GTK for a smartphone is an interesting choice. I'm curious if the Gnome team would ever commit to this.
The use of a stylus in the video is a bit concerning though.
Off topic but I hope Librem releases a high-end laptop in addition to their mid-tier one, ideally with a high performance AMD CPU. Something that could be the open-source friendly version of a Macbook Pro.
The way they disable Intel ME and add hardware toggles for webcams/microphones, and their general good taste is a big draw for me.
> The use of a stylus in the video is a bit concerning though.
I imagine that's just so there isn't a big finger/hand getting in the way of the camera. And because the phone is flat on the table for the video, you can't type on it like when you're holding it with your hand.
A non-issue in my opinion. I wouldn't have given it a second thought.
I want to like Librem's smartphone, I really do. I like having control over my software, and I like being able to really understand how it works and tinker with it and stuff. But I have this sinking feeling that the Librem 5 is just vaporware, or worse, that it'll get released with a half baked UI and a barebones list of apps. On that latter one, I really don't see how they'll fix the problem. On the former, the thing is that the affordences for bad UI design are a LOT tighter on mobile devices and it's taken a long time for desktop Linux to get their game together. Plus, all of Librem's other stuff is over priced and a bit cheap looking, and I'm not sure I want a phone like that.
It's Linux -- if it's missing something, I can just build it. And I don't have to write my code in something horrible like Java, or boot up an emulator to test things, or submit my code to an app store, or figure out how to root it so I can remove all the bundled spyware. If I write a command line utility, I don't need to have a separate codebase between my computer and my phone. It's what I have always wanted my smartphone to be -- just a computer that I completely control, with access to a phone network.
It might be terrible. I'm kind of on the fence about it. The hardware needs to be reliable, the drivers need to be reliable, it needs to be compatible with common carriers. There a few other things. But the software on this device just needs to be passable. I don't need it to be particularly good.
When thinking about the Librem phone, mentally move it out of the iPhone category and into the Raspberry Pi category. Even if it comes out and gets terrible reviews and the software is all half-baked, even if it can't replace my normal phone for most things, I still might be tempted to buy one even just as a secondary device.
Besides the hardware kill switches and replaceable battery, I think this experience is going to compare poorly to LineageOS, which has a lot of the same software freedom and privacy advantages with a much more refined and complete operating system.
This phone is months from shipping and we haven’t even been shown multi touch have we? Is that even a feature being promised?
Ultimately the difference between this and the Raspberry Pin is tangible - $614 to be exact. I know that Purism isn’t out to sell a mainstream phone, but my doubts are seriously strong that they’ll sell enough of these to justify the effort.
I can build anything that’s missing in Linux as long as someone invents the fountain of youth so that I can spent a few thousand engineering years on implementation.
I look forward to leaving LineageOS behind. First of all, Lineage builds for particular phones are abandoned frequently, because they are made by a single maintainer as a hobby and he can get bored and leave. The fact that the Android development environment is so arcane, and the hardware usually closed in various ways, makes it hard for anyone else to pick the project up.
Secondly, there are a lot of things that one can do in an ordinary desktop-Linux environment that are much more complicated in the Android world. On my old Nokia N900, I frequently wrote Python scripts that interacted with the system D-bus to automate things, and I did not have to install a big development environment or download someone else’s solution.
I use LineageOS right now. It's great, and I'm glad it exists, but it still requires me to write the majority of my apps in Java and to test them in an emulator, and it still requires me to use the Android OS for everything.
I know I can theoretically compile my own ROM or something to customize the desktop environment, but... bleh. It's time-consuming for me to build things on Linux, but it is more time consuming for me to build things on Android. Android development is just a pain all around.
This on the other hand, isn't just a >$600 phone. It's a >$600 grimoire. Expensive, and cumbersome, but nearly all magic is. There's an almost subconscious brand (for lack of a better word) that devices of this nature fit into. They're trading user-experience for dev-experience, which is not something LineageOS or Android is optimized for.
As to the rest of your criticism, all of your other points seem completely fair. I'm not sure what I would put the odds at of the device succeeding (note, succeeding means finding a stable niche, not beating any other device). And while I can forgive a lot of flaws in software, for the price they're asking the hardware/drivers need to work well.
I'm very much in the "wait and see" camp. But I do see a lot of potential.
