I got my librem in the mail last week after waiting several years(!!) for shipping. It's complete trash. The software is shit, the battery is shit, it's expensive, low specs, completely unusable.
I'm willing to compromise a lot in order to go with the free software approach, but my god, just do better. There's no excuse for how un-functional this thing is. Try harder.
I had the same reaction with the pine phone a couple years ago. At least that one shipped sooner though.
> Nokia N900 was a beautiful, functional thing running linux and Free software 10 years ago.
All my horses on this being the #1 reason why Microsoft panicked and pushed Nokia into a partnership that ultimately forced Nokia to sell its mobile business to Microsoft, which of course meant the end of Linux in those devices.
I picked up an N900 for next to nothing as-is. The charging port was damaged (but that wasn’t really an issue, this was when you could swap batteries so I just kept two with an external charger) and the software was... damaged. For some reason a lot of packages were broken or removed. Not sure what the previous owner did.
But I didn’t really care. No calculator? No problem, I just fired up Vim while I was bored in class and wrote my own. Contacts book would crash? Again, a simple program I wrote let me quickly grep through a CSV file and add/update/call people.
It was such a joy to use I would carry it around even though I ended up using an iPhone as my daily driver.
I would, on the other hand, leave the CPU as it is, and welcome any change to battery life. That's the only factor preventing me from using one as a daily driver.
It had a well-polished Linux-based OS, developed in-house by Nokia. Excellent hardware for the time, and a keyboard which was a delight to use. It got pretty decent reviews - even outside of the Linux fanboy bubble.
It's just a good phone which happens to run Linux, not a mediocre phone hyped because it is one of the few Linux options available.
> Maemo Leste is an operating system currently in development. It is a modernised and liberated version of Maemo 5, rebased on top of Devuan with a mainline Linux kernel. The first operating system images were released in February 2018.
> It is currently in a usable state with support for various targets such as the N900, Motorola Droid 4 and the PinePhone.
I mean, Android is mostly Linux. I don't think the kernel is the issue. The issue with librem is that they put very little effort into the basic user experience of the software.
The point is that the n900 was Gnu/Linux with glibc like mama said was good for you and worked along very similar lines to your desktop machine.
Maemo was gtk+ based and many Gnome applications ran really nicely. Eg Gnumeric was great. The desktop dev environment worked so you could port or write stuff from scratch then load it onto your phone. apt-get worked - apt was the app store iirc.
But who cares that stuff?
It was as slick and nice a user interface as the then recently released iphone! No really, I'm not making that up. Fold-out keyboard was terrific too. Way, way ahead of android at that time and arguably still. From memory the touchscreen on the iphone was better but nothing else was.
Microsoft bought nokia and killed it. Nothing has been as good since.
Not really, in what concerns application developers.
Only the Java/Kotlin userspace and the small set of NDK libraries are considered official and stable.
Anyone that ventures outside this path and isn't an OEM doing their own Android distribution, is on their own if the application actually works at all when deployed via the PlayStore, as the Termux guys eventually found out.
The saving grace of the Pinephone is that it is only about 1/8th the cost! They set the expectations realistically. It is a hacker phone for hackers.
It feels like as soon as the community is moving in the right direction, the entire state of technology has moved on to something else and they are constantly trying to realign.
Ain't we are all responsible for "duping" Purism into making that phone? Just look at the hypes after the announcement and how people loved it on paper. It's a classic example on why you should never fully trust your consumers for their suggestions.
Based on those hypes, Purism probably thought they can just create a phone for hackers first... and never realized the reality that even hackers loves consumer products. Just take a look on how many hackers owns a MacBook as their main driver despite the existence of many good Linux laptops. People are hiding some truths here, and they wanted a smart businessman to figure it out.
As one of the hypered, I'm sorry, Purism. This is the baseline that I actually wanted from a handheld device:
- Runs open source Android and Linux without bug, maybe also iOS;
- Powerful, as least 8GB of RAM and a relatively fast CPU;
- Big swappable battery, 4~6 hours continue use without a re-charge;
- But most importantly, INexpensive, at least not far more expensive than their consumer counterpart;
- Then you can start thinking about ergonomics, self-serving ability, usability and extensibility (builtin keyboard, swappable modem etc)
I should have listed all this clearly in my hype post...
Also received my librem 5 recently, and it's incredibly disappointing. The software in particular is incredibly buggy and janky. When I open the app store it just fails to load things sometimes. Even the FirefoxOS device I tried a few years ago was better.
> When I open the app store it just fails to load things sometimes.
Considering how many desktop linux distros fall down with this, you may be expecting too much of your beta-level phones. It's been years, yes, but power management isn't exactly easy. Just look at how bad Linux laptops were even just a few years ago - hell, look at how long it took the multinational mega-corporation to get Android's background power consumption down. It was so bad people on XDA set Tasker to turn off BT, location, and even wifi when not in use.
You're right, FLOSS phones currently aren't good. But they finally exist, and that's enough for me to be excited about
> You're right, FLOSS phones currently aren't good. But they finally exist, and that's enough for me to be excited about
Also people forget that android is still FLOSS. You don't need to be running upstream/mainline linux to be FLOSS. It's mentioned kinda off hand as "oh well running android without google sucks" in the article but graphene OS has proven that you can get a very pleasant, well ironed, and secure/hardened user oriented phone while using FLOSS for basically everything except some firmware level blobs for high end phones. And FDroid proves that we can have 90-95% of all the quality apps we could need without even installing anything google dependent (even FCM push notifications which were historically limiting can be replaced with unifiedpush) That experience contrasts with the "FLOSS phones" like the librem or pine phone which show that you can get most if not all of the hardware working without firmware blobs even if the software up the stack sucks.
So what is stopping us from just starting with one of the generic AOSP targets for graphene os and focusing it for one of these FLOSS targets?
