Apparently it is based on Halium, so it's probably not going to run mainline Linux any time soon. On the other hand, I'm less happy with Android than I pretty much ever have been, so having some more options outside of relatively unusable Linux phones and being stuck with modified Android firmware certainly does not hurt.
I am definitely interested, especially if the team working on it is as active as it seems they are. Definitely one of the biggest bummers with PinePhone and PinePhone Pro was realizing that the community was largely on its own; that could work well, but in practice progress has been slow and painful and I stopped paying attention.
I would really like to see more projects doing cellular Linux devices in general. It doesn't need to be a candybar smartphone, if someone can jam a cellular modem into a MicroPC I'm sold at any price I can afford.
I work on this device. We are currently QAing the latest update with our community, which contains a lot of goodies and improvements shaped directly from community feedback. You can see previous changelogs at https://furilabs.com/changelog
I am obviously a little biased, but this is probably the first Linux phone that people can actually use. Someone in our community switched from iPhone to this without much of an issue.
Interesting history! "Motorola became the first company to use Linux on a mobile phone when it released the Motorola A760 to the Chinese market on February 16, 2003." https://en.wikipedia.org/wiki/MontaVista#Mobile_phones
Right, used 2 of them as my primary phone between 2015 and 2024 when the hardware broke not many weeks ago. Just recently ordered their new one, which is a bit cheaper than this Furi. Admittedly a risk purchase because not much is known about it yet.
Also typing this on the ultra-cheap hardware made for the Indian market nearly a decade ago. Well, admittedly they have already announced to discontinue support and I never dared to carry it as my primary phone.
Now the last weeks I was forced to use Android as my primary phone. Really a suprise what crappy piecemeal Android is compared to SailfishOS made by a tiny company struggling economically forever.
You could always protect the signing certificates in the apps with derived passwords, still the length of passkeys practically acceptable to type in on a phone is too short to safely protect a certificate vs a bruteforce attack without some kind of HW assisted storage.
In the end it also boils down to what devices the BankID app providers are willing to support, I have a hard time seeing anything but iOS or Android devices being supported in the near future, Esp as Swedish BankID's now also requires NFC support to read the local police issued ID cards (had to get a new testing-device just due to this requirement).
Note: BankID is the name of personal identity apps that support authentication and signatures in Sweden, Norway and Finland, the authentication is used to access a myriad of both public and private sites like tax office, unemployment, healthcare and gyms. The signatures done via the apps are generally accepted to be of as good legal standing as a signed paper.
I would also have commented: Not supported in Finland. I think I have read articles that they tried to get marketing share, but AFAIK they have failed completely.
On the positive side in Finland you can use SIM Toolkit for legally sanctioned 2FA (mobiilivarmenne). That should be much easier to implement without having Google involved.
Since you guys are running Gnome, you might be interested in helping with (or funding) this native GTK WhatsApp client: https://github.com/tobagin/whakarere/
Since these days something like >90% of people is on WhatsApp, it’s beyond quintessential to have a good WhatsApp client.
My biggest hold-up with that sort of work is that Facebook has previously threatened to ban users who use third-party clients, and then they have gone through with the threat. Is this no longer the case?
Meta also threatened app makers to start a lawsuit. Whatsapp Plus, Whatsup and Mitakuuluu people stopped on that threat. I don't see a reason why Meta would change their stance on this.
There is now a law from the EU, I think it is the DMA, which could change things for people living in the EU regarding interoperability. I haven't read anything about practical follow-ups yet.
By the way, only releasing source code as a blueprint might be fine, that happened before. I might remember it wrong, but it might be lame, the mp3 encoder.
I can't speak for everyone, but I've been using WhatsApp since ~2010, mostly the official client but with bouts of unofficial client usage, and I've never been banned.
Probably not very detectable for Meta since it uses a wrapped webview, but I've been using ZapZap on Linux for a good while now and I haven't seen a ban yet.
if something can happen and haven't happened to you, doesn't warrant a comment i believe. it doesn't help much to figure the data points for the cases were it does happen, since it's implied everyone to whom it have happened would have lots of opportunities to post comments it never happened before.
Has a well-behaved third party client ever been an issue?. A third party client that does things like retain deleted messages, report that read receipts are enabled but never actually sends them or other 'anti-social' behaviors is definitely asking to get banned.
[1]: In July 2021 Facebook sent the developer of Unfollow Everything a cease-and-desist letter because he made a browser extension which helped people use Facebook less.
[2]: In 2020, Facebook shutted down the NYU Ad Observatory by shutting down the Facebook accounts of Cybersecurity for Democracy team members Damon McCoy and Laura Edelson.
Easily probably not the word I'd use. Anything that runs on wlroots should be doable with not a lot of effort. Beyond that, tricky. sxmo doesn't use hardware acceleration though so not too bad.
Is there any phone that runs mainline Linux/upstream Debian, and is available to buy right now?
I'm sick of big companies who removed the word Consent from their dictionaries. I want something I can own and trust again. Being able to run android apps would be great, but given how closed source Android has become, I doubt that apps run on an open system would be reliable.
100% mainline - not really yet, but from those that are close to mainline and on which you can run Mobian (which is upstream Debian with a small repo overlay for stuff not ready to go into Debian proper yet) and that you can buy new right now are Librem 5, PinePhone and PinePhone Pro.
If second market also counts then there are a few more options available - mostly former Android phones where hardware support is spotty though, so do your research first. That said, everyone's needs are different and what's there can already fulfill the needs of plenty of people, so it's a matter of figuring out whether you in particular are in this group yet or not.
Halium is based on libhybris, to my understanding that was developed when Nokia stopped making Linux phones and some of their developers started making their own under the name Jolla/SaifishOS in 2011.
While Nokia could maintain their own Linux kernel that's impossible for a small vendor/team. There is no realistic alternative to taking the Android kernel provided by the SoC vendor.
The SDM845 stuff looks quite promising for mainline but unfortunately the platform is definitely showing its age (and camera is still completely broken on every device as far as I know...)
I want to like it... But, I've been bitten too many times by expensive Linux phones (Oh Nokia N900, how I wanted to love you. but you couldn't even make phone calls without threatening to crash).
It's a little weird that there are very few details on this thing. Notably, one of my #1 concerns when looking for a new phone today is a good camera. And all I can find on the camera is "it's 50MP!!! " But how do actual, non-photoshopped pictures look? Does it do okay in low-light? Does it support any sort of optical zoom?
Their website is currently as slow as a frozen snail (which, if they can't run a basic web server, leads to a lot of doubt they can maintain a long-term Linux distro). So I couldn't click around much, but I didn't find anything really informative.
I tried looking at YouTube videos and only found some weird official videos of a person's hand putting the phone in a sink. Like... Yes, it has been expected for the last 8ish years that my phone will be waterproof, that's not an amazing achievement.
It also seems to be "weird" Linux, like Maemo was for the N900. I don't want weird, custom-built stuff that will be forgotten about in 12 months. Just give me a standard Debain/Fedora/Whatever and an unlocked bootloader to reinstall with what I want.
IMHO they should really look at what Valve did with the steamdeck and SteamOS, It's a custom Linux that works well, but, it's also standard hardware with an open bootloader. There are dozens of SteamOS alternatives.
I agree with you on some points - I do know a lot of people (myself included) have been bitten in the past many years by Linux phones that didn't do what they really claimed they'd eventually do. There's not much I can do there other than continue to work on it and make it the best possible Linux phone. It will speak for itself.
I also completely agree that we should have a camera showcase. We've been talking about the idea of letting customers upload their shots to some sort of gallery so people can see how the real world looks through the camera's lens. What I can tell you though is that pictures look pretty darn good, and low light is great owing to the f/1.57 aperture. No optical zoom for this model unfortunately.
I'm not sure the website being slow speaks to our capability of maintaining a distro, especially when you consider these two things are done by different people. Sure, I'd love to work on the website and make it faster. But I'm busy working on the actual phone.
The video situation will improve as reviews come in. As for just giving you an unlocked bootloader and regular Linux... sure, I can do that. But you're not going to have audio, hardware acceleration, camera, etc. postmarketOS is a good example of this - despite years of tireless work, it's just really hard to get a regular distro working on a phone without leveraging Android drivers, at least for now.
>postmarketOS is a good example of this - despite years of tireless work, it's just really hard to get a regular distro working on a phone without leveraging Android drivers, at least for now.
Does this mean you are not working on mainlining this device? I'll take the slow and steady pmOS approach over sketchy Android shims. That's what ruined the Planet Computers Gemini PDA for me. A temporary solution is often surprisingly permanent.
We are not actively working on it, although some work has been done in that direction as a weekend project. It's not something we are ruling out, but we want to focus on delivering something that works, and works well. If the product does well enough to be able to finance continued effort in that direction, and there aren't any "massive" issues to work on before that, then... I guess it might just happen. Hard to tell right now.
> It also seems to be "weird" Linux, like Maemo was for the N900. I don't want weird, custom-built stuff that will be forgotten about in 12 months. Just give me a standard Debain/Fedora/Whatever and an unlocked bootloader to reinstall with what I want.
I wouldn't say it's a "weird Linux as it's mostly Debian and it's fully rebuildable by the user from the open source scripts[1]. I would likewise love a phone that was just mainline Fedora, but given many of the current challenges around ARM devices with linux, I fully understand why a "custom-built" OS is needed. As long as it's as light-weight as possible and is open source/rebuildable by the user, I think it's a good compromise given the realities around the platform.
Maemo was "mostly Debian" as well, it was one of the big selling points. Except Maemo was quickly abandoned by Nokia and left unmaintained. To further complicate matters Maemo required arcane tools to build it and install it to a device. Custom community kernels quickly became standard for enabling basic functionality which Nokia should have provided, but, didn't. However, the custom kernels were also frequently left unpatched and abandoned because they were being built and maintained by random internet users who found other hobbies.
I think the idea is awesome and I am excited if not skeptical. I hope Furi manages to build a healthy community around this. And doesn't go the path of so many previous Linux phones.
n900 was perfectly reliable, yours had a problem. And Maemo was a wonderful interface for mobile stuff. Clean, functional, tweakable. Regret it every day, facing shitty Android. I dream for a LTE n900 with a larger screen, leaving the rest unchanged : unbreakable resistive screen, mecanical keyboard, fantastic stereo speakers, etc.
> And all I can find on the camera is "it's 50MP!!! " But how do actual, non-photoshopped pictures look?
It's probably all they know about the camera. Not even Apple make their own cameras. It's always someone else's (customized)product. If they're outsourcing hardware, they might not even know the customer code assigned to the contracted OEM.
> I don't want weird, custom-built stuff that will be forgotten about in 12 months
Then none of hardware and many of software features will work. There are no in-box standard implementations for e.g. phone call in GNU/Linux, so it'll have to be supplied in the form of a custom preconfigured distro with bunch of free preinstalled apps.
The fact that product necessitating free preinstalled apps to function leads to needs to supply OS as flavors of your own weird unsupported distro, like N900-specifix Maemo or Steam Deck OLED-specific SteamOS images.
You aren't going to like being instructed how to install and patch a SIP softphone just to make calls on North American carriers with vague hints for Canadian specifics. It's a phone, it'll be lame if it doesn't make calls out of the box. And there comes the 3 years old $DISTRO-$HWREV-$VERSION.img that instantly shows antenna bars right on the setup screen.
Actually I think it's unfair to call Maemo a weird unsupported thing. It didn't seem that far from bone stock desktop Linux.
It's surprising to me that there aren't any details about the software stack.
Whenever I've tried to find a touch-optimized Linux shell, I've come up short. As has been discussed on this forum in the past, most of the attempts are understaffed side projects that have thousands of manhours of work ahead of them.
I'd like to know what shell FuriOS is based on, and what they're doing to make it usable. Even a demo video would go a long way!
