Interesting analysis:

http://www.lieberbiber.de/2017/06/20...s-post-mortem/

Could there be something that Jolla and SailfishOS could learn from Ubuntu failure?

Excellent read, great analysis! In my opinion, especially the first point is crucial for success and that's where Jolla might have better chances. They now target big governments which usually have a lot of cash in their pockets. Moreover, they avoid the bloody battleground of the consumer market.

However, many of the points also apply to the SFOS ecosystem - especially the lack of devices and (as an app developer) the still very unsatisfying status of officially available APIs. If you want to write a nice app which uses some of the features which make a smartphone really smart, you are probably not allowed to use Harbour and end up in OpenRepos (fortunately, we have this great platform).

I hope, Jolla's financial situation will allow it in the future to invest much more in developer experience - which is crucial both for the governmental use case and for the greatest of all communities.

Yea, I was not even close to being a dev but the system was weird and I couldn't be comfortable just going into terminal and using familiar tools. It was only good enough for non-power users IMHO.
There were few useful apps and it reminded me more of encountering a new strange os than Linux.
The N900 was great despite the closed drivers and never being able to progress beyond the kernel those drivers were designed for, same as the Sharp Zaurus for the same reason.
Ubuntu touch could do little and I ended up reflashing to cyanogen to get a usable if still hobbled and less than power user worthy tablet. Convergence never happened and it is all down to lack of drivers for our hardware, otherwise we would all be using x11 and be able to install any ARM compiled distro.
Unlike laptops where hacking on linux is fun and easy with decent driver support there is no longer any good option for tablet and phone users beyond some sketchy Cherry Trail AMD64 stuff, the N900 might have had blob drivers but it used ALSA and X11 so was pretty easy to compile for despite the ultimately life shortening bin blob drivers and corporate problems killing a promising future.

inspite of analyzing ubuntu failure jolla should analyze their own failure Iijoki update. people are facing random reboots while using 4g internet and jolla has done nothing about it, pathetic.

Thanks for the article - it's a really good read!

Also pretty much summarizes my my findings when researching if porting my apps to Ubuntu Touch would make sense.

And for a short comparison with Sailfish OS - there are IMHO possibly the saving graces why Sailfish OS still exists and Ubuntu Touch doesn't:
- developers being ex-Nokia and actually knowing what they are doing
- sane & nice UI
- not putting the cart before the horse (not doing heavy sandboxing when it's already almost impossible to make a useful app due to missing/broken APIs)
- no reinventing everything, using proven software components
- actual Linux distro underneath with the APIs you expect
- being able to install packages, which makes the system much more powerful & enables OpenRepos (Ubuntu Touch had RO root partition with image based updates - impossible to install packages without also breaking future updates)
- sane application lifecycle - apps are suspended when screen is blanked, but can & do run just fine in background if device is active & can even run when screen is blanked via overrides
- Android emulation (yes, really)

You're kidding... Nokia is quite a brilliant example of how not to do things ;-)

Yes Irony, haha.
Business aside, actually having worked towards a product release and releasing it (more than one time) is what MartinK means with "how to do it" from engineering pov in comparison to ubuntu staff.

Exactly - when I say Nokia in this case I basically mean the "old", pre-Elop N900-N9 Nokia, which certainly could ship nice devices (not just the N900 & N9).

Not the current trademark-selling shell of a phone company. Other than that the non-phone parts of Nokia are probably doing just fine, making mobile network infrastructure and tires.

The main reason is a lack of hardware. Android's success on mobile and embedded comes from the variety and quantity of hardware Android runs on, for this reason there is a large user base to create applications and content for. Android didn't start out with a large ecosystem of applications and content, rather by having so many different devices out there this created a great platform. Android runs on millions of phones, there are Android tablets, Android PMPs, Android Navigation units, Android set top boxes and tv devices, Android game consoles and so forth. Like wise Linux on servers has had success for similar reasons and desktop linux in recent years has had increased success for similar reasons.

Producing hardware should be priority number one for any platform.