Slashdot Mirror
Fedora QA Lead Pans Canonical 'Propaganda' On Snap Apps (happyassassin.net)
Long-time Slashdot reader JImbob0i0 shares a scathing article by Red Hat's Fedora QA "community monkey"/senior QA engineer on Canonical's announcement about their application delivery mechanism "snap"...
...and how it's going to unite all distributions and kill apt and rpm! This is, to put it diplomatically, a heaping pile of steaming bullshit... The press release and the stories together give you the strong impression that this thing called Snappy is going to be the cross-distribution future of application delivery, and it's all ready for use today and lots of major distributions are buying into it... The stories have headlines like "Adios apt and yum? Ubuntu's snap apps are coming to distros everywhere" and "Snap Packages Become Universal Binary Format for All GNU/Linux Distributions"...
Now, does Snappy actually have the cross-distribution buy-in that the press release claims (but never outright states) that it has? No... The sum total of communication between Canonical and Fedora before the release of this press release was that they mailed us asking about the process of packaging snappy for Fedora, and we told them about the main packaging process and COPR. They certainly did not in any way inform Fedora that they were going to send out a press release strongly implying that Fedora, along with every other distro in the world, was now a happy traveler on the Snappy bandwagon... They just decided to send out a wildly misleading press release and actively encourage the specialist press to report that Snappy was all set to take over the world and everyone was super happy with that.
Now, does Snappy actually have the cross-distribution buy-in that the press release claims (but never outright states) that it has? No... The sum total of communication between Canonical and Fedora before the release of this press release was that they mailed us asking about the process of packaging snappy for Fedora, and we told them about the main packaging process and COPR. They certainly did not in any way inform Fedora that they were going to send out a press release strongly implying that Fedora, along with every other distro in the world, was now a happy traveler on the Snappy bandwagon... They just decided to send out a wildly misleading press release and actively encourage the specialist press to report that Snappy was all set to take over the world and everyone was super happy with that.
Until everybody learns to play well together, Linux and other great open source projects will continue to be fractious tech that people will, rightfully so, find hard to take seriously.
Maybe both sides are right, maybe neither side is... but as long as people take sides, draw those battle lines, polarize the issues to extreme levels, open source projects can never mature (indeed, maturity is a word we can not associate with such feuds). Compromise and communication seems to never be considered.
That's not to say it can't be done, but the types of personalities that stand in the forefront of Linux, for example, seem very bull-headed - obsessed that their way is right, and never willing to accept constructive criticism or the possibility that there may be a better way of doing things.
Shuttleworth pulls this shit regularly. Systemd, Mir, Unity, etc.
Pretty sure systemd was pushed -- at least initially -- by Fedora/RedHat folks.
Unification is what linux desperately needs in order to make it possible for third party closed source vendors to target the platform. Otherwise they just make ubuntu binaries, and that's it.
The few times that I've ever needed or wanted to install anything that was only shipped in binary form, the packages that were put out for one or both of the popular distros (usually Ubuntu and Fedora) could easily be packaged and run on others distros (such as Gentoo and Arch which I use). Actually, by the time I wanted to install those things on my distro, somebody else had already made an ebuild or packaged it in AUR, so.... do we really "desperately" need whatever this snap apps thing is? Doubtful.
I wonder if snap is coming to Windows? It would be great to install snap packages regardless of OS since Ubuntu is coming to Windows 10 anniversary update
http://saveie6.com/
Shuttleworth pulls this shit regularly. Systemd, Mir, Unity, etc.
What? Are you confusing systemd with upstart?
Flatpak’s (Redhat's preferred alternative to Snap) developers have been communicating with technical conference presentations and blog posts and trying to build a dialog with application developers and distributors
That explains how systemd worked, too. Systemd talked a lot with the people who write startup scripts, at both redhat and debian. They tried to be responsive to their concerns, and give them what they wanted, which is why systemD succeeded.