There is Anbox for running Android applications under Linux, and I've heard some people claim it's "less clunky" then an AVD. I haven't yet tried it myself but will soon.
I don't disagree that there are some suboptimal things about Android but to bootstrap a new mobile OS platform from almost nothing, and being on-par not even matching Android especially with regards to security (sandboxing, a proper permission model) seems like it would require a dedicated team several years of development time, which Purism doesn't have.
A lot of Nokia N900 users were very happy with a “barebones list of apps” and the only reason that they had to eventually abandon that phone was because the closed hardware required a specific kernel version and certain other libraries could not be upgraded for interoperability and security’s sake.
If the Librem offers a more upgradable device, then that is already enough for the hacker community. Most of my time is either spent in a web browser, Emacs, or sending SMS. Also, many Android users who eschew Google apps and install LineageOS with F-droid as their app source, would probably find Librem a comfortable enviroment.
I totally understand that these phones are very much intended for highly tech literate people with specific priorities, but Librem will not see success selling a $650 phone that can essentially only barely use a texting app, a web browser that already exists, and stumble its way through Gnome like these videos are doing.
I also need to challenge you to look up a video of LineageOS and compare it to these videos. LineageOS would be like going forward 20 years into the future compared to this. This is a very rough tech demo at best, and this thing is supposed to ship this year.
LineageOS is giving you 95% of the privacy and freedom benefits while giving you a nearly infinitely better software library and UI along with it.
As a $100 tech toy phone I can see success but not $650. That’s iPhone money.
Locked bootloaders and proprietary components is what holds us back with alternative projects. In that sense Purism is giving us gold here.
I’m nervous though because hardware projects can fail so easily. Shipping and iterating on hardware isn’t the same as shipping software. In my opinion it just has to feel a little more “done” than this.
I'm also stoked about the PinePhone. The PinePhone goals are a little more modest, and so is the price. If the Librem 5 fails to be interesting, at least it'll have a competitor at a much lower price. I'm holding off preordering either, but I may order one or the other next year once the bugs and whatnot have been ironed out.
I'm stoked to have a phone that I can tinker with like I do my Linux desktop.
I have been following this for a while and am likely to buy it if/when it is released. Even though that will require me to change to a much more expensive cell carrier.
The software seems good. After all, it is just GNU/Linux/GNOME - all you need to do is get a C compiler and hardware drivers and then it just works. An older video did show major stutter during scrolling but that can probably be mitigated.
What's concerning is the hardware side of things. I don't think many people expected it to actually ship on the original planned release date, but the delay has been getting quite large. All the stuff on their blog is about software running on the dev kit; we have no idea what progress is like on the hardware. Turning it from that bare PCB into a phone is a lot of work.
Their marketing guy also released a video a month or two ago comparing the Librem boot time to that of an old Android phone, complete with OEM crapware. If they have nothing better to do than make such stupid comparisons, then it is hard to believe it isn't vaporware.
I highly recommend looking into an MVNO (mobile virtual network operator).
There are inexpensive pre-paid plans that use the major cell network towers, and you get the same speeds. I recommend Straight Talk; their plans are very cheap for the data you get, but there might be other cheaper services.
I agree it's taking them a while, but I've never built a phone before, so I am cutting them slack.
I find it easy to believe that it is not vaporware. They have videos of working devices, something that can be replicated in mass, not renders etc. They've shipped dev boards.
I've been following them for only a few months, sure, but I've seen that link and others, and watched many of their videos. It's extremely concerning that they use a stylus, they clearly don't have multitouch support, copy/paste is based on a control key, screen tearing is significant and some of the UIs aren't scaled properly, I haven't seen them do text selection and the keyboard didn't give feedback. That's just a few issues.
Also, the videos are legit for the software, but they are all running on the Dev Kit, not the actual phone hardware, which we haven't seen and have no idea what the progress on is like. So the hardware kinda does look like vaporware. In the link you have, for instance, the best they can do is take a fake circuit board image and fade it into a fake phone image. That's not encouraging.
> On that latter one, I really don't see how they'll fix the problem.
I don't understand what the problem is exactly. There already exist phones that have the latest apps. I don't think Librem is trying to be a market leader, just build a device that's more open, so to say. The demographic isn't Google or Apple users, it's those who wish more control. They're not going to fix the app gap problem because it's not a real problem, it's a meaningless metric for phone advertisements. These videos are to demonstrate that they are working on filling the mobile usability gaps as well as building a phone that can run Linux software, not just mobile UI software.
