The UI concepts of Sailfish OS were great, the gestures were simple and intuitive. However, it lacked so many other features. The Android emulator was only compatible with apps for Android 4.4. Although being marketed as secure and privacy-aware, users have no control over apps, as there is no mandatory access control or any other permission system. Remember, this OS has been in development for over five years by now. Also, many parts of the UI stayed closed source, although Jolla promised from the beginning they want to release more open source components. If they wanted contributors, they would have more likely contributed to user-visible stuff like the mail app — not to some lower level middle-ware that is open source.
All of this and some more reasons caused me to finally pull the plug and switch to Android. I wrote about it in lengthy details here:
If anything is closed, except drivers they'd likely be forced to accept binaries for, I wouldn't even consider the OS. Who do they think their target audience is? I imagine most people looking into an alternative mobile OS are well aware of the slippery cliff that is proprietary software. It's the whole point of looking around. I'm looking for full control so I don't get stuck with someone elses bad choices yet again.
Large corporate clients like the Russian postal service. While Jolla has always drawn on bug reports and some contributed code by hackers who are passionate about owning a cool, supposedly open-source phone, the company has never seen that audience as big enough to generate profit. Jolla has been striving for years to win adoption from sectors that don’t really care if the code is completely open or not.
Only partially, the ASOP does not run on anything because of the vast amount of blobs and basic drivers missing and people are maintaining open-source reimplementation of the gapps to make even the most basic apps working.
Yours is an argument why all struggling OSS mobile OSes are struggling, not particularly why AOSP has problems. Surely if even AOSP is struggling to solve these issues, what chance do the others have?
Not to the same degree, Linux still has some blobs to run on modern computers but nowhere near as much as Android. Android needs blobs for absolutely anything. And you don't need an equivalent of the Gapps on Linux, it would be like something as common as systemd would be proprietary.
> Yours is an argument why all struggling OSS mobile OSes are struggling, not particularly why AOSP has problems. Surely if even AOSP is struggling to solve these issues, what chance do the others have?
Google does not care any more about open-source on Android, the current open-source bits are just there because of legacy history, they close all the open-source parts step by step. The Gapps used to be open-source at the beginning. AOSP isn't struggling, it's just there's no open-source movement.
Those few proprietary apps for which there are no free alternatives (spotify, duolingo) seemed to work fine without the play services.
This is not to diminish what is being done, as I do appreciate the option of a wide array of OSS items decoupled from the Google eco-system, but it's nowhere near independence from Google's eco-system. I would hardly say that it's a replacement so much as a novelty at the moment, even though there are a few absolutely golden apps to be found there. (Free advertisement for PassAndroid, the Android equivalent of Passbook from iOS, which does exactly what it advertises)
The saddest thing here is those alternative OSs innovate a lot. WebOS in particular was years ahead of both iOS and Android in terms of UX (after all they had Matías Duarte on board... ). Sailfish is more interesting technically, in that it was probably the fastest OS to do "zero to non-trivial native mobile app" on thanks to its development tools (the Qt environment, with QML, a standard build VM...).
I backed the tablet as well, but I don't hold a grudge. What they are trying to do is incredibly hard, failures have to be expected.
If you want to meet the developers from Jolla, I think they have a BoF room at FOSDEM every year. I sat in to listen for a while in 2017, it was pretty interesting.
Back in the early '10s however it had things like usable widgets on the home screen, application cards, gestures instead of buttons... which all landed in the two major OSs years later.
The HP WebOS devices had wireless charging. It took many years for this to properly appear again in the market.
The webOS UI from 8 years ago is still way better than the latest from Android or iOS today. The apps and implementation may not have been that great from a performance perspective (even then), but the UI was/is absolutely better.
If you want a specific feature, I'd point to first-class multitasking and cards, real menus, good combination phone/tablet UI via Onyx (I'm looking at you Android), JustType, Synergy (all your services working together instead of being separate islands in your phone), gesture zone, etc.
USB tethering requires manual installation of a Sailfish package via CLI
works out of the box
works out of the box: A2DP, file & contact transfer, perhaps more I haven't tested yet. Works just fine old Nokia headphones, and with Roidmi via the Android app.
Only issue I had is some apps not working.. but none important.
I keep seeing report after report that Play Services in particular can be a real drain.
Play Services is choke full of features, so it tends to be very high in the battery useage list.
Replacing it with MiniG that uses the same API with way less features can only decrease battery consumption.
What is this? I tried searching for it, but got nothing useful. Only things related to crypto mining, and other unrelated nonsense.
It sounds interesting.
Soo on Android many non basic features are enabled through Play Services.
It is basically a huge library from Google that auto updates itself.
Chromecast is a good example of a feature delivered entirely through play services.
Google login also works that way.
Many apps leverage Play Services in some way. It is common to check your play services version at startup and ask the user to update them if they are out of date compared to what your app targets.
Micro-g reproduces the Play Services API and tries to deliver an open source equivalent to some of its features.