As someone unfamiliar with Furilabs, it wouldn't occur to me to go looking for a dev blog to see if you've written anything about the shell. I feel like that's information you'd want to link from the marketing page (even if the details are on your blog).
Recent phosh is packaged on Fedora. I’m critical but also pretty happy with it. Biggest disappointment is that many gnome apps are not as touchscreen friendly as was promised.
I might be interested once there's videos of it doing daily stuff, and especially of how its autofocus (speed and functionality) compares to, let's say even an old Samsung Galaxy S7.
We work with the same ODM that supplied the hardware for the Gigaset GX6. We do not resell Gigaset phones directly, but we do use the same reference design.
I think it's fair to say that whoever buys this is in for a bumpy ride. I mean, that is to be expected, however, I think they should be more honest about the current state, like for instance PostmarketOS does at https://postmarketos.org/state/
I'm currently using a Volla 22 with Ubuntu Touch and I'm happy with it. However if Furi is using Plasma Mobile (they're mum about what interface they're using, but if it's really a 'lightning fast user interface' then I don't know what else it could be) then it's unfortunate that I discovered this too late to get it instead.
Can anyone estimate what it would cost to develop a new phone that's like the N900 with updated hardware (slide out keyboard, camera cover, maybe drop the input stylus, a bit flatter and larger) and mainline Linux support for all components?
I'm sometimes wondering if this is something I could have Pine64 do if I accidentally get rich enough. I think I would spent the rest of my days just writing the software for this.
First question is; which SOC would you want to use? Is there any SOC with full mainline support that you would agree to? There is a reason Pine64 and Purism went with SOCs out of the ordinary path.
I believe everyone has similar fantasies. Your best bet is a custom made mini laptop with a x86_64 SoC with perhaps 18650 batteries or thin easily replaceable battery packs..
Considering there's a bunch of stuff you'd have to manufacture from scratch I'm gonna say... 500 million USD? If you're okay with outsourcing a bunch of stuff to China and treading on gray areas where patents are concerned. Unfortunately it's gonna take a few more years before various mobile phone patents start expiring.
$500 price point is a lot for an open source enthusiast phone, especially one unknown.
The Pine64 PinePhone is $150-$200, with similar hardware provenance, but various open source projects that might at least give better software longevity and software trustworthiness. https://pine64.com/product-category/pinephone/
An additional consideration: as much as we like and respect our global colleagues, unfortunately I could see both hardware brands getting banned in some locales and by some organizations, like some other telecom brands have been banned. A phone that can't be used as a phone has little value, no matter how open it is.
For somewhat different provenance, the Purism Librem 5 is coming in at $800+, which is too high a high price point for most open source enthusiast contributors. https://shop.puri.sm/shop/librem-5/
I'm thinking that better for the Western open source developers and non-wealthy enthusiasts would be something more palatable to Western governments, and usable as a daily driver, like Librem 5, but priced somewhere in (guessing) $200-$400.
no not at all Pine64 hardware is pretty crap (performance wise, very normal independent of the OS at that price point)
Like you are comparing things like 3GB LPDDR3 SDRAM (Pine64) to 6GB LPDDR4X which is world apart (not only does it move from a "very tight on memory" to "somewhat acceptable memory" (very big deal) it also is much faster (~2x).
Similar on the Pine64 you have 4 old little cores here you have 2 big + 6 little cores and all more modern (end even just the little cores are faster when comparing one core with another, ignoring that you have 2 more little cores and 2 big cores).
Storage moves from a quite tight 32 GiB (through I guess if you only do stuff like mail it might be enough) to a likely more then enough 128GiB (also faster).
Noticeable faster GPU too, I guess.
And way better camera, at least by by numbers.
So not comparable at all.
But what matters most is not the absolute differences but moving it from a "so tight on RAM,CPU,GPU,Storage that you run all times into problems" to a "enough for all common tasks". I.e. from you need micro optimized apps for it to work nice to yes it works if your app is imperfect territory. And that is a pretty big deal.
But also lets be realistic. Atm. _any_ Linux Phone is a pure enthusiast product balancing various problems (price, small production lines, only android diver problems, proprietary blob problems etc.). And many of the enthusiast earn enough money to afford a 500$ toy, through maybe just rarely and with thinking twice about it.
The original pinephone was an allwinner soc that had like 4x a53 cores, a weak GPU, and pretty meh ram. This phone has a fairly modern mediatek soc with a78 cores, a fairly modern Mali GPU, ufs storage, and lpddrx5 ram.
It says "Wired/Wireless and NFC combo" under battery charging, but there's no mention of NFC under "Connectivity". Is it charging only? I'd like a linux phone but the Purism phone eschewed NFC too.
There's NFC, although the Linux stack support for NFC is a little limiting. However the NFC hardware is passed through to the Android container, so if, for example, your bank doesn't demand that you pass SafetyNet, you can pay with the phone. (I have done this)
I've used nfc on linux with several readers via pcsc-lite and it seemed to work fine. Do you mean support for the nfc chipsets used on phones? Does the android container deal with that differently somehow?
That's a good question, sorry I missed your comment.
I honestly don't know of an "easy" way to do that. If you have enough time on your hands you could decompile it and take a look. You could also try running it on a spare phone with a ROM that doesn't contain Google Play Services (or where it's been disabled). However, SafetyNet isn't the only line of defense. These apps often have their own root checks, which doesn't pose a problem for us because the Android image isn't rooted. (But it runs inside LXC, and you can attach a root shell into it.)
Another thing they don't like sometimes is the system build fingerprint not being recognized. We have an overlay(fs) system that allows users to customize everything inside the Android container without the hassle of rebuilding the system images or having to stay on the ball when we update things. It's not hard then to pick up the build fingerprint line from a "trusted" device and just drop it in build.prop.
All of this is a bit of a hassle, to put it mildly. But if we try to talk to any bank or contactless payment provider right now, they'll tell us to kick rocks. I guess it's a matter of time until we grow or a very smart person figures out a very clever hack. :)
Something can be fast, but lack efficiency or the ability to utilize the speed.
For example, a car can have 1,000hp, but utilizes a large turbo which takes a long time to spool up, leading to poor performance. It may also have poor handling characteristics.
On the other hand, let's use a car like a Mazda Miata, which only has 181hp. It is not fast, but it is an incredibly performant car, with exceptional handling and driving dynamics.
Speed and performance are not directly correlated, and are independent metrics. Performance is more of an indication of the overall dynamics of action, intention and results.
280 grams is a lot even for today's phones. 12 mm is half a inch. Very thick.
The other dimensions are 6.7 x 3.2 inches.
Ok, this might be a phone for people that want to prove a point but it's too cumbersome. How about starting with something reasonable and fit it with only what fits within those bounds? An example: 160 x 80 x 9 and 200 g. That would be a big phone but not bigger than many phones on sale.
> Ok, this might be a phone for people that want to prove a point but it's too cumbersome. How about starting with something reasonable and fit it with only what fits within those bounds? An example: 160 x 80 x 9 and 200 g. That would be a big phone but not bigger than many phones on sale.
Because in that universe your comment would be about how awful the specs were and how tiny the battery is. I'm reasonably confident that price-size-performance is a "pick at most 2" situation.
It takes lots of engineering effort to make things svelte. That makes it incredibly hard for "indie" phone makers to come up with something competitive in this regard. Likewise with price --- Android phones with similar specs can be found for only $200.
Can you actually do this with the numbers Furilabs probably expects to hit? Surely there's a minimum number to order before the companies you mentioned will even take you serious.
Is there prior art that includes a removable battery, headphone jack and IP68 at those "reasonable" bounds? I wonder how large the battery could be in that case.
Do you seriously think weight and thickness are parameters that hardware designers just artificially inflate all the time just to be mean to users???
You want 1/3 of thickness and 53ml in volume and 80g in weight just removed because methinks that'd be much better??? A phone isn't a steak cut, you... can't do that. You can if you knew beforehand that there are suitable replacements for major components like displays and batteries, if you don't, you just take what hardware engineers give you. What are you even thinking.
I'm so desperately looking for a DIY portable device that does not suck to build my portable audio player, best case supporting android with any of the following custom roms:
LineageOS
CalyxOS
GrapheneOS
/e/OS
I even considered making a player myself using the T113-S3 SoC, unfortunately not having the required skills. There are so many people trying to build prototypes, but this seems to be a space that is not easy to manage.
Here is my personal favorite Audio Player base device, best I found so far:
Thank you. Yeah I did know that. For my use case it unfortunately is too clunky and it does not have headphone remote (afaik), but I really considered buying one, just in case it gets added.
In Android, headphone remote works out of the box... that's why I would like to use one. There is the HiBy M300, which in my opinion would be worth having a custom rom :-) Indeed there seems to be efforts.
Took me ages to connect the fact the phone was flipping around at hyper-speed on the webpage because I had the cursor over it whilst scrolling down. Had to move to the right side of the page to then scroll down.
It says "F-Droid, Aurora Store, and other sources'" so I'm guessing there's no Play store and probably no GMS and other increasingly mandatory proprietary dependencies. Open source Android is in rough shape as a result of Google's tightening the ratchet.
Edit: In a different comment one of the devs clarified that it does have microG, so quite a few apps will actually work.[1]
Afraid not. These things usually require hardware attestation to ensure you're running the exact software Google wants you to run. Now, if someone figured out a way to bypass it, that would probably work for a bit. But it's a cat and mouse game.
It's not just what Google wants you to run, it's what Google promises to some of its customers through a face it has on one of numerous facets(facing banks in this case) as what they let users do with phones, and a humanware chain of trust that allows the server and the phone to be viewed as sort-of secure computing environment.
From an idealist standpoint, it shouldn't be done that way by taking away rights from users, but trivializing it as how specifically Google wants you to do it isn't accurate either.
Can we get a Linux smartphone that works before worrying about nonessential use cases? One the powers-that-be will never support until step zero is shipped at volume.
Why does this question always come up? There's zero technical reasons you can't electronically pay with all-open software stack e.g. Ethereum on GNU/Linux, it's pure matter of trust.
With currently available payment standards, it's like asking for a live credit card with someone else's private key and an open debug port, all at the same time. Things don't work that way.
Because that's the only question that matters. I have just a few apps installed, some of them are easily replaced, like TOTP apps, others are impossible to replace and without them I might as well get a feature phone.
I'm down to one app that I really do need, MobilePay, but it's a payment app, heavily controlled by the banks. That is not going to be installable on CalyxOS or anything without either Apples AppStore or the Google Play Store.
Phones doesn't matter, they are mostly the same, the operating systems are interesting, but without a widely supported app store, it does not matter, the phone will be useless for the majority of people. More and more people are using their phone as their only computing device, they need to have access to banking apps, mobile payment solution and government application, none of those are not going to exist on some random Linux phone.
Then the question should be "which bank is backing this ROM", not "is app A for bank B work on this ROM".
The answer to the latter is and will continue to be an instant negative for quite a while, and it's pointless question to make other than as a really indirect nudging for big bank(lol).
Who said anything about sticking anything into credit card terminals? And I'm not seeing any wires sticking out of the phone. I'd say it doesn't particularly look like a bomb either, but given recent events, that probably doesn't say much.
You'll be asking the same thing by trying to pay through a totally custom phone. As I'm repeating here, the blocker is lack of trust from banks, not technical reasons. THEY have to trust YOU, and then banking apps will work. Otherwise any tricks to get past security check to manipulate your bank accounts will be just short-lived hacks.
If the King of England walked into the Bank of England and wrote a hand-written bond for GBP1k with a bic pen on back of copier paper, they'll take it. Electronic platforms has nothing to do with that. That's why it's stupid to keep asking if banking apps would work.
>If the King of England walked into the Bank of England and wrote a hand-written bond for GBP1k with a bic pen on back of copier paper, they'll take it.
what has gone wrong in your brain that you think that?
