Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

Unreasonable? Dude you sound like "we have to support jolla at any cost"... commercial firm + their decisions + their additional work = their problem, if not sailfish why not tizen... what happened to go with what fills your need. By buying their product you are already risking your money and supporting them enough seeing its a new ecosystem and a new firm. Im not saying we should not support them, i just dont see a reason why if theyre unwilling to make needed efforts. Cmon see samsung and tizen, im not asking jolla to throw cash though, just saying how much theyre lacking to make developers intrested in their ecosystem in a much crowded fierce competition.

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

Yup, unreasonable to expect them to be the sole arbiter & driver of large projects that benefit far more than just Sailfish, completely unreasonable.
As said, they've already made some significant strides towards a solution for the entire community (not just Sailfish) when it comes to Python/Qt bindings.
Hopefully others will jump on-board & help with that, I've no doubt they will, it always happens....
It'd be great to see them make a similar start with XWayland, unreasonable to expect them to do it completely & utterly on their own though...

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

Then i cant help but say "each to their own", still you have a point there. I do wish luck to jolla though, and i am curious to what lies ahead especially about wayland and qt5, in future i may even get a jolla device if it turns out right. But incase i wasnt clear this is the point i was trying to make....

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

Yup, I'm basically just waiting on THP to publish the updated PyOtherSide code, so I can start playing with it. :)

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

I disagree. The current situation for mobile developers from Maemo/MeeGo/Nemo Mobile to Sailfish pathway is IMHO :

1) softwares in C/C++/Qt 4.7-4.8 (Qt Quick 1) : about 80-90% of MeeGo Harmattan and Nemo Mobile softwares, just some % of Maemo 5 softwares. Anyway due to migration from GUI written in (Harmattan, etc) Qt Quick components to Sailfish Silica (Qt Quick components for GUI in Sailfish OS), there is a need of small-medium source code revision & rewrite. There are 2 options for Qt5/Wayland migration :

1.a) XWayland (X11 running as a Wayland client) be available by Jolla and/or community, so all Qt 4/X11 softwares will remain compatible. Jolla said "just to clarify, while we don't support xwayland atm, we're not ruling out the possibility in future...". It seems XWayland source code isn't long, just some hundreds of lines of code. So, as Jolla itself is open to include XWayland (with the community help, i.e., in Nemo Mobile Wayland effort) and XWayland is a lot smaller than Wayland, I think it is possible to have XWayland ready on Sailfish OS until the end of 2013 (but not included in the Sailfish SDK beta which is soon to be released);

1.b) migrate to Qt 5/Qt Quick 2, which seems to be easy, with minor source code revision & rewrite, then recompile. This work may be worth to also release Qt 5 versions for Ubuntu Touch, Android and iOS (changing from the Qt components from Silica to Qt Components for Ubuntu or Qt Quick Controls);

2) softwares in Python/PySide/Qt 4.7-4.8 (Qt Quick 1) : only about 10-20% of MeeGo Harmattan and Nemo Mobile softwares, just some small % of Maemo 5 softwares. But Python is important, allowing party-time developers to code on device (N900, N9, etc), anytime, anywhere, and contribute to the software diversity. The same migration from Harmattan (etc) Qt Quick components to Sailfish Silica is needed, with small-medium source code revision & rewrite. There are more options for Qt5/Wayland migration, none ready at the moment :

2.a) the same (1.a) option, i.e., release XWayland for Sailfish OS, so PySide and Qt 4.8 could be used in Sailfish as in Sailfish SDK alpha;

2.b) port PySide to Qt5/Qt Quick 2, which has no timeline and is a huge task needing contribution. This effort would benefit Sailfish, Ubuntu Touch, Android and iOS developing in Python/Qt5/Qt Quick 2. So it is expected that Jolla will not contribute (alone) for this huge work;

2.c) use a future PyQt 5 version, because the current one excludes Qt Quick 2. There is also the issue of GPL license and release PyQt 5 for Mer/Nemo/Sailfish;

2.d) use PyOtherSide from thp, which will have a version for Qt5/Qt Quick 2 released soon. It is not a Qt 5 full binding for Python, but it seems a way to code (the GUI) in Qt Quick 2 and call Python from it (so Qt5 is not available from Python, only Python is available from Qt5/Qt Quick 2). As advantages, PyOtherSide is smaller, faster to load and is/will be multiplataform (Symbian, Android, MeeGo, Sailfish, Ubuntu Touch, etc).