I don't really buy the argument that Play Services nullifies AOSP in any way but it is interesting to see another alternative (amazon also has a similar project) that is open source.
To get Firefox without being forced into a Google account use Yalp installer.
* Email - yes
* Browser - yes -- the official one browser is based on [an outdated] Gecko engine. Unofficial browsers based on QtWebkit.
* Maps - yes -- the official solution these days is HERE Maps (this requires Android support which is available on Jolla-supported devices). There are several other Map solutions available through OpenRepos (a popular 3rd party repository for Maemo/Meego/SailfishOS software).
While the app ecosystem can in no way compete with the number of apps available in Android or iOS -- if you use your mobile for web-browsing, music, podcasts, weather, timers, etc. it's sufficient.
Do they still provide offline map updates? I have a Nokia N9 and several other Symbian based Nokia phones, but there appear to be no updates to the offline map data.
> Sailfish Browser uses Sailfish Silica Qt components for the browser chrome and gecko engine with embedlite Qt5 binding.
That will satisfy most tinkerers, since they can modify anything they want and even load their own images since the drivers are open source.
Android as an OS is good enough. Just that every single manufacturer has opted to lock it down.
Recently I have been in contact with the topic of eMMC deaths  in Samsung S3 phones and I have to say I am very pissed of by the arrogance of large corporations like Samsung (I have no evidence, but I am confident that Samsung is not the only player doing shit out there):
They sell you a device with a firmware which will probably kill your memory chip within weeks. To fix it, they patch their firmware to let the memory chip freeze as soon as the firmware would kill it. So instead of the firmware killing the integrity of the memory, they risk the integrity of the filesystem and let the user experience freezes and reboots for the rest of the phones life (until the filesystem can't recover).
To sum it up:
- they sell you a fast breaking device
- they patch it to make the phones deaths less likely, but without fixing the real issue
and that all within software.
Sounds to me as if someone wants to make sure the market will not be saturated. I am no lawyer but to me this sounds illegal, but as it is kinda hard to find out that shit (the S3 was over 5 years old when the linked talk was given), it is very unlikely anybody will engage in a lawsuit.
Currently, I am under the impression that the only solution to that kind of a problem would be, to force all manufacturers to open source all their software which is required to use their devices (firmware, drivers, etc. license doesn't even matter). It would be one hell of a change for device manufacturers, but as we can see, otherwise we have no idea what kind of hidden 'features' the devices we buy (our devices?) have.
2. Rooting presents another extremely stupid problem. Apps like Netflix and snapchat and many payments/banking apps will refuse to work. I cannot find a way to explain this except dev incompetency and/or hostility towards users.
3. On Samsung phones there is a Knox bit, which gets tripped as soon as you flash any image. On many models, people haven't been able to figure out how to reset it. This means that once you flash Lineage on your phone, you permanently lose features, even if you go back to the official OS! This is what prevents me from flashing my phone, since I really don't wanna lose Samsung pay and Secure folder etc.
- Netflix: "we see you have root, we can no longer trust our DRM on your phone".
- Snapchat: a service whose core principle was always, "you can send your nude photos to each other, the photos are ephemeral, and they won't be able to save them without you knowing". Rooting your phone breaks that last guarantee.
- Payment/banking: CYA. You have root, some evil "hacker" might "hack" your phone and steal your money, and we don't want to get sued for that.
It's very much user hostility, created by a conflict of interest between the company making the product, and the users of that product.
Yea. And if/when they go away, that UI will die with them, which is also a shame.
Pinning hopes on the purism now for a hackable mobile device.
It all sounds grim, and describing people you failed as “s bit noisy” isn’t great PR. It just sounds like people in way over their heads to me.
It seemed more like they ran out of money entirely, and their options were either give up completely and no-one gets anything, or keep working on Sailfish OS to hopefully get new business going, then make a profit in the future that they can use to pay tablet backers back. And they have been paying backers back as promised when they're able to.
The tablets themselves were manufactured by their Chinese partner and even sold on third party Chinese websites. Funders never got their tablets, but other customers were able to buy the same tablets for below what Jolla had sold them.
Jolla was accepting orders until the very end before they decided to come out and cancel the whole tablet effort.
They have since raised over 12M dollars but ostensibly that is not enough to repay the 2.5M they raised initially.
I have no insights into their engineering or technology since I never got to use their product, but as far as business practices go they have constantly scraped the bottom of the barrel.
If you have some spare cycles go check out the comments on the indiegogo page (https://www.indiegogo.com/projects/jolla-tablet-world-s-firs...)
EDIT - https://reviewjolla.blogspot.com/2017/06/financial-news-joll... claims they have raised 70M USD.
This article also fails to mention Tizen. Has it been abandoned? It was absolutely terrible (millions upon millions of lines of the worst C code I've ever perused) but I thought it still existed, making vaguely similar claims to open-sourcehood as SailfishOS.