If in any doubt about whether to buy this, look at Planet Computers offerings and the quality of their Linux build. They have been offering Mediatek-based Linux phones for several years.
Spoiler: garbage with ancient proprietary kernel, thanks to Mediatek
Furi OS was what turned me away. Never heard of it and that makes me realise we don't have an android alternative OS which is permanent. Do we have any OS which is permanent? Shouldn't we have one?
FuriOS is a derivate of Droidian, which in turn is basically Debian with Halium support. So it's pretty much Debian + some custom patches for usability, hardware support, performance, so on. We intend to upstream as many of our changes as possible as time permits, but right now we are laser focused on improving the experience day in and day out.
Once you turn on full disk encryption, the phone will reboot once and begin to encrypt the entire filesystem in the background. Further reboots will require that you input your FDE password (which can be different from your lock screen password/PIN) in a minimal rootfs-based UI.
I don't like to assume anything when spending 500$.
I have a One Plus 12, which has HDMI out, but it's like they didn't feel like finishing it. It's not really usable, it doesn't have any controls like Dex does.
Why not use the open source Android? I just feel like it would be more mature.
And yes I like many others here have been trying Linux phones for 10+ years, since N900.
Your phone just has to work, it can't crash because you get a call, or hide buttons because the UI is confused by orientation. IT MUST WORK. So that's why I wonder why people don't just use Android.
Open source Android has been increasingly neglected by Google, and is (essentially) unusable at this point unless you are ok with the look and feel of very early Android. It also isn't likely to improve as Google's strategy seems to be to move toward more and more proprietaryness.
I think a much more sustainable approach is to focus on existing Linux distros and get them more usable on phones.
You are wrong about AOSP. I'm running LineageOS (Android 14) on both my phone and tablet without Gapps and it is a fully functional system. (NewPipe for youtube, Aurora/Fdroid for for App Stores)
I don't know what you mean by "look and feel of early android" either, as AOSP based ROMs look just like Google Pixel in the settings, quick dropdowns and dialogue popups.
LineageOS is not the raw open source Android though that the parent probably referred to. Afaik there are quite some changes needed by now to make it really usable.
Still though it kinda proves my point. What's the quickest path to a usable phone? Quite a bit of work on top of AOSP, or even more work ontop of Linux?
It's like using Chromium or whatever clone-a-chrome is hip at the moment. That does absolutely nothing to disrupt the status quo and is very much at the whim of the upstream "owner" whether you want to see it or not.
Disrupting the status-quo is important, but having a working communication device is more important to me. If AOSP is further along towards a working phone then I don't see why we shouldn't focus development efforts on it.
It only benefits the open source market in the long run because a working device will garner wider adoption.
This website is just infuriating[0] to use with the randomly disappearing elements (FF 131 on Fedora 39, even with disabled uBlock). $500 sounds kind of funny when PinePhonePro is $400, but of course on the other hand Librem5 is $1000+, so whatever. Still, (2-gen-old) iPhone price-range.
Couldn't find too many datasheets nor other documentation unfortunately, though it might because of the unusable piece of crap website, which made me rage-close it.
[0]: just noticed: inFURIating, as in Furi labs, haha
I honestly think it looks fine. Nice texture at the back, pretty metal accents around the camera and at the side button, good logo. The uneven bezels (the one at the bottom is larger) are less nice, but that's just a reality of the budget this probably moves in, not even the Fairphone 5 did solve that yet.
... which is $2k... We already know that almost everyone likes the idea of not being Chinese stuff, but no one is willing to pay 4x premium. The Liberty specs are dismal for a 2k phone; as I said (I think response to you as well); I would pay basically anything for descent modern specs (so s24 or so) but running linux, but not 2k for something that will just annoy me in not being practical besides for a bit of lisp (nothing wrong with that though!!!).
What I want is quality to the level of flagship android phones or the iphone with the ability to plug it in to a keyboard, mouse and screen and have it function as my laptop as well.
One single integrated experience. I think everyone wants this. We have the technology we have the demand, it's just companies who make the high quality phones want to completely lock down the market.
You would think the free market would produce a company that would tackle this issue but barrier to entry is waay to high so essentially we have small companies trying to get into this space but they are only creating sub-par products because they can't make something as good as an iphone.
It's basically a legal monopoly. Companies can't get into it because the technology and capital required are way to high and only super mega corps like samsung and apple can pull it off...
Closest thing I've seen to this product I'm looking for is the steam deck. I would buy a steam deck phone if it had the quality of a pixel/iphone device.
You asked for it. I don't want that. I don't want an "integrated experience" and "flagship quality". The first sounds juvenile and the second sounds unnecessarily expensive and probably containing shit I don't need, like fancy cameras to look good on my nonexistent socials.
What I want is a simple, slow, old, efficient, simple phone with the interface of an 80s era 8bit computer that can actually, imagine this, make and, to complicate matters even further, even take calls.
I basically want an open source dumb phone. Do these exist? If not, why not focus on this first? Why go for fancy cameras and apps when we can't even make calls? Looking at your PinePhone.
LTE set that back by a lot. For years there were no VoLTE implementations in common use other than stuffs on Android. Even those had compatibility issues and lots of carriers still block unapproved clients trying to register on VoLTE.
For 3G, you could always do that. You only needed the right modem module with voice call support and audio I/O, like bare PCM pins, and a host micro to handle AT commands.
> there were no VoLTE implementations in common use other than stuffs on Android
What's wrong with the Android one - is it not permissively licensed?
I think the biggest problem of Linux phones is the community's obsession with trying (and failing) to reimplement (multiple times, in parallel) things that Android does really well and can be used as-is.
That's why the PinePhone or Librem 5 still can't even match the usability (at basic things like phone or camera or battery life) of a 2010-era Android phone, despite having similar hardware.
You want a Linux phone that actually works? Start with an AOSP-based phone and provide manufacturer-approved root and escape hatch such as first-party terminal and Wayland/X server app to run Linux apps.
Over time, you can slowly replace Android components with their Linux desktop counterparts when they're ready (or the other way around - the Android bits can just be the commonly-accepted solution to specific problems in Linux - even desktop - distros), but at least you're starting from a solid base.
> What's wrong with the Android one - is it not permissively licensed?
It is not, AOSP based distributions have to kang it from vendor builds. Qualcomm's is mostly standardized but Samsung wrote their own stack and voLTE/voNR won't work on any custom roms.
> That's why the PinePhone or Librem 5 still can't even match the usability (at basic things like phone or camera or battery life) of a 2010-era Android phone, despite having similar hardware.
They most certainly do NOT have similar hardware. You're wrong on thinking it's a software problem when the hardware being interfaced with is notoriously proprietary. The PinePhone and Librem phones are using self-contained quectel modems connected via different interfaces. They are nothing like the integrated soc's of nearly every other device on the market. This dramatically impacts battery life and stability and I don't think it will ever be a solved problem when building devices this way.
disclaimer: just an old grumpy customer, not an industry expert.
There were pointless debates around ITU-T and 3GPP as to whether the LTE is 3.5G or 3.9G or 4G next to the pink fact elephant that _it_ is going to be _the_ next cell standard anyway. That hot debate delayed voice call discussions to post-launch matter that eventually coagulated as the VoLTE.
The Voice-over-LTE was a total kludge together that were(ARE) carrier specific implementations. In Japan at least, it seemed to have had push from KDDI, esp. with hindsight that NTT docomo and Fujitsu, both formerly influential in 3G, both seemed on fire majority of that timeframe, while Nokia being a supernova beyond fire. It would be very natural if Samsung would have made vital contributions, but I don't have much informational pressure from that direction. VoLTE is phone company mannerisms, remnants of retro-futuristic media features of 3G, weird spaghetti codes from carrier labs, and international call exchange system, all homogenized in a blender; the artefacts were co-developed pair of server and client implementations that are standardized in the way there aren't many implementations but not something carefully spec-worked before construction for interoperability. That all happened ~a decade ago. Some point between Android 4 to 10.
I suppose it was all retconned into 3GPP standards after action, but the whole stack is still like a lobotomized Android call app and private fork of Asterisk embedded into the cellular core monstrosity. I guess embedded modem people(like Quectel) had finally got to port functions into modem chip firmwares so they can make calls, after someone done it for non-smartphones(KaiOS, Smarterphone...), I think around 2019 +/- 1 year.
What I'm trying to say is, VoLTE is complicated. It's something like SIM-authenticated SIP/RTP over IP under IP, not even regular 3G data session let alone SIP-VoIP on Layer 3 UDP/IP.
Osmocom project Wiki summarizes it better than I[1]: "Voice over LTE is an adaptation by 3GPP to use IMS over an LTE cellular network. The LTE EPS (Evolved Packet System) provides the functionality of the underlying IP-CAN. ... IMS is much more than normal SIP/RTP. And in addition to that, there is a tight integration between the LTE system and the IMS on top of it."
And it's supposed to be the core feature of a phone, "not".toupper() a party trick... There are better things to do in life for most people.
Holy moly, you must have very different life than most people I guess. I am from the younger generation which supposedly avoids making old-style phone calls as much as possible and I don't think I know anyone who has made less than 100 calls in the last decade.
I think those very total attention expected during phone calls, and the speed at which it happens, contributes to a higher level of anxiety compared to texting. The younger generation hate these elevated stresses and prefer more async communication
Same for me, I am 50 and would happily carry a phone sized tablet without call caps at all. I never need to call or receive call; send me an email or hit me up on a chat channel.
>I basically want an open source dumb phone. Do these exist? If not, why not focus on this first? Why go for fancy cameras and apps when we can't even make calls? Looking at your PinePhone.
Your dream is achievable and has been achieved. PinePhone for example.
I obviously don't mean everyone everyone. More like the overwhelming majority of people.
I have one and it very much struggles to make and especially to receive calls and is still trying to be smart. PinePhone is definitely not a dumb phone. It's expensive as well.
To me a simple, dumb, open source device that can easily be manufactured in all kinds of conditions all over the world sounds like a dream for actual, practical purposes. Like, for example, again, calling. To some degree I have the same issues with "smart watches". Simple, open source, dumb smart watches with just a smidge of 8bit CPU goodness to display, say, something simple like a word or even a letter on the screen would be quite useful. I know there is some movement in that arena using ESP32s but I am not particularly impressed. Alas, alas. Why do we as a civilization tend to go for the extremes and not just get our basic shit together first?
I was being slightly obtuse and I understood perfectly well that you meant all reasonable people. If I can't be obtuse and pedantic on HN though, where else?
These do (or did) exist: feature phones shipped with FirefoxOS. Almost no one bought them and the effort was widely seen as a failure, although KaiOS enjoys ongoing success in the developing world.
I am using a KaiOS phone, a Nokia 6300 4G. It works well as a 4G WiFi hotspot for a laptop and as a small phone for making calls and sending texts. Have a few apps on it for mapping as well as using the browser occasionally.
The KaiOS WhatsApp app will stop working next year though.
I beg to differ. Tried to use a kaios nokia 800 tough. Firefox/KaiOS using a stupid kind of webview for every UI, you get two consequences: first, everything is ugly: wrong padding everywhere, complex layout with non-integer number of list items fitting on screen (this is important, vertical list is the primary widget on small screen dumbphones). And second: everything is sluggish. My true "series 30" nokia launches app instantly, whereas i'm sure the SoC is order of magnitudes slower.
So the UX definitely shows complete absence of culture of design in this kind of space (eg gameboy UI design, classic dumbphone UI, ...). Small screens need very simple widgets (most importantly grid and vertical list), hand made paddings and proper space optimizations, most likely some bitmap font tailored to the resolution.. You can't just slap some dynamic layout and expect it to work nicely on such a small screen. Everything vectorized will be both costly and of dubious rendering quality, etc..
And of course these things only have "decent", but definitely not "1 month scale" battery life. I guess LTE is the real culprit here.
