Flatpak Gets New FreeDesktop SDK 18.08 Runtime

[starshipeleven]

Splitting front end and back end in a single final product in marketing terms means incomplete job and nothing else.

You got it wrong, "splitting front end and back end" means that the backend is a final product, and the frontends are final products.

You can even be delivering the cure for cancer that still no more than 1% tops of the possible clients would even consider looking at it.

That's because the backend clients aren't end users perhaps, but distros or other entities that want to integrate the backend in their pre-existing package manager application, like Gnome Software, or KDE Discover.

The old unix way for sure but... seriously, how many brain surgeons does it take to realize that the model is broken and does not work?

You know right that there are commercial applications that are just backends or just frontends too? Or that in most applications the team working on the backend is not also working on the front end?

It requires different skill sets, so it HAS to be done by different people, period. So it makes sense to actually split the codebase too, or even have it done by different projects alltogether. The real "Unix way" is just a collection of good programming practices.

I really hope that one day Linux programmers will read a marketing book and start working as a team for the whole.

I really hope that one day all the kids ranting in forums will stop assuming they know better and actually read some marketing book too so they can actually understand the reality around them and not post naive bullshit.

If that day ever happens the first thing we'll get is a 'Linux Google Play Store clone' for all Linux apps.

Did you ever visited Flathub in the last decade? It is integrated with Gnome Software and Discover, so if you click on install it will open up these applications to do the leg work.

This isn't exactly new technology either, OpenSUSE has been doing something similar with their OBS (community supported packages) for ages, you click, it downloads a file that is automatically opened by the Yast system management thing that then proceeds to add repo and install stuff.
Same for ubuntu and their own stuff, and probably others I don't know about.

Simultaneously making android apps run on Linux will be top priority, or at least having a working Android app player for Linux. Yes because I hope someday someone sill realize that there are more than 3.300 million Android apps (up to date).
Then and only then there will be the basics to show the general market that Linux is an option because it offers what the market is looking for.

Since you seem to be completely oblivious to the reason the only "Android compatibility layer" that didn't go unmaintained after a year is done by Google themselves, and is still for their own products (Chromebooks), I'll disclose this highly guarded secret just for you:

It would just be working for Google, for free, as we all know that Google Play store is a Google-only source of revenue. You can claim all you want, but If you talk of Android applications that's what it means, giving money to Google one way or another.

And if/when that day comes and the Linux user base grows exponentially, as so will grow donations for the projects and revenue sources (a 'Linux Google Play Store clone' supported by advertising and donations anyone?). But this is nonsense, I really hope I do not have to explain how multiplying the user base can be synonymous of multiplying the revenue, I really hope things are not that bad.

The main issue here is the large investment required to get anywhere near that. It simply does not pay back in any reasonable timescale, especially if you rely on donations and "responsible advertising". Especially if you are basing all your success on an Android compatibility layer, that basically routes all profit and avertising to Google.

There is a reason if Google has made Android itself (and its potential successor Fuchsia OS) opensource with a permissive license, and Microsoft is doing all it can to actually get have a foothold out of the "desktop OS" market. It's not a particularly highly profitable market in this day and age where you can't rely on hardware sales to sell licenses locked to the hardware.

And what about all companies making Linux-based distros? They live off company service contracts, even Canonical finally got the memo that the consumer OS market won't provide them enough donations and revenue sources (without pissing off their audience like they did with the Amazon tracking debacle for example), and ditched all their Unity DE and convergence and mobile effort overnight to focus more on what actually pays the bills (server features).

Sorry for being so harsh but by know smooth talk does not seem to take any effect at all, so maybe some hard love will help change mentalities.

You need to understand that ranting like this in a forum isn't what opensource needs.

Opensource requires time and resources, just like everything else. The reason it's not getting them is that it's hard to actually profit off it in the consumer market, so you see many hobby projects (the music players and note taking applications you mentioned, for example), and many companies that do it on the side of their core market which is server or company workstations.

[andyprough]

pemartins

And the command line died in the beginning of the nineties with ms-dos, so why the hell do we keep on holding our Linux back with this basic mistakes?

I like your trolling style. Impressive.

[Darakus]

Just wondering when the flatpak team will find out that we're in 2018 so Flatpak can get a GUI.

You know, the command line died in the beginning of the nineties and the most popular and used tech devices (smartphones) do not even have a keyboard, so... if one day the Linux developers realize this, for sure the Linux market quota of 1% will grow.
We live in the click-and-done era, people. And even that is ending, the speak-and-done era is taking over. And the command line died in the beginning of the nineties with ms-dos, so why the hell do we keep on holding our Linux back with this basic mistakes?