Do we just not have the specs in the hardware with decent FLOSS support? If that's the case, doesn't that mean we are quite literally steps from the finish line?
Honestly it depends. The PinePhone Pro (which I've only briefly interacted with and do not own) seems like it works well, if a bit buggy, and with a fairly unoptimised UX derived mostly from desktop linux. Given that there's a handful of supported OSes/images (between android and linux) and just as many unofficial ones, it could be that we have only interacted with less polished SW.
Also FOSS GPU support depending on what device/SOC you have really varies. So weak GPU support can end up handicapping the experience.
Personally I'd like to mess with an android pinephone pro (not the stock "linux with android containers") but I'm not willing to foot the bill to set one up to play with.
I know it's not free software, but I find the Jolla/Sailfish phones, the spiritual successors to the OS of the N9, a decent tradeoff between freedom and usability. It even runs android apps!
Valve should try their hands at it. Steam deck already makes up ~40% of linux desktop installations. By the end of this year, it should be more than half of all linux installations.
> Steam deck already makes up ~40% of linux desktop installations
Do you have a source for this? This sounds a bit hard to believe, and I can't find a source. The sources I found show that within Steam installations, Steam Deck users are a significant share, but not the majority [0]
Sorry to hear that. I’m waiting for my Librem 5 myself and I’m also bracing myself for disappointment, especially given the years of waiting.
A small consolation may be the fact that not all the software development effort we’ve paid for is lost. On the contrary, the work Purism has done both in userspace and kernel have immense value. One day, another Linux-based phone will appear, and it’s going to stand on the shoulders of what Purism did. It’s never going to be perfect, or even nearly as polished as the duopoly UX we’re accustomed to. But it’s going to be good enough, no matter if Purism will even be around by then.
Also, the software for our Librems will only become better with time, as annoyances are ironed out.
tl;dr Purism is definitely underdelivering, and we have all reason to be mad at them. It’s only going to get better though.
it simply cannot be permitted. somebody realized PC market/platform made a huge "mistake", letting anybody make software for them and permitting random hardware. this 'mistake' was avoided on smartphones for very pointed, clear, intentions of social control.
there exists NO homebrew smartphone hardware like for PCs.
Your conspiracy explanation doesn’t pass the sniff test. There’s a much much simpler explanation: lack of demand.
Heck, you’d think people on hacker news would be all over open source alternatives to proprietary hardware, software, and services. Yet take a brief wander over to one of the many (many) recent Reddit crisis threads and you’ll find countless naysayers talking about how lemmy and the like will never succeed, despite gaining hundreds of thousands of users in a week and a half.
>Heck, you’d think people on hacker news would be all over open source alternatives to proprietary hardware, software, and services.
Yes, every PinePhone thread has people posting that it's unusable because it's not as good as a flagship iPhone or Android. Or the usual "not ready to be a daily driver" ... because the definition of "daily driver" is apparently a constant of the universe written by God so everyone is expected to have the same one. Meanwhile I've been "daily-driving" my PP for two years at this point because calls, SMS, voicemail, Firefox and the camera work well enough.
This is kind of my biggest gripe with Pine64 (beside the unusable modem in the Pinephones, which crashes all the time): the community. Every time I google something related a pine* product problem, there is at least 1 result from pine64 forum where the answer is along the lines of "if you don't like it, go away. we are happy with it as it is, there is no solution, and you shouldn't be seeking one. you are bad person for even asking. pine* is not made by apple/samsung, you can't expect it to do things".
I have quite a bunch of pine64 devices, but after I saw these interactions, I didn't even dare to ask anything nor to interact with them on any level.
>(beside the unusable modem in the Pinephones, which crashes all the time)
Install the OSS firmware. Quectel firmware is shitty. For me the OSS firmware doesn't crash at all, though I've heard other people say it still crashes for them once a day or so (?). In any case eg25-m will bring it back so it won't be "unusable".
>"if you don't like it, go away. we are happy with it as it is, there is no solution, and you shouldn't be seeking one. you are bad person for even asking. pine* is not made by apple/samsung, you can't expect it to do things".
You only get these responses if you reveal yourself as someone who didn't read the warning labels before buying the hardware and/or if you act entitled to a solution, as AndyKelley did in this thread. We indeed do not have any patience for someone who comes in and says "It's complete trash. The software is shit, the battery is shit, it's expensive, low specs, completely unusable."
I have tried each every available modem fw versions, including the closed source ones, and all possible versions from biktor also (I did run multi-day experiments with each of them, taking notes about their stability, looking for a working version). Every single one of them crashes, the only difference is the time spent between crashes. Maybe it is usable for you. My definition is different - if I can't rely on it for any period of time, it is not suitable for any purpose, but for messing around.
I won't try to convince anyone about the (un)friendliness of the community. If you enjoy its current state, good for you. I, for one, avoid it. I guess everyone can decide what community they want to participate.
While GP's language was definitely coarse, he does make some point. Personally I'm in for the hacking, I care relatively little for other people's (userspace-) software on these devices. But I do see a lot of sentiments around the web saying things like "look how good phosh/kde/XYZ is running on the PP, wow, such speed, many responsiveness" - even though it is objectively not running good when tried in person. Such statements can definitely trick users who just want a novelty phone - they read these reviews, and think that finally there is a viable alternative.
(Disclaimer: while I'm not a fan of some aspects like userspace sw and community, I'm a big fan of most Pine64 hardware, and I not only don't regret any cent I spent with them, I plan to throw more money at them.)
>But I do see a lot of sentiments around the web saying things like "look how good phosh/kde/XYZ is running on the PP, wow, such speed, many responsiveness" - even though it is objectively not running good when tried in person. Such statements can definitely trick users who just want a novelty phone - they read these reviews, and think that finally there is a viable alternative.