I half agree. I dont need the bullshit bells and whistles. However a good camera doesn't mean you want to post on instagram or tiktok - maybe you just want a memory of a trip you went on or technical detail of something you want to document. People took pictures before social media.
Most every phone can already do what we want but the hardware is undocumented or locked behind NDA and the firmware is hostile. Without those locks the gigacorps cant keep you in the data-mining garden. We wont have open phones until this thinking changes. Keep dreaming until then.
I agree I would love this option, but how much would you be willing to pay for such a device? Would you pay $800? $1,000? $1,400?
That's where I have a hard time. I would pay that kind of money, but I would need something well polished and fully capable of being a "daily driver." I think many people are in the same place I am, and thus we have a real chicken and egg problem here.
Didn't they? The Lumia 950 launched at about $500ish and supported Continuum.
Continuum wasn't exactly full convergence, but it was kind of close. As I recall, they had a desktop dock and a laptop like dock, that you could pair your phone to (wireless or wired).
Microsoft was big on their 'universal windows app' concept at the time, where universal meant actually only ran on a small fraction of windows devices. sigh
aw. I completely forgot about that. Why didn't it just work? Was it... during those "we invest all our cash to everything needed for total Win32 dominance but boldly let it rot in the backlot" days of Microsoft... yes yes it was.
Hmm. If I bought a phone for ~$600 and a laptop for ~$1000, I guess I’d pay ~$1700 for the converged device (adding a little bit because of the intangible benefit of not having to manage my files anymore).
I'd pay up to 3k if it was quality delivered the experience of a top quality phone and a top quality laptop.
I mean the technology to do this is already here. If apple or samsung wanted to do this... they can.
Right now I use an iphone, but if samsung made a phone that felt like say linux, windows or macos when in laptop mode... I would switch off iphone in a heartbeat.
And i mean it has to feel like macOS. None of that bloat is acceptable. SteamOS actually pulls this off but in a gaming form factor.
> And i mean it has to feel like macOS. None of that bloat is acceptable.
I suspect that every single thing you like about macOS is something I consider bloat (or would if that was the only related word I was allowed to use).
My first task on my recently acquired M3 MBP was to remove/hide/disable as much of macOS as I possibly could; only then can I use it as a productive development environment.
> and a top quality laptop.
> I mean the technology to do this is already here.
I think it really depends on what precisely you mean here. I don't think it's currently possible to get the same performance of my M3 Max MacBook Pro in the form factor of a phone.
Asus did that in 2011 with the PadFone, it was an Android mobile phone that you could plug into the provided tabled and then plug a keyboard into that.
> What I want is quality to the level of flagship android phones or the iphone with the ability to plug it in to a keyboard, mouse and screen and have it function as my laptop as well.
>
> One single integrated experience.
Purism achieved that with their Librem 5 phone [0]. However having a polished experience like with iPhone requires to invest billions in software. It can be used as a daily driver [1] but there's a lot to improve yet. Sent from
my Librem 5.
I can do something not unlike that with the second-hand displaylink adapter I currently have gathering dust... I would just need to install a desktop environment on termux for it to be complete (not sure if I could get 3D acceleration with this). The only real problem there is the adapter I have cannot charge the phone while connected.
> One single integrated experience. I think everyone wants this
This is insane to me. Like not just no. Hell no. I want my phone to be a gadget, like a watch, not a whole ass computer.
I want my tablet to be its own thing from my laptop, which is different from my PC(s). I absolutely, positively do NOT want general purpose computing to be the typical paradigm. I don’t want, need, or even like the idea of my phone, that I typically use like a wallet, to be a general purpose computation tool. Lock that shit down, please!
I think adults should read 'majority' when someone says everyone. I will keep saying it when talking with adults as it is just easier than put all the exceptions; if some people get annoyed by it, that's not really my problem.
Alternatively you could use 'most people', or any one of several equivalent expressions, and preserve the meaning of everyone as every one. There is an aroma of manipulation around using language that gives the impression of absolute consensus even if all parties involved understand its vernacular meaning. It's embellishment in service of influencing people and rightly denounced in any context where accurate communication is important.
But we are not writing a scientific paper here; we are having a discussion. I guess if it matters in the scientific sense, one would be more careful; so if the discussion would be about a new type of medicine then 'cures all people' would be a stupid thing to say. Here however, it doesn't matter. But point taken: I guess I commented as HN has especially has this nasty about of commenting on
things where it indeed doesn't matter, stuff like 'citation needed' or 'but it cannot be most people, because I know blah blah blah'. Pointless let's say for this phone discussion; might be relevant for a medicine/vaccine or so.
> I want my tablet to be its own thing from my laptop, which is different from my PC(s).
But what's the point in owning three devices when you can own one doing everything? How about in addition an mp3 player, a camera and a flashlight, too?
I like tools to have defined roles, not multipurpose.
I mean the top post is a ~200USD “smart” pomodoro timer. I’m not even close to that.
I use my phone for communication, quick photos/ notes and sound. It carries my cards. I want it to be completely locked down because it gets used in public and could be stolen.
I use my laptop for personal business including coding. I want it to be secure and streamlined.
I have a gaming PC, it runs widows and gaming rootkits and all kinds of questionable software. I don’t want my sensitive data around that computer.
I do art on my ipad, I want it to be about art. I dont need the phone stuff there. I don’t want my personal stuff there.
> flashlight too
lol I actually carry a flashlight with me always. Phones are absolutely garbage compared to a modern High CRI flashlight. I also have a FF camera and take it with me very frequently.
Apparently it is based on Halium, so it's probably not going to run mainline Linux any time soon. On the other hand, I'm less happy with Android than I pretty much ever have been, so having some more options outside of relatively unusable Linux phones and being stuck with modified Android firmware certainly does not hurt.
I am definitely interested, especially if the team working on it is as active as it seems they are. Definitely one of the biggest bummers with PinePhone and PinePhone Pro was realizing that the community was largely on its own; that could work well, but in practice progress has been slow and painful and I stopped paying attention.
I would really like to see more projects doing cellular Linux devices in general. It doesn't need to be a candybar smartphone, if someone can jam a cellular modem into a MicroPC I'm sold at any price I can afford.
I work on this device. We are currently QAing the latest update with our community, which contains a lot of goodies and improvements shaped directly from community feedback. You can see previous changelogs at https://furilabs.com/changelog
I am obviously a little biased, but this is probably the first Linux phone that people can actually use. Someone in our community switched from iPhone to this without much of an issue.
Nokia N900 and N9 ran on Linux in 2009 and 2011.
N900: https://en.wikipedia.org/wiki/Nokia_N900
N9: https://en.wikipedia.org/wiki/Nokia_N9
The Motorola A1600 ran Linux in 2008. But even then people didn't want a flip phone.
Interesting history! "Motorola became the first company to use Linux on a mobile phone when it released the Motorola A760 to the Chinese market on February 16, 2003." https://en.wikipedia.org/wiki/MontaVista#Mobile_phones
The Sailfish OS phones are in some use, and they've been running full blown linux since 2013.
Using a Sailfish OS phone since 2013 - just saying. ;-)
Right, used 2 of them as my primary phone between 2015 and 2024 when the hardware broke not many weeks ago. Just recently ordered their new one, which is a bit cheaper than this Furi. Admittedly a risk purchase because not much is known about it yet.
Also typing this on the ultra-cheap hardware made for the Indian market nearly a decade ago. Well, admittedly they have already announced to discontinue support and I never dared to carry it as my primary phone.
Now the last weeks I was forced to use Android as my primary phone. Really a suprise what crappy piecemeal Android is compared to SailfishOS made by a tiny company struggling economically forever.
> this is probably the first Linux phone that people can actually use
I'm actually using GNU/Linux phones since 2008. Never relied on Halium.
My gripe with any of the open source phones is that id apps such as bankid don’t work without the secure token storage of Android. Is this Solvable?
You could always protect the signing certificates in the apps with derived passwords, still the length of passkeys practically acceptable to type in on a phone is too short to safely protect a certificate vs a bruteforce attack without some kind of HW assisted storage.
In the end it also boils down to what devices the BankID app providers are willing to support, I have a hard time seeing anything but iOS or Android devices being supported in the near future, Esp as Swedish BankID's now also requires NFC support to read the local police issued ID cards (had to get a new testing-device just due to this requirement).
Note: BankID is the name of personal identity apps that support authentication and signatures in Sweden, Norway and Finland, the authentication is used to access a myriad of both public and private sites like tax office, unemployment, healthcare and gyms. The signatures done via the apps are generally accepted to be of as good legal standing as a signed paper.
I've never heard of BankID in Finland - perhaps the common name is something else?
I would also have commented: Not supported in Finland. I think I have read articles that they tried to get marketing share, but AFAIK they have failed completely.
On the positive side in Finland you can use SIM Toolkit for legally sanctioned 2FA (mobiilivarmenne). That should be much easier to implement without having Google involved.
Since you guys are running Gnome, you might be interested in helping with (or funding) this native GTK WhatsApp client: https://github.com/tobagin/whakarere/
Since these days something like >90% of people is on WhatsApp, it’s beyond quintessential to have a good WhatsApp client.
My biggest hold-up with that sort of work is that Facebook has previously threatened to ban users who use third-party clients, and then they have gone through with the threat. Is this no longer the case?
Meta also threatened app makers to start a lawsuit. Whatsapp Plus, Whatsup and Mitakuuluu people stopped on that threat. I don't see a reason why Meta would change their stance on this.
There is now a law from the EU, I think it is the DMA, which could change things for people living in the EU regarding interoperability. I haven't read anything about practical follow-ups yet.
By the way, only releasing source code as a blueprint might be fine, that happened before. I might remember it wrong, but it might be lame, the mp3 encoder.
I can't speak for everyone, but I've been using WhatsApp since ~2010, mostly the official client but with bouts of unofficial client usage, and I've never been banned.
Probably not very detectable for Meta since it uses a wrapped webview, but I've been using ZapZap on Linux for a good while now and I haven't seen a ban yet.
if something can happen and haven't happened to you, doesn't warrant a comment i believe. it doesn't help much to figure the data points for the cases were it does happen, since it's implied everyone to whom it have happened would have lots of opportunities to post comments it never happened before.
Has a well-behaved third party client ever been an issue?. A third party client that does things like retain deleted messages, report that read receipts are enabled but never actually sends them or other 'anti-social' behaviors is definitely asking to get banned.
[1]: In July 2021 Facebook sent the developer of Unfollow Everything a cease-and-desist letter because he made a browser extension which helped people use Facebook less.
[2]: In 2020, Facebook shutted down the NYU Ad Observatory by shutting down the Facebook accounts of Cybersecurity for Democracy team members Damon McCoy and Laura Edelson.
[1]: https://12challenges.substack.com/p/how-to-deal-with-receivi... [2]: https://cyber.nyu.edu/2021/08/21/facebook-disables-ad-observ...
Can i run my own gui like sxmo on it easily?
Easily probably not the word I'd use. Anything that runs on wlroots should be doable with not a lot of effort. Beyond that, tricky. sxmo doesn't use hardware acceleration though so not too bad.
Anyone tried running KDE?
Not on this device in particular, but I've seen it done in the Droidian community. Similar steps should work, although I heard it's a little janky.
I don't care about running stuff out of the box, but NixOs support would be awesome!
Does anyone have NixOS+Halium working?
Possibly the OnePlus 6? This isn't an area I'm too well versed on though
Is there any phone that runs mainline Linux/upstream Debian, and is available to buy right now?
I'm sick of big companies who removed the word Consent from their dictionaries. I want something I can own and trust again. Being able to run android apps would be great, but given how closed source Android has become, I doubt that apps run on an open system would be reliable.
100% mainline - not really yet, but from those that are close to mainline and on which you can run Mobian (which is upstream Debian with a small repo overlay for stuff not ready to go into Debian proper yet) and that you can buy new right now are Librem 5, PinePhone and PinePhone Pro.