I agree with most of your post, but lacking apps is definitely a real problem.
As someone who preordered the Librem 5, I know I'm not going to get an operating system with the polish of iOS or Android. I'm willing to live with that; the phone is attractive for other reasons, both practical (easy to hack) and ideological (freedom!). Heck, even if the final product turns out so poorly that I end up never using it at all, I won't be too upset: even if my $600 didn't end up benefiting me personally, at least it went toward a serious attempt at achieving a worthy cause (an open phone).
Still, I would like to try actually using the phone as a daily driver, and for that to be viable, I'll need to be able to access the services I depend on. Luckily, I don't use that many services, and many of the ones I do use mostly have web versions; as much as I hate the experience of using webapps, I should be able to get by with them. But I can think of several that I'll have trouble accessing at all:
- iMessage (I guess I could try to set up the Matrix bridge.)
- Slack? Discord? (Both have desktop web versions, but not sure if they have usable mobile web versions. That said, I have bitlbee set up to connect to Discord already and I know there are bridges for Slack.)
I'm lucky that none of these seem to be hard blockers for me... but other people may not be so lucky.
And, well, if things work but are a poor experience, at some point I'll probably get fed up with it and switch back to my iPhone. Some people are more ideologically committed than I, but many are less. Though I'm not counting on it, I would love if Librem's OS gained momentum and became a sustainable ongoing project, rather than dying off as a prototype like so many prior attempts (webOS, Firefox OS, Ubuntu Touch, etc.). For that to happen, it will need users.
For those that haven't seen it, here's a bullet list of the Librem 5 differentiators:
+ Does not use Android or iOS. The Librem 5 comes with the mobile version of our FSF-endorsed operating system PureOS by default, and is expected to be able to run most GNU+Linux distributions.
+ CPU separate from baseband, isolating the blackbox that the modem may represent and allowing us to seek hardware certification of the main board by the Free Software Foundation.
+ Hardware Kill Switches for camera, microphone, WiFi/Bluetooth, and baseband.
+ End-to-end encrypted decentralized communications via Matrix over the Internet.
+ We also intend the Librem 5 to integrate with the Librem Key security token in the future.
Just shockingly ugly. If you're going to do a pixels-under-glass UI, you really have to start with a toolkit that isn't awful, or you end up with the openmoko which looked exactly like these demos but 12 years ago. You end up with, in short, a phone where the user adjusts the volume using alsamixer in an xterm because the UI wasn't really thought through.
I'd say its better than most modern phones with their stupid flat UIs. I use KDE on the desktop for its featured richness, but gnome is still much more 3d and I envy its appearance. Buttons very clearly look like buttons and text boxes look like text boxes and the whole thing just looks 'nice.'
I’m sure people will disagree with this but I think they should focus on getting an easy way for people to target their devices with React Native, NativeScript, PWAs and every other cross-platform framework.
That way they can much more easily acquire software. Seriously making an app from scratch is no small task, and often time isn’t worth it for such a small market.
Have you seen what development for this is like? You can slap together simple apps (the UI anyway) with a few dozen lines of C. It's extremely simple, lightweight and fast, everything react/PWA's/etc aren't. This is what excites me about librem the most, it's so easy to just write code with no bloated IDE, 15 xml files, AbstractFactoryBuilders and the rest of the android crapfest.
PWA's and the like can stay away, I want software from developers that actually respect their users, software that doesn't waste resources, software that actually looks native.
Yes, FirefoxOS was certainly web centric and mostly did not succeed. That doesn’t mean it failed /because/ it was web centric. Every mobile OS has struggled other than iOS and Android because getting adoption for a new OS is hard.
Some additional thoughts:
1) maybe shipping a new OS with hardware, catering to a privacy focused niche is key to succeed this time
2) maybe web technology wasn’t ready to have wide adoption for making cross-platform apps at the time FirefoxOS was launched, but is now
3) maybe PureOS will fail regardless of how apps are built. In my opinion, the best chance of succeeding will be from actually being useful. In order to be useful, it needs software. Unless, of course, you seriously believe that getting hackers excited by not using web technology will be a better solution.