Just as notable is who is missing from the dialog: the actual users. Which explains why systemd made startup-script writers happy, and a bunch of users upset.
"First they came for the slanderers and i said nothing."
Would greater consistency between the Linux distros make it an easier platform to target with binaries? Of course! But what we've seen happen with systemd, GNOME 3, and now these "snaps" is that this unification is happening around the worst technology around.
A typical Linux distribution today consists of the Linux kernel, which is actually quite good, and everything layered on top of that is shit.
We hear about systemd causing one problem or another far too often. For example, it was just a few weeks ago that we learned that a systemd change broke software like screen and tmux!
X is a trainwreck, but Wayland is even worse. With X we get a prehistoric relic, but at least it's somewhat usable. With Wayland all we get wheel spinning pointlessly in mud, going absolutely nowhere.
Then there's GNOME 3, which is by far the worst user experience we've ever seen from a Linux desktop environment. It's a tablet-oriented UI that nobody actually uses on a tablet, yet desktop users who try to use it are forced to endure the tradeoffs made to try to cram a desktop environment onto tablets.
People were attracted to Linux in the first place because it offered software that was better than what we found in Windows or Mac OS at the time. But that was 15 years ago, and times have changed. OS X and Windows provide a much better user experience these days for desktop users. FreeBSD is now indisputably better on servers.
These efforts to unify Linux have ruined it, I think.
"Unification" is exactly what GNU/Linux *doesn't* need. Innovation creeps up in the FOSS world through other projects, NOT through iteratively improving one core. I'm not a fan, but it's undeniable that systemd didn't become popular simply by building on initscripts. Even OpenRC doesn't do that; it uses sysvinit, but the scripts are entirely its own. Iterative development is not enough to create large breakthroughs. In the systemd example I gave it's really not that large because it doesn't actually do anything new, but it's still such a large change that it wouldn't be possible if it wasn't its own project.
In the "wonderful" unified Linux world, you'd be stuck with what you're given. If upstream doesn't want your contributions, you no longer have the (realistic) option to fork. You either get it into the mainline release, or your innovation or improvement simply doesn't get used because everyone's already "agreed" on this base.
Unification destroys the concept of distributions, destroys the ability to fork and build, destroys the ability to greatly change and innovate. I may not be a fan of some of the recent changes (such as the systemd cult), but I absolutely respect the shit out of the environment that makes that change possible or even easy. Whether you're running OpenRC, systemd, sysvinit, runit, or whatever, you're still GNU/Linux at the end of the day. The ability to pick and choose your software in Linux is a feature, not a bug. If you don't want configurability or choice, go use a BSD or literally any other OS.
The main issue with "targeting" is itself. Software should depend on as general a base as possible. For example, it shouldn't be using low level dependencies with unstable APIs. Most applications (games being the biggest examples) can get by with dbus, udev, xorg/wayland, and SDL. Other libraries such as libav, libpng, and so on are rather stable if you target them correctly.
If we want to point the fingers at anyone, it's the closed sources guys modifying the libraries that they use and expecting every distro to use their modifications. In the same breath, Ubuntu is one of the worst distros to target *because* they modify the libraries so much. The rest of Linux land doesn't suffer all that much, if at all. Build your shit right and it'll work.
Trying to force GNU/Linux land to unify will destroy its software ecosystem. Hackers will move on and/or simply refuse to target the unified stack. I've heard of some developers actively putting code into their applications to exit early if it detects software that the author disagrees with. Things could get ugly fast.
Fuckin hypocrites. RH/Fedora is the outfit that gave us Poettering of PulseAudio and systemd infamy. And *they* think they're gonna call out anyone else after those debacles? Methinks they doth protest too much... somebody needs to go over to RH and tell them to STFU.
C|N>K
Unification would be the death of Linux. It's all snow, you not being able to shovel it because each snowflake is individual, is a different problem.