Either you're saying those reviews are lying / wrong, in which case the problem is not the product but the reviews, or you're saying those reviews are accurate but don't reflect the use case of the buyer for whatever reason, in which case it's not the problem of the product nor the reviews.
In my very first comment I said that he was welcome to come onto the IRC etc and ask if the phone can do X, Y, Z that he needs it to do, and he would've gotten an honest answer for that. You've decided that you don't even want to do that because you've been soured from the community for whatever reason, and frankly I don't care to convince you otherwise.
In the same way, AndyKelley spending a $1000+ without doing the research and then whining about it is his loss, not mine. Even in https://news.ycombinator.com/item?id=36414414 he's still going on about "The issue with librem is that they [Purism] put very little effort into the basic user experience of the software." because he thinks it's an iPhone or Android that only has one choice of OS or UI.
The naysayers are jaded folks who have watched two generations trade all privacy, freedom, honesty, transparency, openness, and even ownership of their own data for convenience.
Judging by consumer behavior it appears that ease of use is more important than virtually any other concern.
People who really care about things like privacy and freedom are too few in number to tilt the market, and this is compounded by the fact that as a rule FOSS people don’t pay for things. So the market here is a small number of cheapskates.
I think the intersection of people who were outraged at Reddit’s API change and “average consumers” is very small. They honestly aren’t important to me because for the most part they’re lurkers anyway.
The people I care about are the power users, the posters, the creators. People put effort into building communities. Those folks are very few in number but extremely important. They can create a community anywhere they go.
They’re also willing to put in effort to begin with so it’s generally easier to get them to move to lemmy when the experience may not be as polished.
but I'm not trying to imply anything. I merely wonder if such a thing (to create or destroy demand of something) is possible; and if it were, how would you do it?
surely it involves marketing? but I also want to reflect on the market cap of convenience...
Linux phones are a niche within a niche. Marketing won't do anything to change that. Most people are satisfied with iPhones and Androids, and the few that aren't are mostly satisfied with degoogled Androids.
It’s not a conspiracy. It’s just economics. There is no money in it so there is no funding available for it. Things like Librem are practically basement operations compared to Apple, Samsung, and Google.
Mobile software is far harder and more costly to produce than desktop because the small form factor and rapid casual use cases mean there can be zero tolerance for bad UX. The amount of effort that goes into just getting things like touch keyboard and touch gestures just right is monumental. The tolerances are very tight.
PCs with big screens and keyboards can support I/O paradigms that are much less demanding of perfection or require much less polish to make usable. A PC can just put some buttons on the screen or use a command line and that works fine. A phone UI needs obsession.
What efforts like Purism should do is greatly reduce their scope. Make a device that does the basics which would be phone, text, basic web, and maybe a media player or support in the web browser for this use case. At first anything else could be accessed via the web, and maybe look at introducing PWAs as a way to do basic apps.
Keep the GUI as simple as possible with the smallest number of options and features as possible and focus more on perfecting what is there. Ship a more minimal better polished product. If you don’t have the resources to perfect it, drop it.
Make anything more custom or advanced configurable via ssh, but make the core features entirely usable without doing that. At first you might not even have an App Store, just a CLI package manager you can access from a PC with ssh.
It’d basically be a hackable minimalist phone. Someone who just wants a privacy oriented minimal phone could use it as is and power users could ssh in and hack on it.
At this point I think the task is simply impossible, the minimum required features on a phone are too high now. If it doesn't have all day battery life, maps, and have a good camera, it's useless. So all you'll still have a real phone with you anyway.
We may as well be talking about an open source, community developed trip to the moon. There are no viable intermediary steps. You either have billions of dollars to find it, or you can't do it.
If you only care about a portable hackable device, the steamdeck is probably the product for you.
I believe there are some limiting laws/regulations around radios that make DIY, free, hackable ones usually not allowed. So, modems, WLAN cards, wireless routers... tend to be quite proprietary.
>The software is shit, the battery is shit, it's expensive, low specs, completely unusable.
>I had the same reaction with the pine phone a couple years ago.
Anyone who writes "the software" like you and TheAceOfHearts did, as if it's an iPhone or Android, doesn't understand the first thing about a Linux phone. Please resell them to someone that's less PEBKAC than you. Thanks. They'll appreciate it.
And in the future I recommend asking people who have them before you buy them, so that you know ahead of time not to keep these phones from going to people who will make better use of them. For example, we in the PinePhone IRC channel will happily listen to what you want from the phone and tell you whether the phone meets those requirements or not.
They shouldn't need to understand it. I don't understand the first thing about how my phone works or mobile development in general. It just works and does what I need it to do.
You don't need to understand it according to Librem. They are marketing a "user-friendly"(their words) replacement for iPhone and Android. If this was a product sold for the most hardcore Linux users who dream about bash scripts then yeah this isn't a product for me and that's fine. But, that isn't what they are trying to sell and that is the problem.
>Beta Edition PinePhones are aimed solely at early adopters. More specifically, only intend for these units to find their way into the hands of users with extensive Linux experience.
I have a Braveheart edition Pinephone for experiments. I know that Pine64 devices are very bluntly a work in progress. I'm more specifically talking about Librem in this case.
The only people who wonder about the year of the Linux desktop are the people who don't use Linux :) We Linux users are happily using it already and don't really care what other people use.
>> We Linux users are happily using it already and don't really care what other people use.
And yet here you are hectoring the plebs, putting us in our place. It sure seems like you care, because if you didn't you would have kept quiet and got on with the good life.
As someone who has been running Linux as sole OS for over a decade now, I wouldn't say I am "happily" using it. It is acceptable and all alternatives are even worse in a lot of ways, but the Linux experience is definitely still lacking quite a bit.
It's still less user-hostile than ads in your start menu and asking the overlord for the privilege of having rights to use your own computer. That's good enough for me.