If second market also counts then there are a few more options available - mostly former Android phones where hardware support is spotty though, so do your research first. That said, everyone's needs are different and what's there can already fulfill the needs of plenty of people, so it's a matter of figuring out whether you in particular are in this group yet or not.
Halium is based on libhybris, to my understanding that was developed when Nokia stopped making Linux phones and some of their developers started making their own under the name Jolla/SaifishOS in 2011.
While Nokia could maintain their own Linux kernel that's impossible for a small vendor/team. There is no realistic alternative to taking the Android kernel provided by the SoC vendor.
I want to get one of these to play around with soon: https://wiki.postmarketos.org/wiki/OnePlus_6T_(oneplus-fajit...
The SDM845 stuff looks quite promising for mainline but unfortunately the platform is definitely showing its age (and camera is still completely broken on every device as far as I know...)
Still, it's something!
Not true for the Poco F1 (although AFAIK you still need to build your own kernel on postmarketOS to enjoy this): https://fosstodon.org/@joelselvaraj/112621744555315631
Shift6mq is supposedly close, with OnePlus6(T) its more difficult due to I think the way the sensor is connected.
Currently, the Pixel 3a is the most promising device in terms of Camera Support on a Qualcomm device with mainline kernel from what I can gather.
I want to like it... But, I've been bitten too many times by expensive Linux phones (Oh Nokia N900, how I wanted to love you. but you couldn't even make phone calls without threatening to crash).
It's a little weird that there are very few details on this thing. Notably, one of my #1 concerns when looking for a new phone today is a good camera. And all I can find on the camera is "it's 50MP!!! " But how do actual, non-photoshopped pictures look? Does it do okay in low-light? Does it support any sort of optical zoom?
Their website is currently as slow as a frozen snail (which, if they can't run a basic web server, leads to a lot of doubt they can maintain a long-term Linux distro). So I couldn't click around much, but I didn't find anything really informative.
I tried looking at YouTube videos and only found some weird official videos of a person's hand putting the phone in a sink. Like... Yes, it has been expected for the last 8ish years that my phone will be waterproof, that's not an amazing achievement.
It also seems to be "weird" Linux, like Maemo was for the N900. I don't want weird, custom-built stuff that will be forgotten about in 12 months. Just give me a standard Debain/Fedora/Whatever and an unlocked bootloader to reinstall with what I want.
IMHO they should really look at what Valve did with the steamdeck and SteamOS, It's a custom Linux that works well, but, it's also standard hardware with an open bootloader. There are dozens of SteamOS alternatives.
I agree with you on some points - I do know a lot of people (myself included) have been bitten in the past many years by Linux phones that didn't do what they really claimed they'd eventually do. There's not much I can do there other than continue to work on it and make it the best possible Linux phone. It will speak for itself.
I also completely agree that we should have a camera showcase. We've been talking about the idea of letting customers upload their shots to some sort of gallery so people can see how the real world looks through the camera's lens. What I can tell you though is that pictures look pretty darn good, and low light is great owing to the f/1.57 aperture. No optical zoom for this model unfortunately.
I'm not sure the website being slow speaks to our capability of maintaining a distro, especially when you consider these two things are done by different people. Sure, I'd love to work on the website and make it faster. But I'm busy working on the actual phone.
The video situation will improve as reviews come in. As for just giving you an unlocked bootloader and regular Linux... sure, I can do that. But you're not going to have audio, hardware acceleration, camera, etc. postmarketOS is a good example of this - despite years of tireless work, it's just really hard to get a regular distro working on a phone without leveraging Android drivers, at least for now.
>postmarketOS is a good example of this - despite years of tireless work, it's just really hard to get a regular distro working on a phone without leveraging Android drivers, at least for now.
Does this mean you are not working on mainlining this device? I'll take the slow and steady pmOS approach over sketchy Android shims. That's what ruined the Planet Computers Gemini PDA for me. A temporary solution is often surprisingly permanent.
We are not actively working on it, although some work has been done in that direction as a weekend project. It's not something we are ruling out, but we want to focus on delivering something that works, and works well. If the product does well enough to be able to finance continued effort in that direction, and there aren't any "massive" issues to work on before that, then... I guess it might just happen. Hard to tell right now.
> It also seems to be "weird" Linux, like Maemo was for the N900. I don't want weird, custom-built stuff that will be forgotten about in 12 months. Just give me a standard Debain/Fedora/Whatever and an unlocked bootloader to reinstall with what I want.
I wouldn't say it's a "weird Linux as it's mostly Debian and it's fully rebuildable by the user from the open source scripts[1]. I would likewise love a phone that was just mainline Fedora, but given many of the current challenges around ARM devices with linux, I fully understand why a "custom-built" OS is needed. As long as it's as light-weight as possible and is open source/rebuildable by the user, I think it's a good compromise given the realities around the platform.
[1]: https://news.ycombinator.com/item?id=41840036
To be fair (insert meme here)
Maemo was "mostly Debian" as well, it was one of the big selling points. Except Maemo was quickly abandoned by Nokia and left unmaintained. To further complicate matters Maemo required arcane tools to build it and install it to a device. Custom community kernels quickly became standard for enabling basic functionality which Nokia should have provided, but, didn't. However, the custom kernels were also frequently left unpatched and abandoned because they were being built and maintained by random internet users who found other hobbies.
I think the idea is awesome and I am excited if not skeptical. I hope Furi manages to build a healthy community around this. And doesn't go the path of so many previous Linux phones.
n900 was perfectly reliable, yours had a problem. And Maemo was a wonderful interface for mobile stuff. Clean, functional, tweakable. Regret it every day, facing shitty Android. I dream for a LTE n900 with a larger screen, leaving the rest unchanged : unbreakable resistive screen, mecanical keyboard, fantastic stereo speakers, etc.
Maemo was indeed fantastic!
SFOS is somewhat getting where Harmattan was before Nokia imploded. It's already better in some ways.
> And all I can find on the camera is "it's 50MP!!! " But how do actual, non-photoshopped pictures look?
It's probably all they know about the camera. Not even Apple make their own cameras. It's always someone else's (customized)product. If they're outsourcing hardware, they might not even know the customer code assigned to the contracted OEM.
> I don't want weird, custom-built stuff that will be forgotten about in 12 months
Then none of hardware and many of software features will work. There are no in-box standard implementations for e.g. phone call in GNU/Linux, so it'll have to be supplied in the form of a custom preconfigured distro with bunch of free preinstalled apps.
The fact that product necessitating free preinstalled apps to function leads to needs to supply OS as flavors of your own weird unsupported distro, like N900-specifix Maemo or Steam Deck OLED-specific SteamOS images.
You aren't going to like being instructed how to install and patch a SIP softphone just to make calls on North American carriers with vague hints for Canadian specifics. It's a phone, it'll be lame if it doesn't make calls out of the box. And there comes the 3 years old $DISTRO-$HWREV-$VERSION.img that instantly shows antenna bars right on the setup screen.
Actually I think it's unfair to call Maemo a weird unsupported thing. It didn't seem that far from bone stock desktop Linux.
The website (which is faster now) links to a review with a nice photo showcase, https://blog-d.luigi311.com/furilabs-flx1/. Not perfect, but better than I thought!
FWIW I wrote about this a few months ago, after trying the actual device at devconf.cz in July.
It's on HN here:
https://news.ycombinator.com/item?id=40805156
Are you referring to the one line comment you wrote, or do you write for The Register and that was your story submitted?
They seem to have written the original article going off the articles author name and the name in their bio.
That is correct. I have also stated in my bio that I work for the Reg.
I wrote the Register article.
It's surprising to me that there aren't any details about the software stack.
Whenever I've tried to find a touch-optimized Linux shell, I've come up short. As has been discussed on this forum in the past, most of the attempts are understaffed side projects that have thousands of manhours of work ahead of them.
I'd like to know what shell FuriOS is based on, and what they're doing to make it usable. Even a demo video would go a long way!
I'll write something in the dev blog, thanks. We use Phosh, with some minor patches. Upstream Phosh is pretty good as it is!
Glad to see you engaging the community!
As someone unfamiliar with Furilabs, it wouldn't occur to me to go looking for a dev blog to see if you've written anything about the shell. I feel like that's information you'd want to link from the marketing page (even if the details are on your blog).
Community is what got me here in the first place. :)
Fair enough! Noted.
I haven't been able to try Phosh because it's badly packaged on NixOS.
Maybe one day I'll take a stab at fixing their package.
https://github.com/NixOS/nixpkgs/issues/278204
Recent phosh is packaged on Fedora. I’m critical but also pretty happy with it. Biggest disappointment is that many gnome apps are not as touchscreen friendly as was promised.
I might be interested once there's videos of it doing daily stuff, and especially of how its autofocus (speed and functionality) compares to, let's say even an old Samsung Galaxy S7.
What kind of scenarios would you like to see the autofocus working in?
is this actually their own HW design, or is it just some random alibaba HW with their software on it?
edit: looks like same HW as "Gigaset GX6" https://www.gigaset.com/hq_en/gigaset-gx6/
We work with the same ODM that supplied the hardware for the Gigaset GX6. We do not resell Gigaset phones directly, but we do use the same reference design.
review for the android version: https://www.notebookcheck.net/Gigaset-GX6-smartphone-review-...
seems pretty favorable, keep in mind that it's already two years old. it's also a brick at 280g, almost double a pixel 5, otherwise pretty similar
Given the open issues at
https://github.com/FuriLabs/issue-tracker/issues
I think it's fair to say that whoever buys this is in for a bumpy ride. I mean, that is to be expected, however, I think they should be more honest about the current state, like for instance PostmarketOS does at https://postmarketos.org/state/
> https://github.com/FuriLabs/issue-tracker/issues/78
If SMS doesn't work, I have serious concerns about everything else...
I'm currently using a Volla 22 with Ubuntu Touch and I'm happy with it. However if Furi is using Plasma Mobile (they're mum about what interface they're using, but if it's really a 'lightning fast user interface' then I don't know what else it could be) then it's unfortunate that I discovered this too late to get it instead.
I think I've seen multiple references to Phosh.
It is Phosh. It can run Plasma Mobile too, but I'm told it's still a bit early in development.
Can anyone estimate what it would cost to develop a new phone that's like the N900 with updated hardware (slide out keyboard, camera cover, maybe drop the input stylus, a bit flatter and larger) and mainline Linux support for all components? I'm sometimes wondering if this is something I could have Pine64 do if I accidentally get rich enough. I think I would spent the rest of my days just writing the software for this.
First question is; which SOC would you want to use? Is there any SOC with full mainline support that you would agree to? There is a reason Pine64 and Purism went with SOCs out of the ordinary path.
I believe everyone has similar fantasies. Your best bet is a custom made mini laptop with a x86_64 SoC with perhaps 18650 batteries or thin easily replaceable battery packs..
I believe the GPD Pocket is also x86_64.
Considering there's a bunch of stuff you'd have to manufacture from scratch I'm gonna say... 500 million USD? If you're okay with outsourcing a bunch of stuff to China and treading on gray areas where patents are concerned. Unfortunately it's gonna take a few more years before various mobile phone patents start expiring.
> Fast, performant and cheap. You wanted all 3? Now you got it!
Fast, Good, and Cheap. Performant is just a synonym for fast. Is it any good?
Also call me a miser, but $499 is not "cheap". I recently bought a new Nokia G42 with the same camera and storage for <€150
"privacy centric" then uses google services as part of web page. easy fix - quit referencing gstatic.com
That said, it is awesome they are making this phone, and even if you don't use it, more choices are better for everyone.
Good catch. We'll fix that. Thanks!
> Fast, performant and cheap. You wanted all 3? Now you got it!
What's the difference between fast and "performant"?
Maybe one is latency (how optimized the software is) and the other is throughput (how many cores it has).
Although more likely just weird marketing.
It's marketing drivel, and marketing loves the "rule of three".