How is a package manage cross platform when they don't even know how to package apps for a Fedora distribution yet? One would think, possibly, that questions of how and where to put apps and their dependencies in the various filesystems would be the first thing you find out, no? Or does it just chuck everything into opt?
Patiently awaiting the APPS guy.
I predict it will be the SNAPPS guy this time.
I should use this sig to advertise my book ISBN-13 : 978-1501515132.
Unification is what linux desperately needs
No, Linux's strength comes from its diversity and flexibilty. Linux is the dominant operating system in the world right now, and it's because of its flexibility that it was able to easily be adopted on platforms as different as bionic/android, and SLURM/Linux.
If someone tried to enforce their system ideology on everyone, then these variations wouldn't be possible, and Windows phone might very well be the main competitor to iPhone (all the OEMs wanted that....or something, and Google hit first with Android).
"First they came for the slanderers and i said nothing."
Apart from the toxicity, aren't they likely to melt when you put them on the hob?
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
In Soviet Russia, apps guy awaits COWS!!!.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
I don't get why people consider Android to be Linux. It's nothing like traditional Linux distributions.
Sure, a customized version of the Linux kernel is used, but everything else typically associated with a Linux distribution has been thrown away. That's probably why Android is successful: the Linux kernel is marginalized to the point where almost nobody even realizes it's there, and the traditional GNU/Linux userland is thrown away and replaced with what's essentially proprietary software, even if the source code is available.
Android isn't successful because of Linux; it's successful despite using Linux. Android would be just as successful were it built around the, say, NetBSD kernel instead. It's not the kernel that makes Android useful; it's the custom userland software that Google has created that makes Android successful!
Android is as much "Linux" as Windows would be "Linux" were Microsoft to replace the Windows kernel with the Linux kernel. That is, it would not be "Linux" in any meaningful sense.
I read the article (I know, I know) and Adam's rant seems to be completely unconnected to Canonical's announcement earlier this week. He seems to be ranting about headlines he's read in _response_ to Canonical's announcement and not what the developers are actually saying.
Adam is countering claims that rpm and deb packages are dead, but Canonical doesn't seem to have any such plans. He's countering the idea snap packages have universal support, but Canonical does not apepar to be saying they have, only that they could with help from other distros.
This seems to be yet another example of Adam going bat crap crazy over something without a foot in reality. I honestly wonder why Red Hat employs him, he does far more harm to their brand than good.
Watching the folks who brought us systemd argue with the ones who are bringing is snap gives me warm fuzziest.
Mandatory XKCD: https://xkcd.com/927/
It's the absolute truth of everything, as shown by the past DECADES of various standardization attempts.
Meanwhile, the concept of snaps is broken and WRONG. The reason is quite simple. Let's take OpenSSL for example. A lot of software uses it, links to it. In a snap-only world, if you had a dozen snaps on your computer that used OpenSSL, each and every one would have to update their packages INDEPENDENTLY. You have zero control over your computer. You cannot pre-empt, you cannot patch before the upstream does, you're at the mercy of that huge blob becoming available for each snap.
And NOTHING stops closed-source commercial software vendors from shipping their software with BUNDLED libs, and/or statically linked, right now. Except one little thing. They don't care about less than 1% of the user base. Case in point: Steam.
So there is absolutely NOTHING in snaps, no benefit whatsoever, over the existing delivery methods. You want a centralized app container for all your distros? Tarballs. You can tarball your bundled and/or statically build application if that's what you want, even today. And guess what, EVERY *nix operating system on the planet supports those.
And good luck with the isolation/containerization part, until all the distros agree which one to use. LXC, LXD, docker format, rocket format, next-container-wonder-du-jour.
Do you know how many different versions of Linux run across a thousand server environment?
Yes. Our datacenter has many thousands of boxes, and we happen to run CentOS on our machines. At any time, there are at most two different versions of Linux across our datacenter. When we roll out updates, some boxes will obviously remain on the old version while others are upgraded until they are all the same version again.
