberry pi’s as an example. they are out of stock and people can’t get them now, so the prices have 2-3x on them, and people are actually spending 2-3x for a RP…there’s demand
the L5’s waiting list is how long now? is it still 1-2 year wait to get a non US L5?? yet, the prices for resell are selling for about retail price. not good demand or resell value on a product that is unavailable and has a long waiting list
i think the reason theres no demand and no resell value…is because the phone is way overpriced for something that has way lower specs. pretty much…if you buy an L5 you’re paying for an “idea”. you’re paying for the dev work of “that idea”. and you’re paying a hefty premium for low spec hardware because of that idea. thats the way i see the project now anyway
its like @amosbatto mentioned, hardly any people (except hardware enthusiasts) are going to spend a ton of money for something with crap comparable tech and far behind buggy software.
thats why i brought up the idea of purism needing some big rich donor sponsors…instead of getting mainstream peoples (who are mostly broke at this point) to pay for “that idea”, we/they probably need to find a way to reach some big sponsors
1 Like
Yeah, BUT: even 25 years ago company execs were laughing about Free Software (and Linux) and look at the world now. Free software has revolutionised the world of computing. Who knows, maybe Free Hardware starts the second revolution? The problem I see is that lots of stuff gets virtualised and what people use more and more are just services, not programs on their own computers. Other problem I see is while smartphones have been the staple of personal computing (and many other things) since the first iPhone, mobile and personal technology is diversifying. This can be both good and bad. What we’ve seen so far is that playing catch-up usually leads nowhere (Librem 5 may still disprove me, I hope it does). Perhaps it’s time to take initiative and do some rethinking. Perhaps packaging decentralised services for easier use, open hardware virtual reality platforms, etc etc? That’s just my whishlist.
1 Like
Yeah, Elon Musk could have invested 1 billion USD in a Linux phone instead of intending to buy Twitter for 40 times this amount.
Apart from that. Guys, stop taking the price as a measurement of the product costs.
The high price of Librem 5 means:
- Elegant way to reduce the number of orders as the backlog was already way too long and people are already waiting for years to receive their phones.
- Elegant way to motivate people to go for the Made in USA model. The lower the difference in price, the higher the motivation. 2000 compared to 1300 looks way more attractive than 2000 compared to 600.
Even if Purism could sell L5 for let us say 300 USD, this would not make sense as then the orders would skyrocket and they will sell faster than deliver.
And we have to take the topic of security of supply regarding electronics very seriously.
Nobody cares about security until a catastrophy happens.
I am happy for all your 10 times faster Chinese made Android and iOS smartphones. But let us imagine that one day we have a Third World War and suddenly you can’t buy a smartphone or a notebook anymore just because nobody in the West manufactures such products.
And the ones that you already have stop working due to a large scale malware attack using either flaws in some closed source apps you use for convenience or uses even some hardware level backdoors built in by our beloved Chinese manufacturers.
Yes nobody cares until it happens. It is like Chernobyl or Fukushima or the war in Ukraine. Until it happens it is very very very unlikely. But when it happens, it is suddenly a reality and a mess to solve.
5 Likes
Before picking up a MNT Reform I would completely agreed with @amosbatto here. Even after doing that and using the device for over a year, I still partially agree with him.
BUT…
What the Reform has shown me is that the SoC in the Librem 5 is more capable than one might initially imagine. I can use my Reform for all but maybe 5% of software developer needs. With SSH/VNC I can complete the rest of that. The point is, while the SoC is capable of so much, it just doesn’t do it as fast as the competition.
The Reform uses a single PCIE lane to enable NVME based SSD storage. This is something that greatly helps the device to feel faster as the CPU is not waiting for data as much.
So while I’ve been notified that my Librem 5 should be shipping any day now, I am VERY excited about it. I know that I will be able to use the L5 just as well as I can use the Reform. For me it really will be a pocket computer and I can’t wait.
The bottom line for me: The SoC being used in the L5 is good, and still has a lot of life left in it. It really can be used as a convergent device.
However, any device can be improved by a processor bump of sorts, and this is no different. I’m just not suggesting that this needs to happen in a year or sooner. We have some time still, and I think that there will be no reason why the L5 will less useful than it is right now in 2 years. If anything, and if history and statistics tell us anything, the exact opposite is going to be true.
That is a big plus for the Librem 5.
8 Likes
Aren’t you afraid one of those computers is going to start collecting dust after you obtain Librem 5? 
PS Just checked pics of MNT Reform … I hate the right SHIFT key already (but, I do like the mechanical keyboard).
No not really. The Reform really has an excellent design, and the mainboard is designed to upgrade the SoC. I think it will have a long life still. I need to find a good keyboard to use with the L5 though.
3 Likes
Let me correct this. I did a web search and I found out that the libre-soc.org actually did submit files to TSMC to make a 180nm test chip in July 2021, but the only info I could find about it is on this archived web page:
https://web.archive.org/web/20210812174204/https://openpowerfoundation.org/libre-soc-180nm-power-isa-asic-submitted-to-imec-for-fabrication/
I couldn’t find anything else so I’m not sure if the chip actually got made, but it looks like Luke Leighton has gotten enough funding to keep working on the libre-soc.org processor. The chip doesn’t even support floating point numbers, so the design is definitely in the early stages. Right now it has roughly the complexity of the original 80386, and it is using a process node that is 24 years old. I wonder if it will be possible to keep using free/open source tools at smaller node sizes.
Power ISA has a bit more legacy cruft than RISC-V, but they are both clean and streamlined ISAs compared to x86. POWER9 processors use a lot of energy, so they were only good for servers and workstations. IBM open sourced its experimental Microwatt core based on Power ISA, so in theory the ISA can be used for low-energy applications. Currently only IBM and libre-soc.org are designing new chips with the Power ISA. In contrast, there are over a hundred different companies designing chips based on RISC-V, so it is the ISA generating all the buzz.
Initially, I thought Luke Leighton was not being practical when he decided to not use one of the existing free/open source cores for RISC-V, but when I read more about his hybrid CPU/GPU/VPU vector processor, I realized that he did have to start from scratch, so he might as well use an ISA that doesn’t require signing an NDA and is more open to additions from outsiders like Leighton.
I don’t know anything about chip design, so it’s hard for me to evaluate whether Leighton’s experimental design is a brilliant idea or not, but you can read the details here:
7 Likes
Thanks for attaching the presentation. On slides, this looks awesome. It seems that it was the funding which prevented use of RISC-V (because of NDAs). Interesting. Well, I hope IBMs ecosystem wins this time round. Similar situation to that from the mid/late 80’ and the proliferation of the “IBM compatibles”.
Maybe @lkcl could shed some light on this, I’ve been curious about what the outcome of that 180nm chip was too.