This is the right thing to do and I hope it is not too little too late. Free/libre OSS won the battle for the server, lost the battle for the desktop, but desktop is not (that) relevant anymore. FLOSS phone ecosystem will be hard to next to impossible, in my opinion, but worth going into. Self-hosted server side software has a very good chance of succeeding. For a long time I essentially wanted something like Synology and QNAP but completely open, with a single package installation interface and ecosystem of phone apps working with my self hosted software.
> Free/libre OSS won the battle for the server, lost the battle for the desktop, but desktop is not (that) relevant anymore.
Desktop won't stop being relevant in the foreseeable future - there is nothing to replace it with for a whole variety of tasks. And usage of Linux on the desktop is gradually growing.
There are still obstacles that prevent someone from switching - lack of Free alternatives to software used in their working environments, lack of knowledge about the existing alternatives, even the lack of understanding of the difference between Free and non-Free software.
If GNU's idea of free software and open source are lumped together, I'm not convinced that the battle for the desktop and mobile are over due to the direction Microsoft is currently heading and the way in which Apple's closed source ambitions are increasingly subject to criticism on the basis of being closed source. Of course the caveat is I'm looking at it more as the 100 Years' War [1] than the 30 Years' War [2].
The reason I take the long term view is that it seems to me that open source development provides advantages that play out over the long term. At the very least, Google and Red Hat's hybrid models look more and more like the main stream default.
The interesting part of it is that those applications are mostly written by GNU/Linux developers, thus decreasing the likelihood of GNU/Linux ever becoming a relevant desktop in world wide market share.
> FLOSS phone ecosystem will be hard to next to impossible, in my opinion, but worth going into
You can run Android without gapps fairly well, I believe there's even a project that replaces gapps and uses xposed to fake signatures so that they look legitimate to replace Google Maps and other in-app widgets frequently used on Android. Or you can run with GApps if you're willing to tolerate some level of proprietary software.
FLOSS on phones is actually in a quite good state, F-Droid is pretty good, many well maintained open source roms. Better than most other platforms.
> Free/libre OSS won the battle for the server, lost the battle for the desktop
I don't think it lost the battle on desktop - if you don't rely on platform(or interest)-specific software like the adobe/microsoft suites and some 3A games then you'll be more than fine on the linux desktop.
Shame that the battle for the desktop is actually (mostly) being fought on the laptop - and Linux sucks there. The default settings are usually extremely poor for powersaving. Why doesn't powertop+TLP get installed, configured and activated by default when Ubuntu/Fedora detects its installed on a laptop? Also, sleeping still isn't reliable (if I close my Macbook with some work on I can be damn sure it'll be exactly as I left it when I open it. With Linux, its still not 100% reliable). Wifi still also isn't nearly as reliable (packet loss, connection drops etc.) as either Windows or macOS.
The fact that TLP exists is a great step. It wasn't that long ago that I had a dozen different custom hacks in place to get better power usage while on battery, now TLP does all of them and more.
Still kind of disheartening that I get 10-14h battery life on my 13" 2015 Retina under macOS, but it drops to 4-5h on untweaked Ubuntu. I can push that to about 7-8h with the help of Powertop and TLP. That's still a vast, vast chasm to overcome for Linux. Yes, I know Macbooks do non-standard things with Wi-Fi and Thunderbolt controllers (special ACPI sleep modes etc.), that isn't a bad thing, Linux should attempt those as well. To end on a positive note: I do agree TLP (and other powersaving tools) are good progress.
It's not you personally, but I see this a lot in tech, in particular from the FOSS crowd 'Well, if you don't need x or y, then its fine!'
FOSS will always be second tier for most people until the plethora of software available is at some relative parity, similar to that of macOS at least. (I'm talking strict availability). The fact that there isn't more cohesive efforts to have big software vendors port their software to FOSS systems I feel like is part of the problem.
I personally feel the FOSS community doesn't have good evangelism to actual software developers that develop non-idealistic software. Adobe is a good example of this. So is Microsoft. So is Autodesk (With some exceptions, granted, but I think those come from the need to run in parallel processing environments that Linux supports quite well, and Maya). Heck, so is Wunderlist. Or 2do. 1password. They don't make native apps or even some semblance of apps for any FOSS platform, and I see no attempts to evangelize coming out of FOSS communities to get these places on their side. I'm sure there are exceptions I'm missing or perhaps some of these examples listed have progressed but my point still stands.