If you're struggling with heterogeneous machines, maybe you should consider making your datacenter more homogeneous.
I solved the problem by switching to FreeBSD. Saves a lot of time not dealing with Linux kids reinventing an octagonal wheel every 3 months. Plus you get a first-class port of ZFS. I highly recommend giving it a try.
The imbecile responsible for the abomination that is systemd, is employed by Red Hat.
Fuck you systemd and fuck you Red Hat.
Diversity is something of a double-edged sword. It's great for people that have a very specific itch to scratch and of course the open source nature makes Linux great for customization, but I think the amount of fragmentation on Linux tends to dilute efforts at fixing some problems. So, apparently, Linux needs a new app delivery mechanism? How many does that make now? How many desktop environments are there? Everyone seems to keep heading off and make their own rather than fixing the existing ones.
That's a lot of duplicated and wasted effort. Granted, it ultimately provides more choice, but I think it also tends to make each of those choices much less polished than what it might otherwise be.
Irony: Agile development has too much intertia to be abandoned now.
They do this all the time. Hardly worthy of a response although I am glad to see a clarification regarding RH and snap. The lack of communication is amusing. However, let's not forget Red Hat *cough* systemd *cough*, spits on floor.
I am happy if any vendor intending to sell closed source software stays away from Linux. If you want closed source software, go for Windows or MacOS, and let them treat you like the puny, dependent consumer you are. I for one am happy without any single piece of commercial software. And I am happy with any technically challenged people to not use Linux.
you mean... sudoctl systemctl snapctl installctl windows10ctl
output of the command:
"active/dead: upgrade was ... 9:32 am"
to view the full message, run upgrade again using a 4k monitor.
lucm, indeed.
If unification means GNOME/GTK3, systemd, whatever the hell Ubuntu's UI department shat out recently...no thank you. We don't need core systems with ridiculously egotistical leads wrapping their tendrils around Linux. Linux is about choice and freedom; if we have to trample those to get more closed source programs, what the fuck is the point?
Unification is what linux desperately needs in order to make it possible for third party closed source vendors to target the platform. Otherwise they just make ubuntu binaries, and that's it.
Redhat/Fedora, SUSE/openSUS, Arch, Debian, and other major players don't seem to be impacted by this third party closed source vendor only targeting Ubuntu. If that were the case, there would be no need for Snap, Ubuntu would already be the default distro. More likely, Snap is just what it says it is - a software delivery targeted at mobile (which can also work with desktop). The problem is Ubuntu isn't a big enough player in the mobile sphere and probably even the desktop sphere. By promoting Snap as the unifier of linux, however, might just get enough user base for one of these software vendors to produce for linux. But rest assure, Canonical is not promoting Snap for the benefit of other distros, it is for their own benefit. They simply need the other distros to buy in to get a large enough base to attract attention.
Ironically, Snap isn't even needed for these vendors to support multiple distributions. They could always use SUSE/openSUSE's OBS and have automated builds for whatever platforms they so desired.
Here's the gist of the problem. Canonical gave up on the desktop a long time ago when they pursued mobile. The problem is they are too small of a player in mobile to get noticed and in the time they've been gone from the desktop innovation they were so known for, others have caught up and even passed them by. The desktop is still alive, but most opportunities for growth are in the enterprise. With linux, that battle is predominantly between Redhat and SUSE (and Fedora and openSUSE). The only way Ubuntu can get their foot in that door is if they can convince people to run their Snap packages, regardless of the distro used and then chip away at the installed base. Ultimately, Canonical, like many dot-coms, were focused on finding the next big thing while ignore the golden goose they had. In doing so, their golden egg has become a very tarnished brass.
