6 ms·

Canonical turned a profit, back above 500 employees

143 points by marcodiego 5y ago

marcodiego 5y ago
I miss the days of exciting Ubuntu releases... the 6.06 release with Compiz... the 12.04 release, finally with mature unity... It is now full of boring releases, desktop software often is nowadays, and that is a good thing mostly because of the stability it brings and because it is finally maturing. Most recently increases are probably mainly due to cloud services, WSL and partnerships like the Dell and Lenovo ones, but the feeling of dedicated 'contributors for a cause' is still there. WSL may convince a set of users to stay on windows instead of converting, but it makes Linux more known and more used and that will only make the number of users increase at first. Long term, it is hard to predict. Linux on the desktop is now around 1-2% by some estimates. I don't think it will ever reach the market share of windows, I don't think it will even reach the market share of the 'previous windows release', but it may grow still. If it reaches something similar to MacOS, maybe 10% 15%, hardware vendors and service providers won't be able to ignore it anymore and that will bring a new cycle of adoption. After that, it will stabilize, but at what point is also not easy to predict. As whole, FLOSS is getting good enough in most aspects and needs. It doesn't needs to be 'the best' or 'the favorite among professionals', amateurs doing something outnumber professionals in most areas. Inkscape got good enough, audacity (or what ever fork replaces it) got good enough, krita and blender got good enough even among professionals, Libreoffice is good enough if you don't have to handle MSO-born documents. The situation continues to improve and gimp 3.0 will be better, 3.2 will likely be good enough even for professional photographers. The landscape seems for video amateurs, animators musicians... Around a decade ago, Linux on the desktop was adrift. Now, there are routes which will take some time to mature but seems like good bets: PipeWire will bring good enough latency like jack with the same features of PulseAudio for audio AND video; snaps took centralized software distribution, Flatpak took decentralized software distribution and AppImages took direct distribution... that alone made software distribution on linux much better; Wayland will fix remaining Xorg problems... Oomd will fix lockups when out of memory... Wherever there were serious problems, now there is a route. So, things are going in the direction of 'good enough' soon for the general case and most niche cases too. For me, it has been good enough for more than a decade.
- criddell 5y ago
  > Around a decade ago, Linux on the desktop was adrift. If you wanted to ship a commercial Linux desktop application today (ie something that can print, takes advantage of hardware video acceleration, handles multi-monitor / multi-dpi setups, can be remotely managed across a company, is distro agnostic, erc...) could you do it? I'm specifically thinking of native applications and not Electron or Java.
  - Nursie 5y ago
    Why not? I have a variety of pieces of commercial software on my debian workstation, including MS Teams. Most of them are snaps. What is the specific problem you think might prevent that?
    - criddell 5y ago
      Teams is Electron, is it not?
      - NikolaeVarius 5y ago
        No. Edge Webview
        ripley12 5y ago
        Not yet. WebView2 is not available on Mac+Linux yet (but that’s coming).
        criddell 5y ago
        That's basically the same thing. It's still a web app (which makes sense).
        la_fayette 5y ago
        Teams 1 is electron, teams 2 will apparently be based on edge webview...
      - 5y ago
k__ 5y ago
I used Ubuntu for years, but switched back to Windows for easy gaming. If I need a Linux for work, I run it in the cloud, which is usually good enough. Cloud gaming on the other hand is too sluggish for my taste.
- mlok 5y ago
  I just discovered this and it looks promising : https://news.ycombinator.com/item?id=27870399 https://news.ycombinator.com/item?id=27870399
  - k__ 5y ago
    Thanks!
owaislone 5y ago
Happy for them. Back in 2004/2005, Ubuntu was what got me into the Linux wonderland. It was an amazing time. Canonical's work on the desktop towards the end was so exciting for me personally. Unity as a concept even though just partially realized is still my favorite ever DE. To this day I install little tools on my mac that give me some Unity elements like super+{1,2..N} combo. Part of me still wishes they or someone else can come back and finish that vision. The original Unity design docs were so good but I can't find them anymore. Does anyone have access to them? They detailed things like workspaces, "windicators", HUD etc. It was an exciting time even though there were a lot of flamewars because of Gnome fans not being happy with Canonical doing their own thing.
- simion314 5y ago
  If I remember right Unity was based on user testing and not on some pretend designer preferences and weird philosophy.
  - owaislone 5y ago
    I don't know if it was based on user-testing but yes it did involve a lot of user-group studies where they'd bring in random people with no prior Linux experience and tell them to use Unity
mybrid 5y ago
It is interesting to consider what happened to Browser as OS market. It is also interesting to consider with what happened to OLPC, one laptop per child, initiative. If all one does is use a device as a web browser then it doesn't matter what that device is. Browser games are gaining traction. As a Linux developer I use Linux as my primary OS and have Windows 10 loaded as a VirtualBox VM but the inverse would be easily doable if my brain fell out of my head and I decided to become a Windows developer. Windows is trying really hard to walk the tight line of making Linux run well for server applications natively in Windows but not for a Linux desktop or web browser. What I'm getting at here is that how users think about these things could easily break a market-changing usage in the next ten years.
unnah 5y ago
$138M revenue isn't half bad. I didn't see a product breakdown on a quick skim of the financial report. Anyone have an idea which products they are actually making money from?
- ris 5y ago
  My complete guess is extended support for otherwise-EOL ubuntu releases.
- bryanlarsen 5y ago
  They get a cut for every Ubuntu image spun up on AWS. I imagine that's the bulk.