Someone should do a Linux phone that scales way back:
1. you can dial phone numbers somehow
2. you can talk on the phone
3. you can hear the person on the other line
4. the cell reception on the phone is decent quality
5. antenna is properly placed for very good reception
6. basic SMS works, all the time
7. has been extensively tested to work with a single carrier, any other carrier is no warranty
8. boot time and shutdown time are so small that it opportunistically shuts down instead of suspend
9. bonus goal: contacts, can maybe play snake
Librem 5 fails #5 and #8 (and can't really succeed at #8 unless you just drop into a command line, but then you wouldn't be able to input phone numbers). I've read some reports that #1, #2, and #3 have improved enough to be usable. On #7, they offer a bundle with a simcard for a carrier that was tested with the phone. (Doubt it was extensively tested, though.)
Of the little I've read about Pinephone and Pinephone Pro, they've had problems with #3, #6, and #7, and #8.
Overall, however, both appear as a smartphone complete with messaging and browsers. But Librem 5 can't do hardware acceleration for Firefox ESR (because the etnaviv driver doesn't have recent enough OpenGL features for webrender to work).
Pinephone is too slow to browse.
I'm not sure about Pinephone Pro (though I'd be quite surprised if it passes the Firefox webrender tests).
I'd much prefer a much more basic device that gets all the basics right, and then be pleasantly surprised that I can play mp3s on it, too.
The meaning of the word "phone" has shifted in the recent years, and your requirements are a good representation of what that word meant in circa 2007; recently however the requirements 1-8 have so little relevance that it's possible to operate a device without satisfying any of them [0]. The new meaning of the word is an expanded version of #9 whereas it's a connected gadget that runs apps from the well-stocked built-in store.
Any newcomer, Linux or not, would immediately face the most obvious problem, which is a complete lack of a well-stocked store.
[0] I sometimes take my smartphone abroad for a combination of offline Google Maps and WiFi hopping.
For Linux devices, if you have a powerful enough phone (Pine64, can we please have a really beefy PinePhone?), you could bundle Waydroid + Aurora/F-Droid.
From the original list, I really don't care at all about 1/2/3 (phone calls) and only care vaguely about SMS.
My requirements are for an excellent browser and an app store with support for messaging platforms (telegram/signal/skype/slack/teams/etc) and email support, everything else is either "extra" or an outright distraction I'd probably be better off without (twitter, reddit, etc).
My phone exists mostly for YouTube, WhatsApp, and providing a wifi hotspot for my desktop computers. (It is irritating that much about the phone itself assumes that the mobile broadband is a fallback, and that I have a separate internet connection to give me wifi. I don't. It sometimes refuses to install updates over mobile broadband, and insists that I find a wifi connection from somewhere, by going out to a café or something.)
I think that's exactly the point: someone sells something as a phone, but doesn't even have the most basic features every phone since way before 2007 has. Sure, for an actually marketable product, you also need wifi & bluetooth, a good camera, a nice design, a UI/UX people actually want to use, apps for at least a couple common things and probably quite a bit more, but you also have to get the baseline of being a phone right, which at least most of the points are about. You simply can't skip those if you want to sell a "phone".
> You simply can't skip those if you want to sell a "phone".
I recognize that's true for you, but it's definitely not true for me, and I expect many others, too. I'd honestly be fine if my "phone" dropped the "phone" part entirely. The only purpose of that feature is for spammers to make my phone vibrate a couple times per week. In the very rare (less than once per month) instances I need to actually talk to a human on the device, I'm sure someone could invent an app that allows for that.
I've been daily driving a PinePhone running postmarketOS + phosh for the last month and here is what I've found:
1. Yes, has a dialpad. Works as you would expect
2, 3, 4. Yes, works well on my carrier
5. The antenna is along the side of the phone I believe and the reception seems good, I haven't had issues with this
6. I would say SMS only works about 95% percent of the time. The phone does randomly disconnect from the celluar network maybe once every day or two
7. Works well on my carrier, no issues here. I have a data plan and it works fine
8. It takes maybe a minute or two to boot, but to be fair my Samsung A52 Android takes longer to boot, maybe 2 minutes
9. Yes, there is a contact app, don't know about snake :)
Other:
10. The phone is slow, firefox works and all the websites I visit (including google maps, uber, etc.) render and work correctly, but slowly. I wouldn't say too slow to browse, but 100% it is slow
11. Be CAREFUL inserting the SIM card, it is extremely easy to break the SIM pins and brick the phone if the SIM card is not inserted correctly (I learnt this the hard way)
12. Surprisingly a large amount of apps available via flatpak/flathub, including both foss apps for things like email and 3rd party clients for things like spotify and whatsapp (text only, no video or audio calls). All the core linux stuff and lots of the foss apps via apk
13. However not all apps on flathub are compatible with the PinePhone since PinePhones uses aarch64 architecture, not x86_64
14. Very poor battery life out of the box, about 4 hours tops. However, after installing tlp and irqbalance (via apk) and setting auto-suspend to 1 minute, the battery life is about 8 to 12 hours of "normal" usage, so quite usuable
Pinephone Pro here, I wanted to use it as a home automation controller around the house, I think I bricked the keyboard case attachment by using the wrong charging port (why is that even a thing??) because it doesn't work, but to be fair, it never did work, and I've tried troubleshooting it for a few hours with no joy, so now I'm just kinda too frustrated with it to sink more time into it and it's sitting in a drawer.
I should probably try postmarketOS on it and use it as a burner travel phone.
You only mention installing, but did you have to configure tlp & irqbalance? Battery life is all that's keeping me from using my pro as a daily phone :)
If you're going as basic as OP wants you can get 10+ days on a typical phone battery regardless, just by implementing battery saver. I've done it with both smart and feature phones.
That's what I had in mind. (See my edit above from earlier.)
