Jolla’s Sailfish OS has reached release 1.0 | Hacker News Reader

Jolla’s Sailfish OS has reached release 1.0

jollausers.com

46 pointspiokuc33 comments

Among the new entrants to the mobile market Jolla and Sailfish have the best chance of success based on the team's experience, the capital behind them, the maturity of the software, and the technological approach of not relying on Web apps, having a Qt-based native app stack, and Android app compatibility.

All that makes them good platforms, but how about desirable ones…

I disagree, I see not relying on web apps as a big minus for Jolla. There's probably not room in the market for 3rd native platform, and there's definitely not room for a 4th. Jolla is fighting a current they can't keep up with.

The web is guaranteed to continue to exist and keep getting better, albeit perhaps slower than we would like. The speed in which it gets better is dependent on the number of platforms backing/hedging it.

The compatibility with Android apps mitigates this to some extent.

But gives them the OS/2 problem. Nobody wrote native apps because it could run Windows 3.x software.

Nobody would have bought OS/2 without any compatibility

Right. With a market this thoroughly dominated by other players, you have three options:

1. Be compatible with their software (and nobody will write anything native) 2. Don't be compatible with their software (and nobody will use your OS) 3. Be really amazing and marketed perfectly so that you can build your own ecosystem.

Unfortunately, 3 is much harder to do than the first two.

It seems like Jolla is going for a combination of #1 and #3. They're aiming to be compatible with existing Android hardware, they've made it easy to preview their OS on existing devices without making the switch outright, and they're targeting select countries first so that they can use their momentum to move into other countries (rather than overextending their reach from day one).

It's in addition to web apps of course. The point was that Qt provides an excellent native stack and toolset which the other competitors don't have. You can also develop for Android and iOS with the same tools. Web apps are not there yet.

> It's in addition to web apps of course.

No it's not: https://sailfishos.org/about-technology.html

> We are investigating Cordova Qt and compatibility with Firefox OS APIs for this purpose.

No mention of "add to homescreen" style web apps that require little/no extra effort from developers. They are going all-in with QT it seems, and it's going to fail.

Generic web apps don't need anything more than a shortcut on the desktop, launched with a few command line arguments to run without chrome in fullscreen. I'm pretty sure they already support that, although the default browser isn't very good. FirefoxOS packaged apps would be even better.

This is a bit OT, but I have a theory about the Web being a loser for phones and tablets: There is some empirical evidence that the Web-as-app-runtime works best on bigger devices that are stationary or have large batteries, and that have screens that are a bit too far away to touch.

Chromebook is a winner. WebOS on LG TVs is great. Both of these can have larger memory and batteries and more-powerful and power-hungrier GPUs. They also both have pointing devices. They're not trying to drag the Web into a touch UI.

I think this is an accurate observation (re: computing power and battery life), but it's only valid for now. In fact, it is less true than it was a year ago, and will be even less true in a few more years.

Empirical evidence or simply history? The web started as a platform on bigger devices which were stationary and a bit too far away to touch - of course it still works best on those platforms.

However, "The web" is simply a stack made up (at the client end) of HTML/CSS for layout and javascript for logic. Nothing more, nothing less. Both of those technologies are entirely screen resolution, power source and interaction method independent (with the exception maybe of the aspects of javascript which relate to mouse pointers).

There's no reason that websites can't be built from the ground up for mobile devices. The reason that "the web" (according to your empirical evidence) doesn't work for mobile devices is because the vast majority of the web wasn't designed for mobile devices. Mobile browsers are still mostly dealing with sites designed for 4x the screen size and designed with very little thought for client side execution time. Many popular "native" apps use web technologies (HTML/CSS and javascript) in a way which you wouldn't even know wasn't native - the reason it works is because the markup and JS are targeted towards mobile devices rather than desktop browsers.

> There's no reason that websites can't be built from the ground up for mobile devices.

This is true in principle, so let's see if my theory holds up when generalized to other pre-touch stacks:

Windows? Failed at repeated attempts to touchify.

Linux/BSD? Only systems where the whole app runtime was replaced succeeded at being touchified.

I think that points to a future where, if the Web is to work on touch mobile devices, there has to emerge a consensus around a manageable handful of frameworks with common UX conventions, and apps have to use one of those frameworks.