It's claiming that it already *is*. And that it is "enabling a single binary package to work perfectly and securely on any Linux desktop, server, cloud or device", when it currently does nothing of the sort, because confinement is disabled on other distributions and cannot work effectively on X11 (which is still the default for every major distribution) in any case. That's a direct quote from the press release - note *present* tense. It does not say Snappy will some day "work perfectly and securely on any Linux desktop". It says it does so right now. This is patently *not true*.
For the record, Slashdot, while I *am* an RH employee and a Fedora QA team member, this was a personal post, as the first words of it explicitly claim. It's not posted on behalf of Fedora or RH and does not reflect the official position of either RH or Fedora.
This is a perfectly reasonable point of view, sure. Do you think the best way for a company to try to achieve unification behind their system is - before holding any meaningful discussions with other distributions - to issue a press release massively overstating their system's current capabilities and heavily implying it has already *achieved* unification? Don't you think it might be better to, oh, I don't know, actually talk to other distributions and try to achieve some sort of consensus? And be honest about what their system is currently actually capable of, and the challenges involved in making it a truly viable cross-platform solution?
Unification is what linux desperately needs in order to make it possible for third party closed source vendors to target the platform. Otherwise they just make ubuntu binaries, and that's it.
One might argue that third party closed source software is bad because its closed, but this is how the world works unfortunately, and linux won't get any hold on the desktop market if you can't even port your closed source application to it because each distro is its own special snowflake.
I'm not concerned about market share, but you are correct just for useability sake. When pressed for time, I often just use Wine and a Windows app rather than walk through the build process of non-repo apps.
The shepherds did so well protecting the flock that the sheep no longer believed that wolves existed.
In the old days, these guys were using the same distro as me and would also show up in the help forums.
Most of them finally moved on. Thank you systemd! Thank you RedHat.
Innovation creeps up in the FOSS world through other projects, NOT through iteratively improving one core.
Let me continue on your systemd example. You are completely right, innovation works this way. But once the new innovating product is there, and it is obviously superior, people will want to adopt it. Thanks to systemd you now have an unified way to specify services, and an unified way to start and stop them. It is good to have it.
I am welcome to any project which builds on systemd, and keeps the user {CLI,configuration} interface.
If upstream doesn't want your contributions, you no longer have the (realistic) option to fork.
Having trouble convincing one maintainer to add your contributions? Well, It'll be surely harder to convince multiple, if you want a feature inside the whole ecosystem.
In the context of package managers this might be binary diffing. You want the linux package managers to do binary diffing? You'll have to convince each and every little distro king to accept your patch.
The ability to pick and choose your software in Linux is a feature, not a bug.
That might be right, to some extent. But what is fucking wrong is that you have multiple different package managers, each with their own way to invoke them via CLI. Arch even has two, yaourt AND pacman!
And package managers are very simple things. Generally each distro does the same things, just calls them differently.
I do NOT think that this is good.
Yes, its good to have a choice between desktop environments, and between web browsers, or file managers. That's all great. But the amount of lack of unification just harms linux IMO.
Why does each distro need its own package managers? Why does each distro have to re-invent the wheel for this?
All we get from this is just bugs over bugs, and waiting for each package manager to implement the same improvements like binary diff downloading or displaying a progress bar on download.
In fact, with most distros you are missing the choice when its about the package manager. You have to take what the distro gives you.
Unification [...] destroys the ability to greatly change and innovate
The linux kernel has been constant target of innovation, and it has stayed monolithical all the time, without any forks.
Software should depend on as general a base as possible.
I do agree with that one: but many features are not available via abstraction layers, and then you suddenly need to add ugly desktop environment specific code if you ever wanted to implement that feature.
The main issue with "targeting" is itself. Software should depend on as general a base as possible.
So you want to give support to your users via phone. Should you be required to know how each and every distro does things?
In windows you can say "go to start, type 'abc', then click this entry then click yes then no then ok".
IDK about the software industry, but in other industries press releases often overstate the truth or make it easy for journalists to misinterpret them. Its not a good practice but it does happen.