Especially since both Pinephone and Librem devs are essentially dumpster diving for their hardware. Even Dell with all its resources couldn't get fast suspend on Ubuntu working (or whatever it's called). And that was just x86_64. But even I know how to power an RPI on/off! (I can even run the systemd thingy that reports startup times for various processes, which already puts us years ahead of the competition.)
If we work smart I honestly believe we can have a Linux device with the feature set and battery life of my LG VX5200 by the year 2030.
Edit: here's a sheet on that wonderful little indestructable device:
What do #1-#8 have to do with Linux? As far as I can tell, nothing. A flip phone will cover your use case nicely.
Personally, I have no interest in making phone calls. 99% of the time when my phone rings it’s a spam call. The other 1% of the time it’s one of my uncles or aunts, all of whom are in their 80’s or thereabouts. After they pass on, I will have no reason at all to make a phone call, except perhaps to book a doctor’s appointment with some doctor who hasn’t switched over to online booking.
Phones really need a hardware button or key combination for this-- tap out SOS on the power button or something.
It may well be the last phone call one makes in their life, but the ability to do it at all is predicated on the integrity of a piece of glass. It's insane.
Obviously situationally dependent (where you live, why you're calling) and of course you should have the ability to defend yourself if the cops fail to preform, but all that aside, that is not a universal truth. I called my local PD # a few weeks back because someone on a scooter got hit by a car in front of my house and I called the police (local PD directly, not 911, 911 would have been slower) and they had a unit at my house in UNDER TWO MINUTES. It's literally four minutes to drive (sans lights and sirens) from the police station to my house.
I know my experience is anecdotal and not universal, but neither is the "cops won't show up" one.
Not advocating for this, I do not own one, but I'm pretty sure this substantially covers all goals listed minus #8 (plus an unlisted one - Signal integration).
I investigated and considered this option carefully. Ultimately ended up doing a Pixel running GrapheneOS (no microG) as a stop-gap until I buy either a Pinephone or Librem, once those are a little more usable.
Punkt if it had an open software stack would have a sizable market, not main stream but definitely something reasonable.
Something else that platforms like that, and for everyone else, is a standardized means of communicating with banks. That way it is not up to the bank to build and maintain apps for seperate platforms. That one devices like Punkt could build alternative apps to do these kinds of functions.
Great in theory, a nightmare in practice. It would open up the means of fake apps to steal details. Also as the XKCD comics said - You have 15 standards, lets make a new better one, there are now 16 standards.
Side note on Punkt. They have one of the best monthly emails and I would highly recommend. Usually only a 2-3 minute read, lot of promoting their alarm clock (whatever) has a decent book recommendation and occasionally a link to a short article.
shutting down means you cannot receive phone calls.
This is actually one of the major problems linux phone operating system hackers have been trying to solve — idle power consumption is really really important for a cell phone, but most hardware open enough to run custom operating systems on don’t have easy ways to suspend to a reliable low-power state.
I have one, played snake on it, liked it (hurray for real buttons). Haven't bothered set it up as my daily driver because 90% of my phone usage is Signal and so far as I know that's not yet working.
If somebody has the bandwidth to hack on it in a way that will move the project forward, I'll send them mine for free. I love the idea but it's just not my kind of tinkering.
#7 - so you're suggesting it only work in a single country?
I think that to be practical, you're going to outsource the cellular comms to a prebuilt module that's got the carrier certification already sorted out.
Just package virtualbox in a midrange android phone with an OS image that does what you want. Maybe some port forwarding to make phone calls and stuff.
I expected this to at least mention the pine phone, the librem, or the Ubuntu phone (which was charming). It’s just about de-googled android and how it doesn’t work very well due to how everything depends on google play services
When I wrote it, from the options you mentioned, only Ubuntu phone existed and it was after Canonical abandoned it. So in 2019 it really felt like Android was the only option.
> ...de-googled android and how it doesn’t work very well due to how everything depends on google play services
Even that aspect is wildly overblown in my experience of using Google-free Android devices for the last 12 years. Yes, the thing will prompt you to install some Google-thing because it supposedly does not work without it - there are currently two such notifications on one of my devices telling me so. The thing is, nearly everything I throw at the device works even though some things complain about the supposed lack of Google services. I click away the warning if it appears and just use the software, this includes e.g. BankID - a Swedish electronic ID provider - and similar apps. I do not have microG installed on my main device either, I just forego on using Google services altogether.
All in all I think the current situation with de-Googled Android is quite useable. If ever some mobile thing comes along which more resembles the way software is installed on personal computers - i.e. install Linux or BSD of choice, add required software, configure accounts and you're in the game - I might move to that but until such a time I'll keep on using AOSP-derived Android distributions in combination with self-hosted services.
> All in all I think the current situation with de-Googled Android is quite useable.
It is, mostly. So far, I've still been able to reject having a Google account on new phones. Then I delete the "first time" app Google sticks on there, disable Google services, install F-Droid, and use only F-Droid apps.
Does that still work? It's been three years since I started a new phone?
Yes, that seems to work still, I recently configured my wife's new Galaxy S21 FE and had no problems telling Google and Samsung to take their accounts to where the sun don't shine.
On a similar note I've been wanting to buy and use a fully "free" phone - one where I have complete control over most if not all things. But it appears to me that unless you use a regular "phone", there is no way you can DIY your way into building a "modern" phone bit by bit like how a desktop could be made.*
Calling/4g/5g require specific hardware that seems difficult to integrate at a hobbyist level (unlike say installing Linux which is possible in an afternoon even for some not familiar with it.) And even though you could get android to run on a raspberry pi, it appears impossible to use something like google pay (which likely requires safetynet) on it.
I'm not sure if the reason is the "blobs" that ARM uses instead of drivers but it is disappointing that I can't take something like the Snapdragon 8 Gen 2 dev kit, dualboot linux and android and use it like a phone-laptop hybrid. /Disappointed rant over
* - technically even laptops like with an SBC or the framework board. I can make the hardware but the software challenges seem huge. If anyone familiar with ARM development has any info/suggestions, I'd love to know more.