FOSS is great for users, and has really evangelized well to a certain subset of users in particular (mostly the tech literate enthusiasts and developer crowd, myself included). I however, will never accept this is a good answer. We should be demanding better evangelism to get big name products on the platforms. I do the best I can myself in this regard, but I feel that the entire idea that we should be more inclusive of these vendors/companies/developers is scoffed at in the FOSS/OSS communities and I do not inherently understand why. Yes, they create code that isn't meeting the guidelines of FOSS/OSS, but if you don't create the best toolsets independent of them or have some semblance of balance, then the platform will stagnant at x number of users.
I think idealism colors perspectives but reality is much more grey than either side is willing to cede, perhaps.
It's an interesting point of view, but I'm not sure I agree with the underlying argument. Specifically, I have not seen the free software environment stagnating at all. If anything I have seen it accelerating fairly consistently since I first became aware of it in 1985. There was a time when I was forbidden from using free software in my day job. Now I am encouraged to use it. How far we have come.
This point of idealism comes up a lot and I think it's something of a red herring. Keep in mind that the FSF's goal is singular - to promote software freedom. It is not idealism to hold to that mission. It is their only purpose for existence.
Let me give you a real world example of why a mixed scenario is not realistic. My wife has an Android phone that she bought from the phone company. It is on version 5.02. There is a bug in that version of Android where a log file fills up and the phone refuses to connect to WIFI connections that have had a lot of activity. This bug has been fixed for a long time in the Android code base. The code base is open source. I can inspect it, compile it, etc, etc. I can see the bug. I can see the fix. But I can't load the fix on the phone because the phone is locked. I can't even get root on the phone. I can't fix the problem. The vendor has told me that since the phone is 2 years old (2 years!) that no updates will be forthcoming and I should buy a new phone.
The problem is that no matter how free a piece of software is, it is really only as free as the environment in which it runs. We have seen over and over again, that companies will collude to ensure that their interests trump that of the user. A world in which free software exists only as an extension of non-free software is a world in which it is marginalised to the point where software freedom is lost entirely.
It is not idealism that prevents software freedom from working well in organisations that do not want software freedom.
Where I agree with you is that we have a long way to go to allow users to connect the dots between the problems they have as consumers and the protection that software freedom affords. My own wife thinks it is completely reasonable to spend $1000 replacing a perfectly good phone simply because the vendor wishes it. Somewhat unusually, this is a consumer movement held dear by developers, but virtually unknown to the consumers it seeks to protect. This is clearly a problem. However, we won't fix that problem by abandoning the purpose of the movement. Will my next phone be an open source phone that denies me software freedom, or a proprietary phone that denies me software freedom? Does it matter?
I don't know if my point came across well, so I will try and re-iterate what I mean here in some more detail.
To be perfectly clear: I share your motivations, and I cede that the FSF is a lobbying organization for promoting libre software. I want to note that I'm not specifically taking target at FSF here just generally the OSS/FOSS community in aggregate as I have interacted with them.
Now, to the argument! :)
My basic argument is as follows:
1. Its more important to have a libre/free OS than it is to have all Applications ever be libre/free. This is a matter of pragmatism. Your wifes phone for example, would be much better off if the phone OS itself was libre. This in my opinion is the best approach to take. It gives you the bulk of the freedom that one does seek, without completely alienating developers who, in some cases given the realities of today, rightly (or wrongly), don't want to or can't open source their software. I'd rather have their software on an FOSS platform than have to leave a FOSS platform to run their software. Thats just the economics of the world at work.