Yea, I think PWAs are a great substitute for apps. I wrote a websocket chat that my girlfriend and I use. It works great on desktop and mobile, makes for a seamless transition from commute to work. I'm less hyped about the electron ecosystem, but something performant that allowed those front end tools would be great.
Registered just to post this. Love Purism, want it to succeed. But...
IMHO, the effort is going to fail miserably, unless folks stop wasting time trying to build a phone UI, i.e. another attempt to build Android/iOS ecosystem. There were Sailfish, MER, Openmoko, and whatnot attempts to make a pure Linux-running cell phone. Just stop it. You will burn through your money doing half-baked UI, there will be no adoption, then no developers, then no apps, and thus no users.
As per OP link, Purism just have shown that they wasted precious time on badly looking GNOME Clocks, Emacs, Password manager, a game, a half-baked music player, Torrent client (on a phone!), and Drawing app made with their native UI. What a waste of time to re-write (or port) all of this, all over again.
The only surviving plan for any Linux-phone: make it web-centric. You have to port ONE app: Firefox. Make it fast, make it perfect. Then, automatically you will get:
- Adoption. http://m.uber.comhttp://m.lyft.com/ work out of the box. I am not leaving home without the phone because it enables to access essential day-to-day services. Partner with companies that develop those web-based apps.
- Adoption. Web-based music: Spotify, SoundCloud. Out of the box. Purism doesn't need to waste time on this. Just have pre-installed bookmarks to those apps. Partner with companies that develop those web-based apps.
- Adoption. Endless web apps such as "Clocks", "Notepad", Games and drawing apps already exists. Purism don't need to waste time on this stuff. Just have pre-installed bookmarks for those apps. Partner with companies that develop those web-based apps.
Once basic needs satisfied, those of us who need Emacs, will be able to port Emacs and Torrent clients themselves. Why waste your time, Purism?
Focus on releasing the hardware, be different from other phones, be lean, get immediate adoption, community will fill the blanks.
Sailfish has a lot of closed-source components and it runs on closed hardware that forces specific kernel versions and makes it difficult for the community to upgrade; one is completely at the mercy of Jolla or the device maker. It shouldn’t be compared to this project.
And seriously, basic functionality should be web-based? A phone should not be required to have an internet connection to do such things. While I am on a plane, or when I am traveling in a country but unable to buy a local SIM, then I should be able to expect that my phone’s non-communication-related apps will still work.
> So you still have no idea what they are doing.
> They are making a privacy-focused, portable Linux device that makes calls.
Now understand, thank you.
Before your kind explanation, I indeed did not fully grasp what Purism was doing. I was naively thinking they were doing a privacy-focused device that I could someday use everyday.
But instead they are doing a device that will run Linux and will occasionally make calls. I had one of those: Openmoko. Still have it in a drawer.
Without making a device that is actually useful, which will survive more than one iteration, those are futile attempts. Truth is: one small company will not be able to make a full well-integrated phone software in the time/money the company is allowed to burn. Unless they embrace "outsourcing" apps to existing ultra-portable (web) ecosystem, the attempt is doomed to fail.
Sadly, but if the course continues, I will make my prediction: Purism will successfully release Librem 5, there will be a half-baked UI (like MER/Openmoko/Sailfish/OE), half-baked bunch of native "Notepad" apps, no one will seriously use it, and it will be the last phone Purism will release.
Then, the circle of life: a new company will try the same.
> I don't understand why people seem to have it out for Purism. Did they wrong you? Are you a competitor?
I love Purism. I mean no sarcasm here.
But you have to know history and learn from previous mistakes, and try something different. If Purism's approach is the same as to what Openmoko was, it is not going to end well. I hope they will try something different.
But, so far, they do exactly same thing as to what Openmoko was doing. I loved Openmoko, I bought their phone (and a debug board), but I was never able to use it as a phone. It did make calls. Due to limited resources (and all small companies have limited resources), they could never master native UI. I still don't believe that a small company can make all-encompassing software apps for a phone, thus I suggested to try something different: use existing web apps, don't spread already thin resources on "porting" native apps yourself.
> The only surviving plan for any Linux-phone: make it web-centric. You have to port ONE app: Firefox.
You are aware of the failure of firefox OS right?