Baseband and cryptographic shenanigans are a pox to FLOSS/H.
Particularly on the hardware front, since there is a vested interest/requirement for operating Telco's to maintain backdoor functionality/capabilities for the local government. It's just like payment processing.
Everyone having their own stack is such a great sounding idea, but the Establishment/System is tuned such that it can never happen, and in the event it does, it's guaranteed to be dismantled because every ne'er- do-well will start exploiting it, and the vast majority of people are not interested in protecting themselves... They want someone else to do it for them.
It's one of those nice things we just can't seem to get our shit straight enough as a society to have.
The base-band issue is the Achilles heel of the whole idea. Depending on which government you are under, even trying to implement the exact same functionality via open software is seen as tampering with a radio device and can get you in a lot of trouble. With that one move, they have effectively killed any possibility of a completely open system ever happening.
Are you saying you can build a modern desktop without any binary blobs? Full control of all firmware (including the CPU microcode and things like Intel ME)? What about WiFi cards? I'd say that hasn't been true for at least 20 years.
> Are you saying you can build a modern desktop without any binary blobs? Full control of all firmware (including the CPU microcode and things like Intel ME)? What about WiFi cards? I'd say that hasn't been true for at least 20 years.
Not exactly.
What I mean is the average x64 Intel/AMD device is relatively much more "controllable" than an equivalent smartphone, assuming decent BIOS/UEFI options are provided. If I want to customize the looks with rainmeter, or even system-level tweaks (eg mactype), it's relatively trivial on desktop OSes (possibly excluding recent locked-down MacOs versions). I can even jump between Windows and Linux and Hackintosh/MacOs and BSD as I have a "pretty decent" level of control.
A rooted android phone is also quite decent, however issues start creeping in the further you stray from what your manufacturer provided. It starts of with things like Widevine/DRM not playing nice (unless you have a Titan M chip or something similar), and the further you go (stock rom -> custom rom with gapps/microG -> deGoogled rom), the more painful it becomes to even use ""basics" like whatsapp.
I guess all this is a very roundabout way of saying that desktop OS is a first class citizen (mostly), while mobile OSes (which are necessary for booking an Uber/paying at the checkout POS/navigating with Google Maps with turn-by-turn nav) do not have other freedoms. And while the hardware absolutely exists for an arm board to do it all with the right daughterboards, doing mobile stuff on a desktop is impossible, and doing desktop stuff on mobile is limiting thanks to software.
I strongly depends on what you mean by 'control' and 'made'. In a sense it all depends on the depth of control and making your own; nearly all modern silicon requires proprietary software to initialise it to a usable state, and almost all hardware can't really run on its own and needs supporting devices which can cause an otherwise open design to end up closed anyway (even if it is something tangentially adjacent like a required PMIC).
Some CPUs and SoC can be operated to their fullest capabilities with pure open hardware and software designs where you have all the control in the world (if we draw the line at silicon diffusion and packaging), but those usually lack one thing or another making them less feasible for common modern use. Often these are things like memory controllers and I/O where internal SoC peripherals need special initialisation before they are usable. Even in cases where the driver side of things is reverse-engineered a bunch of Real-Time OS code needs to be loaded at boot time to get things like GPUs, Basebands and storage to work at all. So far, there are practically no cases where those have been reverse-engineered and replaced.
For a lot of internet-commenting-people, 'open' means something specific to them, like changing some design aspects like colours and fonts, or installing cracked software. Those are mostly just end-user features and have little to do with openness. This has a sad side-effect making searching for 'open' information get littered with what essentially boils down to appearance and recompositions of existing software.
Perhaps we need 'levels' of openness or other words to describe what we mean. The Snapdragon example is a good one in that it essentially uses an openly available and targetable ISA, but closed system design which still makes it impossible to use as-is (thus needing the blobs you mentioned). On one hand it's all legal and NDA stuff, on the other hand it's also laziness and cheaping out on the long-term usability. If a company were to spend just a little of their marketing budget on finding out the parts of the hardware that doesn't need all that NDA coverage, making open software to run the hardware becomes much easier.
Good points, ideally I'd say the software should at least be non-proprietary or at least source-available, but in reality I would settle for anything where I can control at least the software (OS level) side of things.
Since 2019 the landscape has deffinitely changed. "Hardcore" Free Software phones like Pinephone, Librem or PostmarketOS are still difficult but if LineageOS is Free-Software-enough for you, don't give up now. There is grapheneOS that should be no-compromises solution with it's sandboxed google apps. (Haven't used it myself.) MicroG now passes safetynet once again and is pretty stable. (Altrough on some phones you may need root to make it believe that the phone is not rooted.) Clean LineageOS without gapps is more usable than ever due to the huawei-google breakup resulting in a lot more apps being published on huawei appgallery without google safetynet requirement. (Unlike their playstore version.) Downloading apk from Appgallery website requires just some url manipulation for which you can use following bookmarklet:
Source: me. I daily-drive 7 year old phone running LineageOS with almost no issues. I also have a backup phone where I multi-boot LineageOS+MicroG and postmarketOS for fun and experimentation.
I think there's a good chance that the huawei version of the app will work even without huawei services except for push notifications. SDM845 is now probably the best chipset for postmarketOS so I hope you'll be able to play with it.
>But pmOS really suffers from reliability issues – both on edge and on stable it seemed like every update broke some things and fixed others, [...] such as the media controls on my bluetooth headphones breaking [...] major showstoppers such as broken phone calls, SMS, [...] all of my icons disappearing from the UI
I would be surprised if these issues were because of pmOS and not because of Alpine. For example everything involved with bluetooth headphones or with calls or with SMS is in Alpine, with the exception of one ALSA UCM config file from the pmOS device-pine64-pinephone package that hasn't changed since 2021. In the ~2 years that I've had my PP (running pmOS Edge), the only time phone calls were broken were because of pipewire updates in Alpine repo. SMS has never broken. I don't use bluetooth headphones so I can't say anything about that.
