"The lack of contributions to the kernel project relative to other distros is worrying for anyone interested in the overall health of the Linux ecosystem. If the most popular Linux distro in the world has nothing to contribute, well, that doesn't bode well for the overall future."
"By Canonical's estimates, Ubuntu has roughly 90 percent of the Linux market."
Red Hat 2013 revenue: $1.3 billion
Canonical 2013 revenue: $66 million
Whether this relatively tiny company contributes to the kernel is not going to have much impact on the future. This "90% market share" comment is also questionable--90% of a market that will pay nothing for its software? 90% of zero is still zero.
I remember after a few iterations (06.10, Edgy Eft) it offered a really great minimal experience. Everything worked well, defaults were reasonable and it was uncluttered. Besides, old Gnome was quite nice and cozy.
With time, Ubuntu has tried to be all things to all people and it became a big crufty thing. Now if you want something minimal you need to use Arch, Slackware or similar.
OS X suffers from the same problem, I think. Tiger-Snow Leopard were simple and beautiful. Now, iOS-ification is taking them through a path I dislike.
I have great hopes about Nix-like systems (NixOS, GuixSD) as they allow simple and complex setups while keeping clutter under control.
I think the cruft coincided with a shift in mentality in the Linux world in general.
Quite a number of under the hood changes i see today seem to have originated with developers working directly or indirectly on Nokia's Maemo devices.
In November 2009 the N900 shipped, a clear break with the 770, N800 and N810.
early 2010 Canonical release a new theme for Gnome2, and a year later Unity.
And the same year as Ubuntu gets Unity, Gnome goes 3.0.
And further down the stack you have systemd (2010) that may well be better at home on a phone or similar usage device than a traditional server or desktop.
And then there is Wayland, that one of the principle developers i seem to recall was directly involved with Maemo development.
Never mind going from APM to ACPI (Torvalds have some choice words about the latter) and more recently from BIOS to UEFI.
There is also a changing of the guard going on somewhat. The old hats are hitting 50 and 60, and the people in their 20s and 30s taking over are bringing in experience from the web way of doing things.
Sometimes the criticism aimed at Canonical feels like a case of tall poppy syndrome or crabs in a bucket. I may not agree with everything they are doing but they are trying different things so more power to them!
That being said, I occasionally use Ubuntu as a development environment in VM. Don't mind Unity, however the search feels clunky.
Ubuntu with Gnome 2 was really a good user experience back then. Gnome 3 is all about the desktop/mobile convergence, yet nobody to this day uses Gnome 3 on a mobile. Was it worth it ? Sometimes, it may be better to narrow your niche instead of trying to be everything to everyone.
b) Motivating ideation is not a plan. It was never the case that 10x10, GNOME 4.0, 20x20, GNOME OS (proposed more than a decade earlier btw) were widely accepted goals of the project. They were thought experiments.
Judge open source projects on their activity and output, not the claims of non-participating bloggers.
With Micosoft Bash on Ubuntu on Windows, the number of users of Ubuntu/NT could eventually dwarf the number of users of Ubuntu/Linux. With Ubuntu already at the top of desktop Linux distros now, they will probably entrench themselves even further.
You probably haven't been to a .NET developer conference, but you would probably be very surprised to see the number of people there running macs. The main reason is that Developers love unix. You're right that this is about Azure, but it's also about Developers loving unix.
I use Ubuntu on my computers only because I can avoid the Unity insanity and just switch to a standard Gnome 3 experience. Gnome 3 is not Gnome 2 but still is more beareable than Unity.
The problem with Canonical is they have some kind of Steve Jobs syndrome: when they think they are right, they just disregard any feedback as change-resisting, perhaps just to follow the suggestion 3-4 years later.
It goes from NotifyOSD to the Unity Lenses, to the software updates pop-unders, to whoopsie, etc. You can see the pattern, innovating for the sake of change, when no change was needed. Whas it worth? What do they sell, the desktops? No, they sell server support.
Looking back, there has never been ever any gain for them coming from a contested new feature. Every time the community protested against something, it turned out to be a bad feature, or at least, something that provided no measurable gain or wide-acceptation.
The Linux community is like an atom. All the heavy stuff is in a small core of contributors, and the rest is a fuzzy shell flying around screaming into the aether. And there are lots of distros that are fully community driven, and they didn't get us Steam or pre-installed Linux computers, or whatever else.
It's valuable to have one that is stubborn, and it's valuable for that stubborn one to be big enough that it can survive its stubbornness.
The community, after all, pushed back on 3D-acceleration-by-default, on a store, on systemd, on pulseaudio, even on what-was-once-called-jockey.
Windows and Mac user here. My next PC will be Ubuntu. If I could run Android Desktop in parallel (i.e., accessible while Ubuntu is running), I'll be in heaven. Google -- hope you're listening.
"By Canonical's estimates, Ubuntu has roughly 90 percent of the Linux market."
Red Hat 2013 revenue: $1.3 billion
Canonical 2013 revenue: $66 million
Whether this relatively tiny company contributes to the kernel is not going to have much impact on the future. This "90% market share" comment is also questionable--90% of a market that will pay nothing for its software? 90% of zero is still zero.