Just my 2 cents as a daily bash user.

Just about all the Linux users I know cite the command line as one of the best features of our favourite OS. A lot of Windows users I know love powershell. I would also argue that Apple switching to a UNIX like system for Macs, along with its powerful command line, was one of the things that brought apple back from the brink and made it enough money to get into the smartphone business and eventually win capitalism. For Linux, if I had to choose between a good GUI and a good TUI, then I'll take the bash friendly option any day.

Fedora 28 allows you to enable Flathub via GNOME Software, and after you do that you get to manage your flatpaks with clicks, just like you want. There are many things that could be improved on Linux, including flatpaks, but saying that flatpak lacking an official GUI is slowing Linux growth, when the two flagship desktop environments already have GUI package managers that support flatpak, is a little bit silly in my opinion.

[RealNC]

So does that mean apps will now not look totally alien compared to my native apps and not throw my desktop settings down the toilet?

[nanonyme]

You don't know what happens when programmers are designing application GUIs right?

Linux follows the good unix tradition of splitting applications in a backend and a frontend. This allows programmers to actually focus on the real deal (the backend) while programmers with a different skill set can work on the frontend.

Btw, also Kde Discover (a Gnome Software equivalent for KDE) supports flatpak.

KDE Discover had issues following deps earlier, I don't know if that's been since fixed. I've been trying to find someone who could actually test this

[nanonyme]

I tried to twice comment about Steam but forum is eating my comments. Rest assured, new runtime is close on roadmap once a severe unrelated bug is fixed
Please help test https://github.com/flathub/com.valve...team/issues/60

[pemartins]

Splitting front end and back end in a single final product in marketing terms means incomplete job and nothing else. It's a bad job, plain and simple. You can even be delivering the cure for cancer that still no more than 1% tops of the possible clients would even consider looking at it.

2018 and the first steps towards one being able to install every Linux app in every Linux distro are just now being given. We still do not have our .exe or our .apk. This is not even worthy of qualifying, it's just a bad joke. And even when those first steps, still baby steps, are being given, it still needs to be integrated god knows where with what, plus doing this, this and that...
The old unix way for sure but... seriously, how many brain surgeons does it take to realize that the model is broken and does not work?

I really hope that one day Linux programmers will read a marketing book and start working as a team for the whole. If that day ever happens the first thing we'll get is a 'Linux Google Play Store clone' for all Linux apps. Simultaneously making android apps run on Linux will be top priority, or at least having a working Android app player for Linux. Yes because I hope someday someone sill realize that there are more than 3.300 million Android apps (up to date).
Then and only then there will be the basics to show the general market that Linux is an option because it offers what the market is looking for.

And if/when that day comes and the Linux user base grows exponentially, as so will grow donations for the projects and revenue sources (a 'Linux Google Play Store clone' supported by advertising and donations anyone?). But this is nonsense, I really hope I do not have to explain how multiplying the user base can be synonymous of multiplying the revenue, I really hope things are not that bad.

Until that happens we'll keep on seeing many more music player apps and note taker apps and similar unneeded projects and few more than that, while having beautiful amazing things like kde and deepin, having light malware-free and fast operating systems given away for free and nobody even considering looking at them.
Sorry for being so harsh but by know smooth talk does not seem to take any effect at all, so maybe some hard love will help change mentalities.

[nanonyme]

I hope that applications like Steam and Blender move to that runtime fast... Mesa is quite old in the 1.6 runtime (18.0.4). And I hope that Mesa gets updated more regualary... sad to see that NVidia drivers are getting way more love than Mesa ones.

I had an answer to this but it got stuck in moderation queue

[starshipeleven]

Just wondering when the flatpak team will find out that we're in 2018 so Flatpak can get a GUI.

You know, the command line died in the beginning of the nineties and the most popular and used tech devices (smartphones) do not even have a keyboard, so... if one day the Linux developers realize this, for sure the Linux market quota of 1% will grow.
We live in the click-and-done era, people. And even that is ending, the speak-and-done era is taking over. And the command line died in the beginning of the nineties with ms-dos, so why the hell do we keep on holding our Linux back with this basic mistakes?

You don't know what happens when programmers are designing application GUIs right?

Linux follows the good unix tradition of splitting applications in a backend and a frontend. This allows programmers to actually focus on the real deal (the backend) while programmers with a different skill set can work on the frontend.

Btw, also Kde Discover (a Gnome Software equivalent for KDE) supports flatpak.

[starshipeleven]

Does it mean that the runtime increased in size and bundles even more Gnome components ?

Yeah it requires every user to buy a 8TB helium-filled hard drive to store it.
/sarcasm