I haven't read the press release myself so I can't really make a judgement on whether it was okay or not.
Maybe you don't like mark, but I think its awesome what he's doing, and that he invests his money into ubuntu instead of some yacht or an apartment in some skyscraper.
I'd say its okay to be pissed off by this considering that they didn't make any consensus beforehand, but IDK, for them it seems to be enough that you can port your application to fedora, not that fedora uses it as its main method.
Would fedora adopt it if nobody used it? Its the classical network problem. An app distribution method nobody has installed gets no adoption by app developers because nobody will be reached with it, and if there are no apps for some package manager, nobody will use it, let alone use it as their main package manager.
I am happy when people like you are run over by drunk drivers who proceed to smash into telephone poles. Classic win/win.
I think the amount of fragmentation on Linux tends to dilute efforts at fixing some problems.
Which problems specifically?
"First they came for the slanderers and i said nothing."
Only the client portion is free software. It only works with a proprietary, Canonical-run package repository. Canonical does not offer source code for the server aspect, and thus, does not offer the ability to create third-party servers. The entire system is subject to Canonical's walled garden.
Let's me start by saying that when you want to run a game, you're more pragmatic and the political/philosophical/meta-technical issues take a second or third seat. I.e. even if snap packages are "evil" game users want to play the game, and Steam is a bit evil for someone who played in the DOS / Windows 95 days (or 8/16bit before that) when you didn't have to log in to some tracking platform each and every time. It's still praised a lot still.
Games are distributed as .deb etc. packages, binaries or Steam. There is GoG but I'd wager it's not very well known. There is also source if you want to fail at compiling it, or fail at compiling dependencies.
If you're bored and want some games you might look into your package manager for distro provided games, it should be safer in a sense and they're there. Of course the games suck, but they are not in great numbers and they are outdated. Especially as users may be running Ubuntu LTS (and its spin off's) or Debian stable. For multiplayer games, an older version usually is a detriment. Emulators lack in quantity and quality.
End result even a pure open source game is easier to run on Windows, since you download and run setup.exe and that's all, latest and most usable version. (and no driver issue)
Hence if there were snap package of games and some place to browse them it probably would be a lot better/easier. .deb : I can't install a 600MB game or smaller at all, that's more than there's free space on the / partition. And running automatically out of / space is already what happens if you don't do apt-get clean in a while (or with useless older kernel versions). Thus .deb games are useless or only work in some cases.
A second technical reason that prevents even me from running games included in the distro or distributed as
IF snap games can be installed on the 40x to 100x bigger data storage partition, problem solved!
I mentioned one already, so let's use that as an example: desktops. The situation is a hell of a lot better than it was just five years ago, but my feeling is that if efforts weren't divided over a dozen different projects, it seems like a lot of these issues would have been fixed much quicker.
Like I said, it seems like it's a bit of a double edged sword. It's really great to have a choice in desktops, but along with that comes the reality that each of those desktops is far less polished than what you see on commercial OSes like Mac or Windows.
Irony: Agile development has too much intertia to be abandoned now.
It's not vaporware. It's just not what they said it was. ROS packages should work great with snap. It's just not portable to any other distrubution except Ubuntu.
Catkin also does most of the work, so it's not fair to call this a win for snappy.
This will most definitely be an Ubuntu first thing.
ok, that's a reasonable place to start, desktops certainly aren't as good as they could be.
Do you think the problem is that there is the division of efforts? If Gnome were suddenly the only desktop available, would that make it a better desktop? Conceivably it could also give the project leaders a power-trip, and lead them into the false belief that whatever decision they make is correct, for example. As another possible counter-point, do you think having a single project could cause things to end up like IE6......without competition and thus completely stagnant?
"First they came for the slanderers and i said nothing."