But even then, native mobile platforms are evolving ahead of Web frameworks. Android's Fragment and ActionBar classes enable UIs to "fold up" and "unfold" across a range of screen sizes. That's part of the expected Android UX now. Capturing that, or creating an alternative UX accepted by the same user base, in any cross-platform system is going to be difficult.

You clearly have very little understanding of the technologies you're discussing. HTML/CSS is a presentation layer which describes a way of drawing things on a screen. Nothing more. To compare it to Windows is tenuous, and to suggest that "Linux" even has a standard "app runtime" suggests a clear misunderstanding of the Linux ecosystem.

Web technologies are a layer below what Android or other mobile platforms expose to developers. They're a platform upon which frameworks can build. I agree that a few frameworks providing UI widgets and allowing easy layout will emerge (as they have in, for example, Bootstrap). I disagree that you can in any way generalise by saying "the web is for desktops" simply because of the history it has.

hello from a jolla :-) it will be hard but 'we' have a chance! there is no other platform which feels like a normal linux distribution and people are very friendly, neither 'inaccesible ad droids' nor grim 'prison guards'. just normal maybe a bit nerdy humans which have built a - imho - great phone. not yet there in certain aspects but (more than) good enough and with ethics that i like and want to support.

Right now the only downside of Jolla is hardware - limited modem which doesn't support many frequencies especially in North America I hope Jolla will address this problem fast.

They aren't yet selling phones for North America at all AFAICT

It doesn't matter whether they sell it or not, but the modem not working there with 3G or LTE is a real usage barrier however. Newest SoCs solve this problem by using a universal modem, but Jolla handset unfortunately doesn't use one of those.

So they either should start selling to those markets and make another run with a different version of restricted modem, or we'll have to wait until their second device which will have a universal one, rendering such issues irrelevant forever.

Although as I'm sure you're already aware, they'll be offering the option of installing Sailfish on a range of Android devices too.

I for one am quite looking forward to trying it on my Nexus 5..

- ed

Mentions of this capability are liberally peppered around the rest of the comments on this page already, I now see!

There is a a potential problem with using Sailfish on just any device. Sailfish strongly relies on gestures which includes gestures from the edge of the screen. Jolla handset (same as N9 and N950) is designed with that in mind. I.e. its ergonomics account for such gestures (flat surface, no bumps and so on). Many other handsets aren't designed with that in mind and can be irritating to swipe from the edge. So if there are good ones which fit this kind of usage, that would be great. At least keep this in mind and test before you buy one for that purpose.

Web applications are nowhere near native ones given current limitations of hardware, performance and power consumption and limitations of browser virtual machines in general.

Given all that, Sailfish is able to use FirefoxOS APIs, so here go your Web applications. Nothing prevents you from using them there.

There is room not only for 3rd of 4th but for any innovative platforms. Android and iOS aren't even innovating anymore, so there is no point to talk about "room" there.

According to their site they are "investigating" compatibility with FirefoxOS' web apis: https://sailfishos.org/about-technology.html

That doesn't say anything about "add to homescreen" web apps, what is the story there? Do they exist? Do they appear as just another app in the task switcher? Do they stay in memory as long as qt apps?

FirefoxOS applications simply wrap the browser and provide a launcher. Some environment integration there may be too. Sailfish browser can run FirefoxOS applications already (being current Gecko based), so I guess Jolla is just looking into integrating it all.

can run FirefoxOS applications already

But without access to the camera, accelerometer etc., right?

I didn't specifically test that. But I guess not, unless the browser itself can access them. That may be one of the areas for Jolla to work on.

> I disagree, I see not relying on web apps as a big minus for Jolla.

Jolla has a web browser right? so webapps work on Jolla. Choice is great,being forced to write "pseudo web apps"(when half of the API is not standard) is not.

I thought mobile web apps advocacy is pretty much dead at this point...

I mean, "the web" has completely failed in creating any impressive app in any category. Technologically it's inferior. It takes control and freedoms away from the user and is typically closed source and dependent on some server somewhere.

How can you honestly promote this?

> Downloadable Android Launcher is also coming to Play Store apparently, and it will allow the Android users to test out how Sailfish looks like and feels like before flashing their devices into it...

That is a great idea.

Press release: http://jolla.com/media/documents/MWC_Jolla_announcement_FINA...

I am seriously thinking of installing Sailfish OS on my Nexus 5 to reduce my dependance on Google.

Can I sideload it on a Nokia X?

Reply on news.ycombinator.com