143 points by marcodiego 5y ago

I miss the days of exciting Ubuntu releases... the 6.06 release with Compiz... the 12.04 release, finally with mature unity... It is now full of boring releases, desktop software often is nowadays, and that is a good thing mostly because of the stability it brings and because it is finally maturing. Most recently increases are probably mainly due to cloud services, WSL and partnerships like the Dell and Lenovo ones, but the feeling of dedicated 'contributors for a cause' is still there. WSL may convince a set of users to stay on windows instead of converting, but it makes Linux more known and more used and that will only make the number of users increase at first. Long term, it is hard to predict. Linux on the desktop is now around 1-2% by some estimates. I don't think it will ever reach the market share of windows, I don't think it will even reach the market share of the 'previous windows release', but it may grow still. If it reaches something similar to MacOS, maybe 10% 15%, hardware vendors and service providers won't be able to ignore it anymore and that will bring a new cycle of adoption. After that, it will stabilize, but at what point is also not easy to predict. As whole, FLOSS is getting good enough in most aspects and needs. It doesn't needs to be 'the best' or 'the favorite among professionals', amateurs doing something outnumber professionals in most areas. Inkscape got good enough, audacity (or what ever fork replaces it) got good enough, krita and blender got good enough even among professionals, Libreoffice is good enough if you don't have to handle MSO-born documents. The situation continues to improve and gimp 3.0 will be better, 3.2 will likely be good enough even for professional photographers. The landscape seems for video amateurs, animators musicians... Around a decade ago, Linux on the desktop was adrift. Now, there are routes which will take some time to mature but seems like good bets: PipeWire will bring good enough latency like jack with the same features of PulseAudio for audio AND video; snaps took centralized software distribution, Flatpak took decentralized software distribution and AppImages took direct distribution... that alone made software distribution on linux much better; Wayland will fix remaining Xorg problems... Oomd will fix lockups when out of memory... Wherever there were serious problems, now there is a route. So, things are going in the direction of 'good enough' soon for the general case and most niche cases too. For me, it has been good enough for more than a decade.

criddell 5y ago
> Around a decade ago, Linux on the desktop was adrift. If you wanted to ship a commercial Linux desktop application today (ie something that can print, takes advantage of hardware video acceleration, handles multi-monitor / multi-dpi setups, can be remotely managed across a company, is distro agnostic, erc...) could you do it? I'm specifically thinking of native applications and not Electron or Java.
- Nursie 5y ago
  Why not? I have a variety of pieces of commercial software on my debian workstation, including MS Teams. Most of them are snaps. What is the specific problem you think might prevent that?
  - criddell 5y ago
    Teams is Electron, is it not?
    - NikolaeVarius 5y ago
      No. Edge Webview
      - ripley12 5y ago
        Not yet. WebView2 is not available on Mac+Linux yet (but that’s coming).
      - criddell 5y ago
        That's basically the same thing. It's still a web app (which makes sense).
      - la_fayette 5y ago
        Teams 1 is electron, teams 2 will apparently be based on edge webview...
    - 5y ago

k__ 5y ago

I used Ubuntu for years, but switched back to Windows for easy gaming. If I need a Linux for work, I run it in the cloud, which is usually good enough. Cloud gaming on the other hand is too sluggish for my taste.

mlok 5y ago
I just discovered this and it looks promising : https://news.ycombinator.com/item?id=27870399 https://news.ycombinator.com/item?id=27870399
- k__ 5y ago
  Thanks!

owaislone 5y ago

Happy for them. Back in 2004/2005, Ubuntu was what got me into the Linux wonderland. It was an amazing time. Canonical's work on the desktop towards the end was so exciting for me personally. Unity as a concept even though just partially realized is still my favorite ever DE. To this day I install little tools on my mac that give me some Unity elements like super+{1,2..N} combo. Part of me still wishes they or someone else can come back and finish that vision. The original Unity design docs were so good but I can't find them anymore. Does anyone have access to them? They detailed things like workspaces, "windicators", HUD etc. It was an exciting time even though there were a lot of flamewars because of Gnome fans not being happy with Canonical doing their own thing.

simion314 5y ago
If I remember right Unity was based on user testing and not on some pretend designer preferences and weird philosophy.
- owaislone 5y ago
  I don't know if it was based on user-testing but yes it did involve a lot of user-group studies where they'd bring in random people with no prior Linux experience and tell them to use Unity

mybrid 5y ago

It is interesting to consider what happened to Browser as OS market. It is also interesting to consider with what happened to OLPC, one laptop per child, initiative. If all one does is use a device as a web browser then it doesn't matter what that device is. Browser games are gaining traction. As a Linux developer I use Linux as my primary OS and have Windows 10 loaded as a VirtualBox VM but the inverse would be easily doable if my brain fell out of my head and I decided to become a Windows developer. Windows is trying really hard to walk the tight line of making Linux run well for server applications natively in Windows but not for a Linux desktop or web browser. What I'm getting at here is that how users think about these things could easily break a market-changing usage in the next ten years.

unnah 5y ago

$138M revenue isn't half bad. I didn't see a product breakdown on a quick skim of the financial report. Anyone have an idea which products they are actually making money from?

ris 5y ago
My complete guess is extended support for otherwise-EOL ubuntu releases.
bryanlarsen 5y ago
They get a cut for every Ubuntu image spun up on AWS. I imagine that's the bulk.