$500 price point is a lot for an open source enthusiast phone, especially one unknown.
The Pine64 PinePhone is $150-$200, with similar hardware provenance, but various open source projects that might at least give better software longevity and software trustworthiness. https://pine64.com/product-category/pinephone/
An additional consideration: as much as we like and respect our global colleagues, unfortunately I could see both hardware brands getting banned in some locales and by some organizations, like some other telecom brands have been banned. A phone that can't be used as a phone has little value, no matter how open it is.
For somewhat different provenance, the Purism Librem 5 is coming in at $800+, which is too high a high price point for most open source enthusiast contributors. https://shop.puri.sm/shop/librem-5/
And Purism's made-in-USA version, the Liberty Phone, is at $2K+. I guess maybe government or enterprise sales? https://shop.puri.sm/shop/liberty-phone/
I'm thinking that better for the Western open source developers and non-wealthy enthusiasts would be something more palatable to Western governments, and usable as a daily driver, like Librem 5, but priced somewhere in (guessing) $200-$400.
> with similar hardware provenance
no not at all Pine64 hardware is pretty crap (performance wise, very normal independent of the OS at that price point)
Like you are comparing things like 3GB LPDDR3 SDRAM (Pine64) to 6GB LPDDR4X which is world apart (not only does it move from a "very tight on memory" to "somewhat acceptable memory" (very big deal) it also is much faster (~2x).
Similar on the Pine64 you have 4 old little cores here you have 2 big + 6 little cores and all more modern (end even just the little cores are faster when comparing one core with another, ignoring that you have 2 more little cores and 2 big cores).
Storage moves from a quite tight 32 GiB (through I guess if you only do stuff like mail it might be enough) to a likely more then enough 128GiB (also faster).
Noticeable faster GPU too, I guess.
And way better camera, at least by by numbers.
So not comparable at all.
But what matters most is not the absolute differences but moving it from a "so tight on RAM,CPU,GPU,Storage that you run all times into problems" to a "enough for all common tasks". I.e. from you need micro optimized apps for it to work nice to yes it works if your app is imperfect territory. And that is a pretty big deal.
But also lets be realistic. Atm. _any_ Linux Phone is a pure enthusiast product balancing various problems (price, small production lines, only android diver problems, proprietary blob problems etc.). And many of the enthusiast earn enough money to afford a 500$ toy, through maybe just rarely and with thinking twice about it.
Good points about hardware capability specs, but I meant to talk about a separate consideration.
I meant provenance as in geopolitical. Later in the comment alludes to this more, and one way it can be a showstopper.
Sorry I was being vague, but it's awkward to mention, and unfortunate that there is ever any potential conflict.
The original pinephone was an allwinner soc that had like 4x a53 cores, a weak GPU, and pretty meh ram. This phone has a fairly modern mediatek soc with a78 cores, a fairly modern Mali GPU, ufs storage, and lpddrx5 ram.
$500 and (hopefully) works is a lot better than $800+ and works or $200 and doesn’t work worth a shit.
It says "Wired/Wireless and NFC combo" under battery charging, but there's no mention of NFC under "Connectivity". Is it charging only? I'd like a linux phone but the Purism phone eschewed NFC too.
There's NFC, although the Linux stack support for NFC is a little limiting. However the NFC hardware is passed through to the Android container, so if, for example, your bank doesn't demand that you pass SafetyNet, you can pay with the phone. (I have done this)
I've used nfc on linux with several readers via pcsc-lite and it seemed to work fine. Do you mean support for the nfc chipsets used on phones? Does the android container deal with that differently somehow?
libnfc has different forks for different hardware and the main lib hasn't been updated in a while. It was a chore to get NFC over I2C working for me.
How would you check what the banking app you are using requires?
That's a good question, sorry I missed your comment.
I honestly don't know of an "easy" way to do that. If you have enough time on your hands you could decompile it and take a look. You could also try running it on a spare phone with a ROM that doesn't contain Google Play Services (or where it's been disabled). However, SafetyNet isn't the only line of defense. These apps often have their own root checks, which doesn't pose a problem for us because the Android image isn't rooted. (But it runs inside LXC, and you can attach a root shell into it.)
Another thing they don't like sometimes is the system build fingerprint not being recognized. We have an overlay(fs) system that allows users to customize everything inside the Android container without the hassle of rebuilding the system images or having to stay on the ball when we update things. It's not hard then to pick up the build fingerprint line from a "trusted" device and just drop it in build.prop.
All of this is a bit of a hassle, to put it mildly. But if we try to talk to any bank or contactless payment provider right now, they'll tell us to kick rocks. I guess it's a matter of time until we grow or a very smart person figures out a very clever hack. :)
> Fast, performant and cheap.
What's the difference between fast and performant?
Something can be fast, but lack efficiency or the ability to utilize the speed.
For example, a car can have 1,000hp, but utilizes a large turbo which takes a long time to spool up, leading to poor performance. It may also have poor handling characteristics.
On the other hand, let's use a car like a Mazda Miata, which only has 181hp. It is not fast, but it is an incredibly performant car, with exceptional handling and driving dynamics.
Speed and performance are not directly correlated, and are independent metrics. Performance is more of an indication of the overall dynamics of action, intention and results.
Hasty guess: fast is responsiveness, performant is performance per watt?
fast is latency, performant is bandwidth
I doubt they're using those words to mean that specifically when they are trying to sell a phone to consumers. No one sells phones on "bandwidth".
> Phone: 171mm x 82mm x 12mm : 280g
280 grams is a lot even for today's phones. 12 mm is half a inch. Very thick.
The other dimensions are 6.7 x 3.2 inches.
Ok, this might be a phone for people that want to prove a point but it's too cumbersome. How about starting with something reasonable and fit it with only what fits within those bounds? An example: 160 x 80 x 9 and 200 g. That would be a big phone but not bigger than many phones on sale.
> Ok, this might be a phone for people that want to prove a point but it's too cumbersome. How about starting with something reasonable and fit it with only what fits within those bounds? An example: 160 x 80 x 9 and 200 g. That would be a big phone but not bigger than many phones on sale.
Because in that universe your comment would be about how awful the specs were and how tiny the battery is. I'm reasonably confident that price-size-performance is a "pick at most 2" situation.
It takes lots of engineering effort to make things svelte. That makes it incredibly hard for "indie" phone makers to come up with something competitive in this regard. Likewise with price --- Android phones with similar specs can be found for only $200.
Why bother when you could pay HMD (aka "Nokia"), Foxconn, ASUS, etc to make a dramatically better phone with your logo and your OS on it?
Can you actually do this with the numbers Furilabs probably expects to hit? Surely there's a minimum number to order before the companies you mentioned will even take you serious.
The Linux phones all fail anyway so if you don't have the volume just don't bother.
I agree that it's definitely too tall but I don't agree that it's too thick. It has a removable battery which fully justifies the thickness imo.
I don't think 12mm is bad. My iPhone in its Otterbox sleeve is just over 11mm thick.
Is there prior art that includes a removable battery, headphone jack and IP68 at those "reasonable" bounds? I wonder how large the battery could be in that case.
I rather have nice specs and battery life than thin.
I'd gladly take a phone that's twice as thick if it shaves a quarter off the width and a third off the height.
Do you seriously think weight and thickness are parameters that hardware designers just artificially inflate all the time just to be mean to users???
You want 1/3 of thickness and 53ml in volume and 80g in weight just removed because methinks that'd be much better??? A phone isn't a steak cut, you... can't do that. You can if you knew beforehand that there are suitable replacements for major components like displays and batteries, if you don't, you just take what hardware engineers give you. What are you even thinking.
https://archive.is/WSQ9G
It's sad that basically a static page is getting hugged to death.
Does the usb c support displayport alt mode?
How's the performance compared to a Linux vm running on a pixel 9 with kvm?
Android 15 recently released, which brought support for graphical acceleration inside the vm, so that's one competitor for the Linux phone.
I'm interested, is there a hands on video or review somewhere?
I can't get their timeline page to load correctly: https://furilabs.com/timeline/
I'm so desperately looking for a DIY portable device that does not suck to build my portable audio player, best case supporting android with any of the following custom roms:
I even considered making a player myself using the T113-S3 SoC, unfortunately not having the required skills. There are so many people trying to build prototypes, but this seems to be a space that is not easy to manage.Here is my personal favorite Audio Player base device, best I found so far:
https://github.com/yuansco/TinyEmbedded-Dual
Not Android, but it's an open-source music player:
https://www.crowdsupply.com/cool-tech-zone/tangara
Thank you. Yeah I did know that. For my use case it unfortunately is too clunky and it does not have headphone remote (afaik), but I really considered buying one, just in case it gets added.
In Android, headphone remote works out of the box... that's why I would like to use one. There is the HiBy M300, which in my opinion would be worth having a custom rom :-) Indeed there seems to be efforts.
https://github.com/reindex-ot/hiby_digital_m300
https://www.head-fi.org/threads/hiby-digital-m300-android-di...
https://store.hiby.com/products/hiby-digital-m300
Check out Rockbox if you haven't already.
Took me ages to connect the fact the phone was flipping around at hyper-speed on the webpage because I had the cursor over it whilst scrolling down. Had to move to the right side of the page to then scroll down.
the linkedin profile linked to in footer doesn't exist. #redflag
Weird. A possibly matching profile shows up in search for me, but upon clicking the link LinkedIn says they can't find a matching page.
Very tempting. What is the deal with play store integrity, i.e. is wallet (and some banking apps) supported?
It says "F-Droid, Aurora Store, and other sources'" so I'm guessing there's no Play store and probably no GMS and other increasingly mandatory proprietary dependencies. Open source Android is in rough shape as a result of Google's tightening the ratchet.
Edit: In a different comment one of the devs clarified that it does have microG, so quite a few apps will actually work.[1]
[1]: https://news.ycombinator.com/item?id=41840021
Afraid not. These things usually require hardware attestation to ensure you're running the exact software Google wants you to run. Now, if someone figured out a way to bypass it, that would probably work for a bit. But it's a cat and mouse game.
It's not just what Google wants you to run, it's what Google promises to some of its customers through a face it has on one of numerous facets(facing banks in this case) as what they let users do with phones, and a humanware chain of trust that allows the server and the phone to be viewed as sort-of secure computing environment.
From an idealist standpoint, it shouldn't be done that way by taking away rights from users, but trivializing it as how specifically Google wants you to do it isn't accurate either.
Use the web browser and a debit/credit card.
Can we get a Linux smartphone that works before worrying about nonessential use cases? One the powers-that-be will never support until step zero is shipped at volume.
Why does this question always come up? There's zero technical reasons you can't electronically pay with all-open software stack e.g. Ethereum on GNU/Linux, it's pure matter of trust.
With currently available payment standards, it's like asking for a live credit card with someone else's private key and an open debug port, all at the same time. Things don't work that way.
> Why does this question always come up?
Because that's the only question that matters. I have just a few apps installed, some of them are easily replaced, like TOTP apps, others are impossible to replace and without them I might as well get a feature phone.
I'm down to one app that I really do need, MobilePay, but it's a payment app, heavily controlled by the banks. That is not going to be installable on CalyxOS or anything without either Apples AppStore or the Google Play Store.
Phones doesn't matter, they are mostly the same, the operating systems are interesting, but without a widely supported app store, it does not matter, the phone will be useless for the majority of people. More and more people are using their phone as their only computing device, they need to have access to banking apps, mobile payment solution and government application, none of those are not going to exist on some random Linux phone.
Then the question should be "which bank is backing this ROM", not "is app A for bank B work on this ROM".
The answer to the latter is and will continue to be an instant negative for quite a while, and it's pointless question to make other than as a really indirect nudging for big bank(lol).
What kind of reply is this? None of the people I want to give money will take Ethereum.
Then what makes you think the same people will let you insert a bomb-looking thing with wires sticking out into a credit card terminal?