Alpine is maintained somewhat "amateurishly" compared to other distros. Eg one of the main Alpine devs frequently makes commits to the packages repo without any sort of merge request or review process, which sometimes introduces new bugs and they then make more commits to fix. I'm not saying this is a bad thing, quite the opposite; it's just what the dev culture is, and it means fixes and changes go in quickly without bureaucracy etc. (And I'm pretty sure ddevault is aware of this.) Also while Alpine does involve pmOS folks in discussions about issues that might affect them (and many of the pmOS folks are also Alpine folks), things do slip through the crack many times, like the pipewire issue I mentioned and many others.
At the same time I expect these things to happen on edge but not on stable, so I'm surprised that he says he experienced them on stable too. I only use Edge myself so I can't say for certain.
In any case, to the overall point of having a phone that is 24/7/365 reliable at calls and SMS, I agree that "PinePhone running pmOS Edge" will not meet that requirement, for both software and hardware reasons.
The hardware and software stacks of all of the components of a smartphone are at the pinnacle of modern engineering, requiring an uncountable number of engineers and consumers to maintain a mind-boggling supply chain.
I'm lucky (?) to live somewhere where 2G isn't going to be turned off any time soon, due to legacy support requirements. Otherwise, I'd have to get one of Nokia's newer "dumb smartphone" devices.
I would have kept my previous phones for a lot longer if it wasn't for them switching of the 2G network, and next year they switch of the 3G network! (Australia for reference)
These 4G dumb phones are cool but the power requirements of 4G can be a burden once you actually start making calls. So you get weird things like 30 days stand by but 3 hours talk time
> a constant notification just so it will be able to send me push notifications
That's because stuff running in the background in a way that cannot be optimized by the system must inform the user it's doing background operations.
I think that's a good idea. Hiding the notification is easy (hold it to enter the notification settings and disable that specific notification category) and it clearly shows which apps are killing your battery life, hopefully making the entire platform less dependent on aggressive task killers.
I have been on a AOSP+MicroG setup with no google apps, services or accounts for almost a decade and can safely and confidently say the author has no idea what they're talking about. I don't know what sorts of applications they need their phone for but I have had almost no problems. I use my phone for everything and I can't think of a problem off the top of my head that I have that Google locked in users dont. In fact, when I first took the dive, I thought it was going to be rougher than it was but worth it, and it has been easier. Ive probably stumbled into 1 or 2 apps over the years that I couldn't use, and I can't remember what they were off the top of my head.
It is not only possible to live google free, it is fucking easy.
No. I've read that Google is not allowed put iMessage-like features into the stock messaging app because it would run afoul of antitrust. Ironically, Apple can do this because they own everything.
This is silly and false. Google's default android Messaging app in the US on all carriers is Google Messages and it uses RCS to do exactly what you are saying they can't do.
What does that have to do with anything, though? Google Messages is available to any Android OEM that passes certification, assuming they want it. Certainly Messages wouldn't be a part of AOSP, as Messages isn't open source.
Google Messages is the stock messaging app, and contains iMessage-like features. AOSP is mostly irrelevant. It's not clear why iMessage-like features would be an antitrust issue for Google anyway.
Alright. Yes. now it is. This is all I've been trying to point out- is that there is a built in messaging app as part of AOSP. My recollection is that Google can't turn this into gMessage because then it would be distributed as part of the base OS and would require a Google account to use.
Now it's abandoned so it definitely won't be anything now.
Zero phones in the US ship with AOSP. 100% ship with Google Play Services. AOSP, in the US, is dead as far as any reality is concerned. China is different but Google wouldn't want to even be there.
It sounds like Google's way around this is what they've been doing for years at this point: slowly replacing AOSP builtins like the stock messaging app with separate, proprietary Google apps.
Right, I don't believe that's part of AOSP but is a Google app. I don't believe all phones ship with that though my Pixel does. Do Samsung phones ship with that?
Yup. What's open source now is basically a framework on which you have to build anything useful and, surprise surprise, the only company building that useful is Google.
I've leaned harder into stuff like degoogled Android since app providers started to get obsessive about unrooted phones.
I'm sorry, but no, you don't need to halt execution simply because I'm an admin on my own device. Particularly if I can do exactly the same thing by just using your web interface (90% of the time this is available).
I hoped Amazon's non-Play devices would be enough to keep most developers from being fully captured by Google. Unfortunately they haven't been popular enough in many cases.
[meta-commentary] HN should require a date field on all submissions and then should only show the date if the article is older than a specified time (like 1yr) (or always show the date but not in the title)
I tried CalyxOS before GrapheneOS, a few years ago... I can't speak to CalyxOS now, but at the time, it seemed to be mostly just a hodgepodge of "this is a 'privacy' app; throw it in" without much coherence or polish beyond that.
Then I tried GrapheneOS, which was driven by Daniel Micay, who seems a hardcore security-minded person. GrapheneOS seemed to be working beating AOSP internals into better security shape, and very slowly tackling the apps. So, with GrapheneOS, it seemed I was starting with as clean and trustworthy an Android as one can get (to the extend Android can ever be clean or trustworthy), and in the interim I could add missing open source apps to my install. (Email, maps, SSH, chat, etc.)
GrapheneOS isn't the ideal I could imagine, but it exists, and I feel less violated by GrapheneOS than by other phone in the last ~15 years or so.
I have been running CalyxOS on a Pixel 6a for 6 months and I'm very happy with it. It does everything I need and found no big annoyances so far. A minor issue is MyFitnessPal not reading barcodes but that's it.