So, non-Python developers have a migration path currently available to Qt5/Wayland on Sailfish. Only Python developers are in a wait-to-see/code situation, the easier one seems to be PyOtherSide (2.d).

About the change from Sailfish SDK alpha (with Qt 4.8 & X11) to Sailfish SDK beta (Qt 5 & Wayland) : yes, it is a bad surprise for developers who have already ported/created softwares using Sailfish SDK alpha. But currently there are a very small number (10-20) of softwares written for Sailfish, see the app list in Sailfish - Mer Wiki. So this 'collateral effect' is small.

I trust the Linux/Maemo/MeeGo/Mer/Nemo/Sailfish community efforts so that more than one good solution will be available for developing Sailfish softwares.

I hope XWayland will be released on Sailfish, so X11 legacy will be available : ssh -X, chrooting Debian/etc, many Linux desktop softwares not Wayland compatible, etc.

No, there are many differences between Android and Linux, not only X11 : lack of glibc (so porting Linux softwares is a difficult task), fake multitasking, no dependencies (like .deb or .rpm, so the user sometimes install the same library 5 times repeated wasting storage space), no open development with community participation, no community repository (for libraries, dependencies, etc), etc.

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

Why would Android allow Jolla to use their apps? Its not gonna happen.

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

Let me play devil's advocate here and add 'why not android'

if you port your app to android instead of jolla, then:
1. you've probably learned valuable(literally) dev skills
2. there's a good chance it will run on jolla
3. you don't need to worry about jolla changing their mind again.
4. you don't have to get involved in porting toolkits or xwayland (which was a serious suggestion by one member!!)
5. your potential audience is 10's of millions

win win.

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

And, how do you propose they're going to stop that, without looking like total pariahs?
(much less wasting a large amount of resources stopping what's mostly a positive thing for them)

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

Regarding open repos; ever heard of things like F-Droid.org? And in regards to wasting storage space: yes that's real world problem, but... after going through DLL hell (again) on Maemo, i now appreciate if upgrading some library for some tool doesn't break a multitude of other apps too. Besides that, i don't care about space, 16 to 32GB seems to be the norm now.

And besides porting Linux applications; without them being touch-optimized there are more issues than missing out on glibc calls.

I give you on the fake multitasking, really. It could be better. Just so much: In the end (real world experiences for 8 months now), it impacted me less than i feared. Because everything that is fake-switched in Androids method of things just works. No RAM shortage, no DLL hell, no lame processor. And if devs implement background tasks, nothing is lost. So much for theories.

Re: Sailfish(Jolla): Ideas/Qns & Concerns/Criticisms
 

I'm the owner of a Asus Transformer TF101 tablet + Asus Eee Dock, rooted, with Ubuntu chroot, etc. And also a Nexus 4, but mainly for Ubuntu Touch. Yeah, I know F-Droid catalog since a long ago, this is an alternative way of installing softwares, while on Maemo 4/Maemo 5 third-party softwares (libraries, softwares with GUI, etc) are first class softwares available in Maemo.org repositories.

My N900 has more than 100 softwares installed, many programming tools, Easy Debian, etc. And I've never had any problem of library conflict breaking other apps.

I really have some libraries 5x repeated in my Android tablet. Some simple softwares take 10MB where 9.9MB are dependencies embedded...

Well, take Linux libraries and CLI tools, without GUI, without glibc it's a nightmare to port them to Android. Some aren't yet even after years of effort from the Android community.

Sorry, but after 1 year and half of Asus TF101 (Android 4.0) experience and even some tests of Android 4.2.2 on Nexus 4, I am the witness of many situations of softwares been closed, losing real state, etc. All my friends who are developers and have Android devices cite the same thing, it happens with all Android versions, all hardwares. A mobile OS (Android 4.2.2) on a quad core CPU and 2GB of RAM (of Nexus 4) showing it can't deal with real multitasking (even using Android default softwares) is a joke, a dumbed down OS.

RAM and CPU aren't features of the OS. And even with background tasks, Android softwares have multitasking limitations (even native ones).

Sorry, I am a ex-Psion (EPOC 16/32) user&developer, where a 32-bit real multitasking mobile OS (with micro-kernel) on a 16MB of RAM was a lot better than current mobile OS in stability, multitasking, advanced features, etc. My Psion Revo+ lasted 2 years without rebooting, even while I was programming on device in OPL anywhere, anytime.

So, from my ex-Psion point of view, Android, iOS and Windows Phone are just dumbed down mobile OS.