Who said anything about sticking anything into credit card terminals? And I'm not seeing any wires sticking out of the phone. I'd say it doesn't particularly look like a bomb either, but given recent events, that probably doesn't say much.
You'll be asking the same thing by trying to pay through a totally custom phone. As I'm repeating here, the blocker is lack of trust from banks, not technical reasons. THEY have to trust YOU, and then banking apps will work. Otherwise any tricks to get past security check to manipulate your bank accounts will be just short-lived hacks.
If the King of England walked into the Bank of England and wrote a hand-written bond for GBP1k with a bic pen on back of copier paper, they'll take it. Electronic platforms has nothing to do with that. That's why it's stupid to keep asking if banking apps would work.
>If the King of England walked into the Bank of England and wrote a hand-written bond for GBP1k with a bic pen on back of copier paper, they'll take it.
what has gone wrong in your brain that you think that?
>Mediatek
If in any doubt about whether to buy this, look at Planet Computers offerings and the quality of their Linux build. They have been offering Mediatek-based Linux phones for several years. Spoiler: garbage with ancient proprietary kernel, thanks to Mediatek
Furi OS was what turned me away. Never heard of it and that makes me realise we don't have an android alternative OS which is permanent. Do we have any OS which is permanent? Shouldn't we have one?
FuriOS is a derivate of Droidian, which in turn is basically Debian with Halium support. So it's pretty much Debian + some custom patches for usability, hardware support, performance, so on. We intend to upstream as many of our changes as possible as time permits, but right now we are laser focused on improving the experience day in and day out.
Are all of your patches open source? Given enough time, could a user build a "factory image" of FuriOS from source?
Yes. It doesn't even require a lot of time, all the build scripts are on our GitHub.
Thanks for the answer. All the best!
Hi, thanks for answering questions here. How do you handle device encryption?
Once you turn on full disk encryption, the phone will reboot once and begin to encrypt the entire filesystem in the background. Further reboots will require that you input your FDE password (which can be different from your lock screen password/PIN) in a minimal rootfs-based UI.
Sailfish is an Android alternative (that can also run Android apps): https://sailfishos.org/
PostmarketOS is trying to be this.
What do you mean by "permanent"?
My specific definition: 20 years later when a retired me boots up the device, it will be able to run the calculator app without shutting down.
That sounds like a hardware problem. But if the hardware works, why would you not be able to open any old app?
I think they are worried about some cloud service shutting down preventing them from using their phone.
I'm guessing one of the engineers was born on October 15, 1982?
No word on HDMI out ?
That's the only thing stopping this from being a full computer.
i assume it would do video over the usb-c port like the pinephone and librem.
I don't like to assume anything when spending 500$.
I have a One Plus 12, which has HDMI out, but it's like they didn't feel like finishing it. It's not really usable, it doesn't have any controls like Dex does.
Dex is an outlier, other than Huawei's Easy Projection I'm not sure anyone does more than mirroring the screen
Motorola has a limited version of Dex as well.
The FLX1 hardware does not support this, sadly (as very many phones do).
Unfortunate name.
yeah, went looking for this comment...
I think we have different definitions of cheap
Finally, a phone without camera bump and with an audio jack. Screen is way to large for my pockets, though.
Just make sure all the critical components as well as baseband are implemented in Rust, and you can count me as a customer ;)
Let's not forget to also get it all audited, too.
Who will be buying these? Developers?
It's a Linux phone that runs Android apps, so they ostensibly don't have the problem of building the ecosystem.
Importantly though, it won't have Google Play and the increasingly mandatory GMS, so a lot of Android apps won't work properly.
Definitely cool though and worthy of some attention.
We do however ship with microG. I have heard of people using Uber, Spotify, Lyft, Signal, etc. So Android app support isn't that bad despite no GMS.
Amazing, thank you! You just sold a phone :-)
No video on usb c? Maybe I missed it but don't see it in the specs?
Why not use the open source Android? I just feel like it would be more mature.
And yes I like many others here have been trying Linux phones for 10+ years, since N900.
Your phone just has to work, it can't crash because you get a call, or hide buttons because the UI is confused by orientation. IT MUST WORK. So that's why I wonder why people don't just use Android.
Open source Android has been increasingly neglected by Google, and is (essentially) unusable at this point unless you are ok with the look and feel of very early Android. It also isn't likely to improve as Google's strategy seems to be to move toward more and more proprietaryness.
I think a much more sustainable approach is to focus on existing Linux distros and get them more usable on phones.
You are wrong about AOSP. I'm running LineageOS (Android 14) on both my phone and tablet without Gapps and it is a fully functional system. (NewPipe for youtube, Aurora/Fdroid for for App Stores)
I don't know what you mean by "look and feel of early android" either, as AOSP based ROMs look just like Google Pixel in the settings, quick dropdowns and dialogue popups.
LineageOS is not the raw open source Android though that the parent probably referred to. Afaik there are quite some changes needed by now to make it really usable.
But sure, LOS proves it can be done.
LineageOS != AOSP. They've put quite a bit of work on top of it too make it usable.
To be clear I have nothing against lineage, and I'm a Graphene OS user myself, but these projects add a lot of value on top of raw AOSP.
Still though it kinda proves my point. What's the quickest path to a usable phone? Quite a bit of work on top of AOSP, or even more work ontop of Linux?
It's like using Chromium or whatever clone-a-chrome is hip at the moment. That does absolutely nothing to disrupt the status quo and is very much at the whim of the upstream "owner" whether you want to see it or not.
Disrupting the status-quo is important, but having a working communication device is more important to me. If AOSP is further along towards a working phone then I don't see why we shouldn't focus development efforts on it.
It only benefits the open source market in the long run because a working device will garner wider adoption.
What control does it have over the baseband / modem or lack thereof?
This website is just infuriating[0] to use with the randomly disappearing elements (FF 131 on Fedora 39, even with disabled uBlock). $500 sounds kind of funny when PinePhonePro is $400, but of course on the other hand Librem5 is $1000+, so whatever. Still, (2-gen-old) iPhone price-range.
Couldn't find too many datasheets nor other documentation unfortunately, though it might because of the unusable piece of crap website, which made me rage-close it.
[0]: just noticed: inFURIating, as in Furi labs, haha
> on the other hand Librem5 is $1000+
$800 now:
https://shop.puri.sm/shop/librem-5/
But yes the "nice" one aka Liberty Phone is still $2000, unfortunately.
From their FAQ:
> Furi is pronounced “Fury”. FuriOS is pronounced “Furious”. We simply couldn’t afford the Fury part of the domain names.
design is kinda ugly + 500 dollars is not cheap
I honestly think it looks fine. Nice texture at the back, pretty metal accents around the camera and at the side button, good logo. The uneven bezels (the one at the bottom is larger) are less nice, but that's just a reality of the budget this probably moves in, not even the Fairphone 5 did solve that yet.
What is their GFX toolkit?
Because there debian, android? What is the bottom of it?
I've been waiting for a viable linux phone for years. Ever since the M$ destruction of the Nokia offering.
This one looks pretty good!
It even has a feature that should make it a great migration away from android:
https://furilabs.com/android-app-containers/
Chinese, unfortunately
There is AFAIK only one non-Chinese Linux phone, and it's Liberty Phone.
... which is $2k... We already know that almost everyone likes the idea of not being Chinese stuff, but no one is willing to pay 4x premium. The Liberty specs are dismal for a 2k phone; as I said (I think response to you as well); I would pay basically anything for descent modern specs (so s24 or so) but running linux, but not 2k for something that will just annoy me in not being practical besides for a bit of lisp (nothing wrong with that though!!!).
Furry-labs or Fury-labs?
OLED or nothing
Finally! (Fingers crossed.) So tired of decade-old unimproved hardware with no drivers from Pine or $800+ handsets from purism—quite old as well.
The starlite tablet from star labs is quite good. Wish all these tiny companies could work together somehow.
What I want is quality to the level of flagship android phones or the iphone with the ability to plug it in to a keyboard, mouse and screen and have it function as my laptop as well.
One single integrated experience. I think everyone wants this. We have the technology we have the demand, it's just companies who make the high quality phones want to completely lock down the market.
You would think the free market would produce a company that would tackle this issue but barrier to entry is waay to high so essentially we have small companies trying to get into this space but they are only creating sub-par products because they can't make something as good as an iphone.
It's basically a legal monopoly. Companies can't get into it because the technology and capital required are way to high and only super mega corps like samsung and apple can pull it off...
Closest thing I've seen to this product I'm looking for is the steam deck. I would buy a steam deck phone if it had the quality of a pixel/iphone device.
> I think everyone wants this.
You asked for it. I don't want that. I don't want an "integrated experience" and "flagship quality". The first sounds juvenile and the second sounds unnecessarily expensive and probably containing shit I don't need, like fancy cameras to look good on my nonexistent socials.
What I want is a simple, slow, old, efficient, simple phone with the interface of an 80s era 8bit computer that can actually, imagine this, make and, to complicate matters even further, even take calls.
I basically want an open source dumb phone. Do these exist? If not, why not focus on this first? Why go for fancy cameras and apps when we can't even make calls? Looking at your PinePhone.
LTE set that back by a lot. For years there were no VoLTE implementations in common use other than stuffs on Android. Even those had compatibility issues and lots of carriers still block unapproved clients trying to register on VoLTE.
For 3G, you could always do that. You only needed the right modem module with voice call support and audio I/O, like bare PCM pins, and a host micro to handle AT commands.
> there were no VoLTE implementations in common use other than stuffs on Android
What's wrong with the Android one - is it not permissively licensed?
I think the biggest problem of Linux phones is the community's obsession with trying (and failing) to reimplement (multiple times, in parallel) things that Android does really well and can be used as-is.
That's why the PinePhone or Librem 5 still can't even match the usability (at basic things like phone or camera or battery life) of a 2010-era Android phone, despite having similar hardware.
You want a Linux phone that actually works? Start with an AOSP-based phone and provide manufacturer-approved root and escape hatch such as first-party terminal and Wayland/X server app to run Linux apps.
Over time, you can slowly replace Android components with their Linux desktop counterparts when they're ready (or the other way around - the Android bits can just be the commonly-accepted solution to specific problems in Linux - even desktop - distros), but at least you're starting from a solid base.
> What's wrong with the Android one - is it not permissively licensed?
It is not, AOSP based distributions have to kang it from vendor builds. Qualcomm's is mostly standardized but Samsung wrote their own stack and voLTE/voNR won't work on any custom roms.
> That's why the PinePhone or Librem 5 still can't even match the usability (at basic things like phone or camera or battery life) of a 2010-era Android phone, despite having similar hardware.
They most certainly do NOT have similar hardware. You're wrong on thinking it's a software problem when the hardware being interfaced with is notoriously proprietary. The PinePhone and Librem phones are using self-contained quectel modems connected via different interfaces. They are nothing like the integrated soc's of nearly every other device on the market. This dramatically impacts battery life and stability and I don't think it will ever be a solved problem when building devices this way.
disclaimer: just an old grumpy customer, not an industry expert.
There were pointless debates around ITU-T and 3GPP as to whether the LTE is 3.5G or 3.9G or 4G next to the pink fact elephant that _it_ is going to be _the_ next cell standard anyway. That hot debate delayed voice call discussions to post-launch matter that eventually coagulated as the VoLTE.
The Voice-over-LTE was a total kludge together that were(ARE) carrier specific implementations. In Japan at least, it seemed to have had push from KDDI, esp. with hindsight that NTT docomo and Fujitsu, both formerly influential in 3G, both seemed on fire majority of that timeframe, while Nokia being a supernova beyond fire. It would be very natural if Samsung would have made vital contributions, but I don't have much informational pressure from that direction. VoLTE is phone company mannerisms, remnants of retro-futuristic media features of 3G, weird spaghetti codes from carrier labs, and international call exchange system, all homogenized in a blender; the artefacts were co-developed pair of server and client implementations that are standardized in the way there aren't many implementations but not something carefully spec-worked before construction for interoperability. That all happened ~a decade ago. Some point between Android 4 to 10.