This answers a question that's been bugging me: Why TF can I not find a Calendar app on the Play Store that does not depend on Google Calendar? I used to have one 5 years ago that was completely separate, but now I'm apparently not allowed to have a calendar app that does not send all my calendar info to Google.
So now I have this very nice Android phone that does not have a Calendar app because Google won't allow such a thing to exist unless they get to see my appointments. Guess I'll have to write some kind of web app calendar thing if I want a freaking calendar on my phone.
On F-Droid, there is Etar, a calendar app, and Offline Calendar, which is pretty much a script that creates local calendars not tied to Google or any other service. Worked pretty well for me a few years ago on LineageOS+microG.
I know it is just trading one mega-corp for another but the Outlook app doesn't use Google calendar. I don't think the Samsung calendar uses google either unless you hook it up to gmail, I had mine against an exchange server.
That explains it. My previous phone was a Samsung and it came with an app called "Calendar" that's not present on my current [non-Samsung] phone. Wonder if I could copy it over with ADB? Guessing it wouldn't work on a non-Samsung phone though.
All the simple tools apps are good and the calendar is quite nice these days. You can sync with google calendar with it but it can be local only if you like. I sync to my Nextcloud server with Davx5 https://play.google.com/store/apps/dev?id=907029638802258926...
Also helps they are open source and can be installed via F-droid if needed
Doesn't LineageOS come with such an app? Even if you don't want to use LineageOS, you might be able to just extract that .apk from it and use it on stock Android.
Giving up on free software phones is the first step. Now give up on software phones and just use a tiny non-computer cell phone that does text and voice. The removable battery will last a week and you won't even notice it in your pocket. The firmware will last the entire life of the phone, never needing to be updated.
Replace the car GPS navigation functionality with a dedicated car GPS and you have most of the same functionality but with better longer lasting devices.
I don’t even want a “phone.” The number of actual phone calls I make is minuscule compared to internet usage, texting, videoconferencing, navigating, etc.
And I guess also carry a camera. And a music player. And a gaming device. And maybe a flashlight too. And you probably want a voice recorder. And a pedometer. And maybe a pen and paper if you need to jot a quick note. And don’t forget your datebook with your calendar. And bring a calculator. You could also pick up one of those portable DVD players from the 90’s if you want to watch a movie on the go. I happen to enjoy a couple of sky gazing apps, so I would probably also take one of those old sky charts with me.
you also "miss" being sent to a known slow and long rout just because the buggy algorithm wants to bucket test some random path close to the daily rush hour.
Yes, accidents always happen in the same spot at the same time on the same days, and new people visiting new places always know about those things.
I have had Google try to send me on slower routes, but I use my brain and drive the more logical way and it updates the map accordingly. Though most people are not logical, and such practices theoretically help everyone overall including the petson who is sent on a different route that time.
I use lineage on oneplus phones and have no problems. I can't do internet banking and a bunch of other non-free things, but really, who needs that? I have a web browser. The only thing that doesn't work is google meet and google docs and google drive because they intentionally break the mobile experience in the web. But there is a simple solution. Simply don't use google services. If/when my employer demands that of me (occasionally I hear murmurings) I just tell him I'm online 9-5 on my laptop and if he needs me he can call. I'm happy with my lineage.
> I can't do internet banking and a bunch of other non-free things, but really, who needs that?
I do, and so do a majority of smartphone users. Good for you that you are fine doing without it, but most people will not accept that trade off.
I ran CyanogenMod back when that was a thing, but at that time SafetyNet didn't exist, and it was a rare event when an app that ran on Google-approved Android didn't run on CM. Nowadays I can't run an alternative ROM without sacrificing things I care about using.
Is it specific to your bank? I have used a couple of the big name American bank apps on my phone and never seen a dialog whatsoever (the blue and the red bank just FYI). Or it could be specific to your setup. Are you using Microg and hiding root (if you have it)? Are you making sure that all of the microg checks pass? Does the microg signature spoof pass?
Perhaps as a last resort, after getting your install set up, close the bootloader again. I don't do that but I've heard of people having trouble from leaving the bootloader unlocked.
I can't believe there's no mention here of sailfishos.org, which runs great on Sony xperia devices and even runs android apps. It's Linux (Intel/Nokia Meego, specifically) with a proprietary UI.
Would I prefer an OSS UI? Yes. But for a real, viable duopoly alternative it is worth the sacrifice to me.
I use CalyxOS on a Pixel 6a and a LineageOS on a OnePlus 6. No micro G on any of them.
Honestly, I have absolutely no issues with CalyxOS and I use mainstream apps like: UK Banking (Starling and HSBC), Whatsapp, Telegram, MyFitnessPal, some airlines and local bus apps...
I guess your mileage might vary. If you're keen on a particular app and you can't make it work, that's a frustration. So far I haven't found any reason to come back to a standard OS, and I appreciate my phone not sending constant signals to Apple or Google to the tune of hundreds per day.
I use calyxos as well and have microg disabled. WhatsApp works fine, signal, telgram works great, fairemail client for emails, davx5 for contact syncing etc. I don't use any google apps expects camera app (which I have disabled data access using firewall), but I have a feeling those may not work so well on calyxos.
Only problem - which I actually think is a feature is that I don't get any notifications for a lot of apps, unless I actually open the app. I really think this is a great thing because most of these apps that use google services for notifications are just sending marketing modifications which I would have blocked anyway. I do sometimes miss WhatsApp call because the notification doesn't work properly but I have told all my friends and family that they can contact me on signal for fastest response.
Is it possible to uninstall all browsers (including Chrome) in Calyxos? Most of the time I waste on my phone is in the browser and I would love to have a browserless but otherwise connected device.
I'm willing to compromise a lot in order to go with the free software approach, but my god, just do better. There's no excuse for how un-functional this thing is. Try harder.
I had the same reaction with the pine phone a couple years ago. At least that one shipped sooner though.