2. If one accepts #1 to be a justifiable position, which I think it is, the singular focus as far as the software part is concerned is to push that operating system forward with that set of ideals. Too often I find in the OSS/FOSS communities that they don't evangelize the platform, they evangelize the ideas, which again, i completely get, and I think for users, esp. of a technical bent, get it, and use it, and benefit from it. However, I don't think the average user, or even the majority of technical users, can get away with the switch. In part, its because the same logic that makes a FOSS/OSS operating system amazing doesn't particularly scale to Applications. Again, I will reference the missing software of Adobe, Microsoft, even task app like 2Do, Wunderlist, or the lovely Pixelmator, or even Zoho business apps are not available as native Applications on FOSS/OSS platforms, even Electron-type apps aren't scaling to Linux quickly. I think this is because the platforms aren't being evangelized. Everyone is concerned about the licensing, whether they can review an apps code, etc. or they get caught up in other mundane details, instead of unifying around a platform model that allows Apps of all types to exist (which i realize they do now, but I'm talking about community voices).
3. I would say to your point, an open source phone that runs 'nonfree' software is better than a nonfree phone OS running nonfree software. Why? Well, see above, but specifically because things like OS updates and OS changes would be viewable at a source code level, thereby if any apps are making modifications to it, you will see that.
4. The one thing, and this is my biggest thing, is that the lack of evangelizing as a platform for users is until recently, there is a huge lack of direction in the FOSS/OSS community. Elementary OS is the first time I've seen where a Linux distro is actually putting users first and adopting modern design and aesthetics. Another recent version of that is KDE's Neon built on ubuntu. Great desktop experiences. I feel the user experience suffers from a lack of focus on evangelizing and focus on a platform, instead of just ideals, that presents itself as a coherent whole that is accessible to develop for with some consistent standards. FOSS/OSS doesn't have to equal completely decentralized missions split between 2000 organizations. Where there has been unification lately, the better those products are becoming and actually entice users on a platform. High Idealism is well and good, but without focus on the platform, its hard to argue to switch. If you build a good platform, and get developers to make apps for it that people want to use, you are, in my opinion, gaining more than you're losing with this approach.
I have never seen this approach until very recently in the FOSS/OSS community. I'm glad its happening. It looks like huge steps in the right direction. Now if we could get some of the larger organizations behind a consistent evangelized message for desktop and mobile you could have some real change.
Thanks for clarifying. It definitely makes more sense to me than what I understood from the previous message. Personally, I think there is room for a variety of approaches. It's one of the reasons I'm happy that the OSI is around to champion open source methodologies as distinct from software freedom. Likely, as you say, there would be some benefit from people concentrating on building user-focussed end-to-end experiences.
Having said that, I don't think the FSF needs to get involved with that as the role of prioritising software freedom in an ever changing world is probably more important and also more difficult than ever.
One thing I would caution is assuming that encouraging non-free applications on a free operating system is always going to be win-win. I worked at Corel when they were doing their Linux distro. They misunderstood free software badly. They saw it as an opportunity to lock people into their proprietary applications by changing the underlying OS (without having to pay for developing it). This led to some pretty strange business decisions and ultimately wasted a huge amount of money.
There were a couple of spinoffs from ex-Corel employees who tried to maintain this way of thinking and it was always an uphill struggle. Probably the only one to make any kind of profit was TransGaming and I'll maintain that was substantially because Gav State is both a talented and genuinely nice guy.
I think in order to to make this kind of thing work, you need to understand free software (and the goals of free software) at a very deep level. From my experience, it's just no compatible in a natural way.
I agree they did misunderstand OSS/FOSS badly. The last thing a company like Corel needed (or needs) to do, and I know this i very opinionated; is to make their own Linux distro. This is kind of my entire point. A platform as such that is coherent and well evangelized (I know i keep coming back to this, but I was around for the 'we're going out of business in 90 days' Apple years. It took a huge level of evangelism to keep that ship moving) perhaps would have encouraged Corel to just make their software for Linux, which to me is the win. Yes, its a proprietary Word processor or Image manipulation tool, but the OS would still be free/libre, and Corel, at the time I'm assuming this happened, could have been a win with the platform.
I reject, as someone else mentioned, as an aside, that macOS is just free software with a non-free GUI on top. That is missing the boat completely.
Your story with Corel is my experience with the majority of companies I worked for, very few if any, did care for FOSS for anything else then cutting down costs without giving anything back.
This is why licenses like MIT are so loved by such companies, now with SaaS and Web UIs it is even better for them.