I suppose it was all retconned into 3GPP standards after action, but the whole stack is still like a lobotomized Android call app and private fork of Asterisk embedded into the cellular core monstrosity. I guess embedded modem people(like Quectel) had finally got to port functions into modem chip firmwares so they can make calls, after someone done it for non-smartphones(KaiOS, Smarterphone...), I think around 2019 +/- 1 year.
What I'm trying to say is, VoLTE is complicated. It's something like SIM-authenticated SIP/RTP over IP under IP, not even regular 3G data session let alone SIP-VoIP on Layer 3 UDP/IP.
Osmocom project Wiki summarizes it better than I[1]: "Voice over LTE is an adaptation by 3GPP to use IMS over an LTE cellular network. The LTE EPS (Evolved Packet System) provides the functionality of the underlying IP-CAN. ... IMS is much more than normal SIP/RTP. And in addition to that, there is a tight integration between the LTE system and the IMS on top of it."
And it's supposed to be the core feature of a phone, "not".toupper() a party trick... There are better things to do in life for most people.
1: https://osmocom.org/projects/foss-ims-client/wiki/VoLTE
> that can actually, imagine this, make and, to complicate matters even further, even take calls
Showing how different different people are. I've probably made... 20? ish? phone calls in the last decade. (I'm 46.)
Holy moly, you must have very different life than most people I guess. I am from the younger generation which supposedly avoids making old-style phone calls as much as possible and I don't think I know anyone who has made less than 100 calls in the last decade.
I always wonder why your generation avoids such calls... Voice calls are so much more efficient to get the things you want from the person you call...
I think those very total attention expected during phone calls, and the speed at which it happens, contributes to a higher level of anxiety compared to texting. The younger generation hate these elevated stresses and prefer more async communication
Same for me, I am 50 and would happily carry a phone sized tablet without call caps at all. I never need to call or receive call; send me an email or hit me up on a chat channel.
>I basically want an open source dumb phone. Do these exist? If not, why not focus on this first? Why go for fancy cameras and apps when we can't even make calls? Looking at your PinePhone.
Your dream is achievable and has been achieved. PinePhone for example.
I obviously don't mean everyone everyone. More like the overwhelming majority of people.
I have one and it very much struggles to make and especially to receive calls and is still trying to be smart. PinePhone is definitely not a dumb phone. It's expensive as well.
To me a simple, dumb, open source device that can easily be manufactured in all kinds of conditions all over the world sounds like a dream for actual, practical purposes. Like, for example, again, calling. To some degree I have the same issues with "smart watches". Simple, open source, dumb smart watches with just a smidge of 8bit CPU goodness to display, say, something simple like a word or even a letter on the screen would be quite useful. I know there is some movement in that arena using ESP32s but I am not particularly impressed. Alas, alas. Why do we as a civilization tend to go for the extremes and not just get our basic shit together first?
I was being slightly obtuse and I understood perfectly well that you meant all reasonable people. If I can't be obtuse and pedantic on HN though, where else?
These do (or did) exist: feature phones shipped with FirefoxOS. Almost no one bought them and the effort was widely seen as a failure, although KaiOS enjoys ongoing success in the developing world.
https://en.wikipedia.org/wiki/Comparison_of_Firefox_OS_devic...
https://en.wikipedia.org/wiki/KaiOS
I am using a KaiOS phone, a Nokia 6300 4G. It works well as a 4G WiFi hotspot for a laptop and as a small phone for making calls and sending texts. Have a few apps on it for mapping as well as using the browser occasionally.
The KaiOS WhatsApp app will stop working next year though.
Wow. Whatsapp struggles on my iphone 13; how's performance on your Nokia?!
Works fine to receive text messages and a few pictures.
I beg to differ. Tried to use a kaios nokia 800 tough. Firefox/KaiOS using a stupid kind of webview for every UI, you get two consequences: first, everything is ugly: wrong padding everywhere, complex layout with non-integer number of list items fitting on screen (this is important, vertical list is the primary widget on small screen dumbphones). And second: everything is sluggish. My true "series 30" nokia launches app instantly, whereas i'm sure the SoC is order of magnitudes slower.
So the UX definitely shows complete absence of culture of design in this kind of space (eg gameboy UI design, classic dumbphone UI, ...). Small screens need very simple widgets (most importantly grid and vertical list), hand made paddings and proper space optimizations, most likely some bitmap font tailored to the resolution.. You can't just slap some dynamic layout and expect it to work nicely on such a small screen. Everything vectorized will be both costly and of dubious rendering quality, etc..
And of course these things only have "decent", but definitely not "1 month scale" battery life. I guess LTE is the real culprit here.
I half agree. I dont need the bullshit bells and whistles. However a good camera doesn't mean you want to post on instagram or tiktok - maybe you just want a memory of a trip you went on or technical detail of something you want to document. People took pictures before social media.
Most every phone can already do what we want but the hardware is undocumented or locked behind NDA and the firmware is hostile. Without those locks the gigacorps cant keep you in the data-mining garden. We wont have open phones until this thinking changes. Keep dreaming until then.
I agree I would love this option, but how much would you be willing to pay for such a device? Would you pay $800? $1,000? $1,400?
That's where I have a hard time. I would pay that kind of money, but I would need something well polished and fully capable of being a "daily driver." I think many people are in the same place I am, and thus we have a real chicken and egg problem here.
Microsoft at one point was running full NT on ARM on phones. I think they could've done it within such price range, but didn't.
Didn't they? The Lumia 950 launched at about $500ish and supported Continuum.
Continuum wasn't exactly full convergence, but it was kind of close. As I recall, they had a desktop dock and a laptop like dock, that you could pair your phone to (wireless or wired).
Microsoft was big on their 'universal windows app' concept at the time, where universal meant actually only ran on a small fraction of windows devices. sigh
aw. I completely forgot about that. Why didn't it just work? Was it... during those "we invest all our cash to everything needed for total Win32 dominance but boldly let it rot in the backlot" days of Microsoft... yes yes it was.
Hmm. If I bought a phone for ~$600 and a laptop for ~$1000, I guess I’d pay ~$1700 for the converged device (adding a little bit because of the intangible benefit of not having to manage my files anymore).
I'd pay up to 3k if it was quality delivered the experience of a top quality phone and a top quality laptop.
I mean the technology to do this is already here. If apple or samsung wanted to do this... they can.
Right now I use an iphone, but if samsung made a phone that felt like say linux, windows or macos when in laptop mode... I would switch off iphone in a heartbeat.
And i mean it has to feel like macOS. None of that bloat is acceptable. SteamOS actually pulls this off but in a gaming form factor.
> And i mean it has to feel like macOS. None of that bloat is acceptable.
I suspect that every single thing you like about macOS is something I consider bloat (or would if that was the only related word I was allowed to use).
My first task on my recently acquired M3 MBP was to remove/hide/disable as much of macOS as I possibly could; only then can I use it as a productive development environment.
> and a top quality laptop. > I mean the technology to do this is already here.
I think it really depends on what precisely you mean here. I don't think it's currently possible to get the same performance of my M3 Max MacBook Pro in the form factor of a phone.
Asus did that in 2011 with the PadFone, it was an Android mobile phone that you could plug into the provided tabled and then plug a keyboard into that.
There was also Motorola Atrix in 2011 with similar idea.
Yeah but it wasn't quality right? I mean the both modes need to have a UX on par with a regular laptop and a high quality phone.
I remember Motorola showing off a concept like this in ~2012.
> What I want is quality to the level of flagship android phones or the iphone with the ability to plug it in to a keyboard, mouse and screen and have it function as my laptop as well. > > One single integrated experience.
Purism achieved that with their Librem 5 phone [0]. However having a polished experience like with iPhone requires to invest billions in software. It can be used as a daily driver [1] but there's a lot to improve yet. Sent from my Librem 5.
[0] https://puri.sm/posts/converging-on-convergence-pureos-is-co...
[1] https://puri.sm/posts/my-first-year-of-librem-5-convergence/
I can do something not unlike that with the second-hand displaylink adapter I currently have gathering dust... I would just need to install a desktop environment on termux for it to be complete (not sure if I could get 3D acceleration with this). The only real problem there is the adapter I have cannot charge the phone while connected.
> One single integrated experience. I think everyone wants this
This is insane to me. Like not just no. Hell no. I want my phone to be a gadget, like a watch, not a whole ass computer. I want my tablet to be its own thing from my laptop, which is different from my PC(s). I absolutely, positively do NOT want general purpose computing to be the typical paradigm. I don’t want, need, or even like the idea of my phone, that I typically use like a wallet, to be a general purpose computation tool. Lock that shit down, please!
I should never say "everyone" anymore. Because everyone is almost never true. I mean overwhelming majority.
I think adults should read 'majority' when someone says everyone. I will keep saying it when talking with adults as it is just easier than put all the exceptions; if some people get annoyed by it, that's not really my problem.
Alternatively you could use 'most people', or any one of several equivalent expressions, and preserve the meaning of everyone as every one. There is an aroma of manipulation around using language that gives the impression of absolute consensus even if all parties involved understand its vernacular meaning. It's embellishment in service of influencing people and rightly denounced in any context where accurate communication is important.
But we are not writing a scientific paper here; we are having a discussion. I guess if it matters in the scientific sense, one would be more careful; so if the discussion would be about a new type of medicine then 'cures all people' would be a stupid thing to say. Here however, it doesn't matter. But point taken: I guess I commented as HN has especially has this nasty about of commenting on things where it indeed doesn't matter, stuff like 'citation needed' or 'but it cannot be most people, because I know blah blah blah'. Pointless let's say for this phone discussion; might be relevant for a medicine/vaccine or so.
Your phone is already a general purpose computer...
> I want my tablet to be its own thing from my laptop, which is different from my PC(s).
But what's the point in owning three devices when you can own one doing everything? How about in addition an mp3 player, a camera and a flashlight, too?
I like tools to have defined roles, not multipurpose.
I mean the top post is a ~200USD “smart” pomodoro timer. I’m not even close to that.
I use my phone for communication, quick photos/ notes and sound. It carries my cards. I want it to be completely locked down because it gets used in public and could be stolen.
I use my laptop for personal business including coding. I want it to be secure and streamlined.
I have a gaming PC, it runs widows and gaming rootkits and all kinds of questionable software. I don’t want my sensitive data around that computer.
I do art on my ipad, I want it to be about art. I dont need the phone stuff there. I don’t want my personal stuff there.
> flashlight too
lol I actually carry a flashlight with me always. Phones are absolutely garbage compared to a modern High CRI flashlight. I also have a FF camera and take it with me very frequently.
But it got electrolytes?
"I love the new furry phone"...
Just like DuckDuckGo I think the name will hamper adoption of this.
It's "fury phone":
> Furious Support from the FuriOS team
They could say it's pronounced "Wendigo" and people would still read it "Furry".
If people read "fury" as "furry" then that's a literacy problem.
You’re right.
And if you can’t understand why almost everyone will read ‘furi’ as furry and not ‘fury’, then you’ve got a phonics problem.
As someone who reads/watches a lot of Japanese media, "fury" isn't how I would first read "furi".
I do like "FuriOS" and think it avoids this issue, but other portmanteaus might be better avoided.
Have to say, main reason for clicking to check it out was I thought it was a Phone for Furries.
Curiosity got me to look.
Maybe it will drive traffic.
[dead]
$499 is not cheap. At all.
[dead]
A great thing about Android and iDevices is their departure from yet another UNIX clone using C and C++ as the main application languages.
As such, most of these attempts aren't that attractive to me as OS nerd, always recycling the same ideas.
Still haven’t achieved a workable floss device with old ideas.