> Purism just have shown that they wasted precious time on badly looking GNOME Clocks, Emacs, Password manager, a game, a half-baked music player, Torrent client (on a phone!), and Drawing app made with their native UI.
They haven't wasted time, pretty much all of these are existing apps, you can run most of them on a gnome desktop.
> Focus on releasing the hardware
That is what their doing and what differentiates them from most previous efforts like sailfish, they are a hardware company making a phone the runs linux, others were trying to re-create android and leave the hardware for others.
1. Allow non-kernel updates without carrier interference. Google has had to learn this lesson the hard way. Let the carriers lock down the lower layers, but Firefox and the user-facing stuff should still be able to update well enough for a few years.
2. Don't launch on crap, low-memory devices. KaliOS works on low-end devices, but that's because it is hyper particular about which apps run and what web APIs they use (and a lot of websites don't run very well). There's no reason a web device can't work for high-end devices. Most people run the same 50 apps as everyone else (and 25 of those ship with the phone anyway). High-end devices also open up other options like running Android in a linux container (like ChromeOS does).
3. Make WebAssembly (then asm.js) first-class. "Native" wasm apps with more direct access to APIs and access to a standard set of canvas widgets (sorry, HTML sucks for a lot of things). New APIs need to be available for more specific cores (especially things like DSP or ML units). Finally, add support for fully-native processes for trusted programs because sometimes 25-50% slower is too much and sometimes a spec (eg, Vulkan) simply isn't available as a webAPI yet but competition is needed here and now.
4. Be different. Make something that tries to clone Android and people will expect it to be Android. Firefox OS would have gone much farther looking and acting like webOS (which is still a better UI than Android more than a decade later). You can't disrupt a free, mature project with an identical, but immature project and a bunch of ideals that 95% of people simply don't understand.
> You are aware of the failure of firefox OS right?
Right. And this was a mistake by Mozilla: they tried to make an OS, and a browser and interact/negotiate with 3rd-party phone vendors. I did not care for FirefoxOS-based phones indeed because they where "same old ridden-with-firmware black boxes".
Mozilla makes a good browser, they'd rather keep doing that.
Purism is a different story. It seems they understand that some folks appreciate open hardware, privacy-focused designs. Cool, one ingredient for the success: check. But in order to be a successful product, the phone has to be useful out-of-the-box. And if the hardware company will attempt to make a whole "native" software stack, they will fail the same way as Mozilla failed at being an OS-company and HW-facing company: not enough resources.
So, how about Purism makes a great hardware (+ drivers), and integrates exactly three apps: 1. a phone app that is able to make/receive calls, 2. a short-text messaging app 3. and a blazing-fast browser to access the world. Done. Basics are there: I can make calls, I can receive/send messages, I can access email, slack, hail uber/lyft, use web-based maps/navigation, listen to online podcasts/music, use web-based calendars, etc. A phone I could use everyday.
Purism pretty much is doing what you're suggesting.
They're working with upstream packages so that when you install Linux on it, it just works, no hunting for additional drivers to make it work.
Those apps already exist. Again, they're working with those apps to make the phone "just work" when you install the apps/programs/packages/whatever. They don't need to reinvent Firefox, they can let Mozilla focus on that, and Purism can focus on what they're doing.
If someone were to develop an iphone-quality design with the same objectives as the librem 5, I'd pay $3000 for it.
How much would the market shrink if that was the price of what would become the librem 5? It's rather niche to begin with, but there isn't any real competition the meets the same objectives, and I assert that the customer base has the money to afford it.
Is there any reason why applications cannot be written for the Librem 5 with Flutter? The Flutter framework compiles to ARM binary and the UI is rendered with Skia, a C++ UI layer. It's not like a bunch of widgets would need to be written and supported on Librem, just the Skia layer. What am I missing?
That would be a massive amount of work. In my opinion, Purism correctly limited the scope of the original proposal -- a FOSS phone that can make calls, and send messages (and maybe use a web browser?). libhandy (the library that allows GTK applications to "shrink" their UI for a phone) is intended to minimise work by allowing Purism to take advantage to existing software (e.g. Settings, Web browser). Even this is a huge amount of work (I've written about it here/Reddit in the past).
I'd welcome more push behind Plasma Mobile for it. There is practically no progress now. It could benefit from moving towards more touch oriented design, Sailfish like, dropping lingering elements inherited from the desktop UI.