Actually no. The disruptive aspect was supposed to be price and pace of development.
I suppose that's a fair assessment, ultimately price didn't matter since OEM licenses for Windows were so cheap that people wouldn't even notice the cost built in to the PC and while there has always been a lot of development going on it hasn't been particularly unified so there is a lot of duplicated effort in order to do a bunch of things in slightly different ways.
I'm guessing what he is referring to is the ongoing debate developers are having with Stallman over accessibility to GCC's abstract syntax tree. Many developers want this in order to be able to do auto-complete, refactoring and other IDE options within programs like EMACS but Stallman is concerned that this could lead to that abstract source tree (AST) being used as input to proprietary compiler backends, which he sees as bad.
This is another page in the book of hamstringing GCC and EMACS users in the name of freedom, or rather of trying to prevent a perceived threat from closed-source programs. We have already seen the long debates about plugins for GCC which ultimately ended with it being doable so long as it exports the symbol plugin_is_GPL_compatible and as David Engster already said, a GPL plugin for GCC could already be written to extract the AST if that's what compiler backend authors wanted to do.
Except on tablets, where it's still illegal.
Does it really matter? I mean Free Software doesn't want proprietary developers standing on their shoulders so they have the GPL to force their position, proprietary product developers don't want Free Software leveraging all the work they have done creating their hardware platform so they lock it down.
If we want Free Software on these kinds of products then the Free Software movement is going to have to address the fact that there is no platform on which to run. Proprietary product companies went out and developed or licensed their own platforms so the Free Software movement needs to do the same.
The dedication to pushing a free desktop is admirable, but it takes a disruptive product to effect real change in an established market. The disruptive aspect of it was supposed to be "freedom" but that wasn't enough. As you highlighted, the new smartphone and tablet categories still don't have entrants from the free software community so in that area it's another case of trying to play catch-up with the only ace up your sleeve being the freedom card.
With these new 'device' categories it is even harder because nobody sells a 'blank' device - a generic platform like the PC on to which you can load any operating system - nor can you easily turn most existing ones into that. Unless an OEM comes along happy to set a platform specification and sell blank platforms (like Raspberry Pi but obviously more easily integratable into small form factors) I don't see much progress here. The only way is to leverage the proprietary vendor platforms and hijack those with some innovative and disruptive alternative.
Their open-source contributions have always seemed to me to be letter-of-the-law vs. spirit-of-the-law.
Their open source contributions seem to be in the spirit of open source rather than the letter-of-the-law, as you say they aren't legally required to release their contributions but they do anyway. Sure they aren't releasing the source to everything that the code is linked with like restrictive free software licensing enforces but that's the great thing about permissive open source, it allows free and proprietary authors to work together and leaves the choice of what contributions to make up to the author and not restrict other authors. Whereas restrictive open source licensing is a "my way or the highway" approach.
You can't build Mac apps on Darwin alone, which is why it makes sense to say that "Darwin is open source" but OS X is not.
I would certainly agree with that. I was only disputing that the parts of BSD that Apple did fork have become proprietary - because they haven't - they have formed the open source basis upon which Apple's proprietary frameworks are built.
which was based on MACH, which was derived from BSD
Mach was developed as a replacement kernel for the existing kernel in BSD operating systems. But that's neither here nor there, of course it has many BSD components but the idea that it was "proprietary forking" is nonsense, the parts derived from BSD and Mach exist in Darwin which is open source.
A kernel is not an operating system.
Obviously. Apple also didn't take "BSD and used it as their operating system", they did however use parts of the BSD kernel in their open source XNU kernel. There may also be BSD parts used in their open source operating system called Darwin. But there was no "proprietary forking", you can freely get and use the source code.
but there's also no end to the small one-person projects that can be done, and people can and do make a living like that, here in 2015.
Yet even some of the most popular projects that are used by both FOSS and proprietary products and companies are woefully under-resourced (OpenSSL) and underfunded (GPG).
Oh, is that all? People just have to "make a distro that doesn't use it". Like that's just a small thing, like finding a typo in the documentation or some such.
It's not even that, they just need to maintain the existing non-systemd codebase. There are distros that don't have systemd in them and if they have plans to integrate it then the non-systemd codebase just needs to be forked and maintained. You can hope this is done for you by volunteers, you can do it yourself or you can pay somebody to do it for you.
This seems like more a case of wanting one's free software to also be free of cost and free of effort.
I said that it "never happens" that you can't run newer Android on older phones for any valid technical reason.
Not only did you not say that but there is a valid technical reason: The drivers don't exist!
If we use your definition then there's no "technical reason" why any OS couldn't run on any kind of smartphone hardware. "Hey IRIX runs on the Galaxy S3!"
They can't run it because the drivers are closed-source and incompatible with newer kernels.
Yes I'm well aware of why they can't run it, but that doesn't change the fact that they can't run it. Telling me it "never happens" is just bullshit, it happens all the time (yes because of closed drivers).
And at that point, it's not "choose systemd", it's "choose not to maintain a fork".
But systemd didn't exist before, the systemd stuff should already be in a conditional that checks whether it is even there.
Their philosophy just doesn't work for Apple to have a majority market share. They and their customers can't stand being that 'average.'
Yet the most common and best selling smartphone in the world is the iPhone. If you want to have a "boutique" device or something other than "average" then you would be using anything that isn't an iPhone.
"Pay no attention to the man behind the curtain." -- Karl, as he stepped behind the computer to reboot it, during a FAT