Yet the only times I've ever needed to install something, even if it was just a newer version of a package already part of the distribution it has resulted in half a day's work. Once it failed so completely that I ended up installing a different distro in a virtual machine solely to run the package in question until the updated package came out in my distribution. Or my least favourite, I offered to fix someone's Ubuntu distribution for a 6pack. If I'd have known that the guy installed a PPA that caused package incompatibilities and somehow got apt to the point where any attempt to fix it made it try and uninstall core packages of his distribution I would have asked for a case of beer instead.
Linux package management is much like blindly hitting next on an installation screen in windows without reading any of the text. When you get your software from a very reputable source that you trust things will go pretty well, even when you don't chances are you'll only have minor issues. But every so often you stray from the mainstream and someone has to spend hours cleaning up the resulting mess.
Will this replace apt? Doubtful. Is finding a pre-packaged dependency free alternative to traditional package management unnecessary? heck no.
I second that. Incompetent design and poor management are what hold the Linux desktop back, not dilution of effort.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
I'd certainly not suggest a *single* desktop would make things better. But there's a pretty big difference between *one* desktop and the many that are in development now. I mean, we've got Gnome, KDE (Plasma) as the big two for a while, but now also Unity, Cinnamon, MATE, LXQt, Xfce, Budgie, Pantheon... and probably more besides that. It feels to me that part of what's driving this is not an effort to improve, but to differentiate. There's nothing wrong with that, but on the other hand, it's not exactly helping drive things forward either.
Like you suggest, though, one great thing about having multiple projects is that you do have a choice. After Unity and Gnome 3, many users would seek alternatives like Plasma or Cinnamon.
Irony: Agile development has too much intertia to be abandoned now.
Let me continue on your systemd example. You are completely right, innovation works this way. But once the new innovating product is there, and it is obviously superior, people will want to adopt it. Thanks to systemd you now have an unified way to specify services, and an unified way to start and stop them. It is good to have it.
Thanks for the best laugh I've had all day. The systemd mess is definitely not an example of much of anything other than large egos and small programming ability. And lots of lying.
I think even if there were only a couple we'd still be in the same polish state. The real issue is that devs only want to work on the shiny new feature, not fix/maintain those that someone else wrote.
"Unification" is exactly what GNU/Linux *doesn't* need. Innovation creeps up in the FOSS world through other projects, NOT through iteratively improving one core.
Google has certainly done a pretty good job of taking Linux, creating a OS platform in Android and making that available for hardware manufacturers and developers to target. Are the GNU/Linux distributions so much more innovative than Android? Seems to me pretty much everything new that has come out has been almost universally panned as being crap: upstart, wayland, mir, systemD, snap, GNOME3, Unity, etc, etc ...
Whether you're running OpenRC, systemd, sysvinit, runit, or whatever, you're still GNU/Linux at the end of the day. The ability to pick and choose your software in Linux is a feature, not a bug. If you don't want configurability or choice, go use a BSD or literally any other OS.
Linux-based operating systems can have the same thing, there are even mature distributions that don't use systemd for example so why do you want to use the ones that use systemd rather than the ones that don't? What I mean is you want freedom of choice and there are hundreds of developers actively working on distributions that support that so maybe you should support them.
Trying to force GNU/Linux land to unify will destroy its software ecosystem. Hackers will move on and/or simply refuse to target the unified stack.
I don't think so, we haven't seen that happen on platforms like Windows, OS X, Android, iOS, etc ... why would it happen with GNU/Linux?
I've heard of some developers actively putting code into their applications to exit early if it detects software that the author disagrees with. Things could get ugly fast.
It it's free software you can just remove the check and redistribute it.
We already had a way to start things. Just drop a script into /etc/init.d and link to it from the various rc?.d. Systemd broke that. Currently systemd will start scripts too, so best bet is still a script if you want all the cases covered.
As for packaging, a lot of commercial user apps come as a tar file you unpack in /opt.
The real problem is in library versions, especially dealing with software compiled against the latest greatest bleedingest edge version.