The real mobile OS to watch and advocate for is postmarketOS, which attempts to get mainline Linux kernels working on phones.
After throwing away the C++ libraries and going to that C code you mention, they have decided to offer .NET Core + Xamarin as alternative to those C++ libraries.
Which I agree is much better than Bada C++ style libraries, but reveals they aren't sure where to go next.
Then at FOSDEM they introduced an IoT OS based on NuttX + JerryScript, and called it Tizen RT, even though it isn't anything Tizen related.
I think based on your argument that Redhat and Ubuntu are reasonable reasonable counterexamples.
Blame hardware manufacturers for this. If they released open specs (not even the drivers sources, just the specs would suffice) of their devices we could have full alternative OS installations in months after each device release, giving new life to old devices and much higher security to newer ones.
Forced obsolescence, pervasive spyware/security problems, being locked into an "app" ecosystem, user-hostile antifeatures, lack of modifiability to deal with hardware degradation over time, etc.
Back in the 90s we suffered from Wintel dominance where if you installed Linux on a PC, it was likely to be a less than pleasant experience due to missing drivers, etc. The modern day equivalent to Wintel is Armdroid. Despite Android using the Linux kernel, you can't just take an off-the-shelf Android phone and run a Linux distro of your choice. ASOP (Android Open Source Project) forks the Linux kernel once and then virtually all Android OEMs create a second fork of the Android kernel fork, apply their own changes and barely anything ever lands in mainline. Jolla's strategy to run Sailfish on top of Android kernel forks with Android blobs inherits the same problems of Android, encourages bad practices of Android kernel development, planned device obsolescence and the throw-away culture when the kernel fork reaches EOL.
There are further problems with Sailfish. It is built on top of the Mer, a Linux distribution built for Tivoisation. Mer refuses to use GPLv3 licensed packages. Mer packages are stalled on the last versions that used GPLv2 before they were relicensed to GPLv3. Most of these package remain unsupported and unmaintained for years and are probably vulnerable to many known exploits.
On top of these problems, Sailfish contains many closed components and the majority of Jolla is owned by Russian investors. Is this an OS that can really be trusted?
Fortunately the situation is starting to change and there are real alternatives which may be usable in the near future. There's PureOS which aims to build a Debian-based OS for the Librem 5 phone on top of mainline Linux, postmarketOS supports mainline Linux on a few devices in addition to Android kernel support, and now there's Maemo Leste which is built from the ground up to run on mainline Linux. These projects are still in their infancy but at least there is hope for the future. Sailfish, unfortunately, does not look like it will be part of that future.
When you buy a Jolla device, you are taken through a tutorial. The gestures aren’t that difficult to get used to. To me personally, they seem a lot more straightforward and intuitive than my wife’s Android phone.
Incidentally, this winter I traveled in a couple of developing countries and bought local SIM cards from an office of the main telecom concerns of these countries. Both times when I said I wanted to use 3G data, the clerks grabbed my Jolla from me and tried to get to where they thought the settings menu was by repeatedly mashing where the Android back button would be. When I said it wasn’t an Android phone, it was like they simply couldn’t absorb that information, they just started mashing harder. That really underscored for me just how much of a mobile monoculture the world is. (The clerks probably could have found their way around an iPhone, too, but I didn’t see many of them in this region.)
Quirky isn't necessarily better if you're trying to woo users who just want a familiar experience without learning new paradigms.
The lack of a back button in Firefox OS annoyed me! Instead of pressing a section at the bottom right of the screen with one's thumb, one had to 'click' on a web toolbar button all the way near the top left of the screen.
It's not that "gestures aren’t that difficult to get used to", it is more do they offer anything more than the tried and true?
In all honesty I think it is irresponsible of these companies to take huge kickstarter payouts to develop hardware or operating systems that are doomed before they have begun.
I think that Google or Samsung cannot make a good android tablet shows how hard this is.
They've become preferred suppliers of the Russian and Chinese governments.
While I don’t personally like the policies of either government, if they throw money at FOSS, that’s still (probably, see Android) a good thing.
Sailfish isn’t completely FOSS and the strategy of licensing it to clients like Rostelecom foresees that a lot of the OS will remain closed source.
Besides, no government will actually check the whole OS themselves, so they have to trust some company at some point.
How so? Jolla’s licensing system allows its clients to bundle just as much closed source, phone home software on their phones as any Android system. They can lock the bootloader, etc.
* The Mail app does not support GPG
* The Chat app does not support any modern XMPP extensions or even group chats (this is because Telepathy does not support it)
* The Calendar has no month view if I remember correctly
But I really wish Sailfish was usable for me, I would prefer to have a "real" Linux distro on my phone.
MeeGo, Mer, Maemo, Moblin, Harmattan, LiMo, SLP, Bada, Leste, Tizen, Nemo...
It they spent more time on making a finished product rather than coming up with yet another stupid new name, logo and website they might have more traction.