Canonical Is At Around 437 Employees, Pulled In $99M While Still Operating At A Loss

[royce]

Varies depending where you work in the world. What sort of costs are those additionally covering? Health insurance?(that's not common where I'm from)

Employee NI contributions, indemnity insurance and a number of other things.

I wouldn't think each staff member is earning in the six figure range either, but apparently the 40-ish million estimate at that was way off since it's over 60 million spent on staff..

Software developers are very well paid in London as there's a massive talent shortfall. So yes.

Feel free to throw that figure on top of what is apparently 62 million to staff costs. Are they paying for that many consultants that the costs are drumming of millions annually? If so they might benefit from hiring more talent if there's a common set of expertise they were getting consulted from.

Contractors are very expensive as well, as they have to shoulder the cost of sick and holiday pay and any stretch of time without work in between contracts. That has a cost for companies who only talent for a certain period of time, while at the same time saving them money in the longer run. Crucially, contractors aren't employees and are counted as expenses.

Yeah sure, and I said I could think of some things that contribute to the cost but not anywhere near another 40+ million. Leasing offices isn't going to cost you tens of millions for a company that size per year :/

I worked at the same building canonical was at at the company that owned it. They were charging canonical a good £1m per annum back in 2012 or around that date.

Computers aren't something that'd be refreshed every year either, you buy the hardware and enjoy benefiting from it via asset depreciation when it comes to dealing with tax. Infrastructure wise, at their size and proficiency/expertise, I'd assume the staff were capable of doing that in a cost effective way as well, such that it shouldn't be costing in the tens of millions annually either...

Another is handled by accountants, which when you have multiple companies such as Canonical has, you can shuffle the debt/tax around. Operating at a loss can have it's advantages when it comes to tax, it may have been more advantageous to adjust financials based on how other parent/sibling companies and the like are dealing with their own financial years, claim the maximum tax benefits.

Being more than familiar with UK annual reports, canonical doesn't seem to be doing anything shifty to avoid taxes. Their numbers look pretty sane for a company that operates entirely in the UK.

[aht0]

Lol that's an employer paradise. Depending on nation, that % can go up to 50% or even more.

No shit. Social security tax here alone means 30% extra. Add some more taxes, it's closer to 60% than to 50%.

[Snaipersky]

All this talk of Debian and Manjaro, no love for opensuse leap. For all the love Linux gets, Valve's moneymakers are on Windows. Leap has shorter but similar release cycles now, is extremely stable, and with minimal effort on Valve's end, could be tailored to gaming: recent Mesa in an appimage (to isolate from system LLVM) fsync kernel, and Steam DE.

[v4vendetta]

This company will be successful while it's run by egomaniac Mark Shuttleworth.
There is not team spirit in the company but only things that devide employees.
Mark likes to humiliate his employees in public, plus most of the employees who were really contributing to company success were either fired back in April of 2017 or left on their own right afterwards.
If before April 2017 that was one of the best companies to work for it became the worst after April 2017.
Technology can be great but it's ultimately people who are responsible for the development, qa, delivery, support, sales, Project Management etc... with this kind of attitude and while Mark in charge Canonical is a Zombie, the Walking dead of Linux, Cloud, OpenStack and containers industry.

[skeevy420]

Michael

I'm unapproved and we're bickering over our opinions

[skeevy420]

You can't have a stable (stable in the ABI sense) and rolling distro. The two are opposite release and support models. A binary compiled for a rolling distro may not work 6 months later without recompiling (which is why partial upgrades are not supported on Arch). That is what is meant by distro stability, and is an example of unstability. Distro stability actually has nothing to do with how often a system crashes or not.

Ideally, that's irrelevant because the kernel isn't supposed to break userspace and it's on whoever develops the userspace to provide the stable environment and that is possible regardless if something is rolling or LTS. Again, ideally.

You also need points at when to do major stack changes.

When it's ready to leave testing. That's when. Those "points" are why Debian/Ubuntu fall behind and have retarded as hell KDE versions.

Out of date kernels/mesa is an Ubuntu/Debian-ism, Fedora uses a release model and still has up to date kernels and mesa. Fedora often has newer kernels/mesa than Manjaro as it's only a few days behind Arch.

Which is why I said Fedora would be a great 2nd choice. There are still a lot more Arch/Manjaro users chiming in to the various Valve GitHub repos and since it makes just as much sense to go with Fedora or Manjaro in regards to software versions, might as well cater to the community.

Valve has the resources to do their own debian-based OS where they'd be in charge of mesa, kernel etc.

Valve has the resources to do whatever they want. They could just as "easily" go with GhostBSD since it ain't that far behind Debian these days...Mesa, some kernel GPU drivers, some love to BSD's Wine, and rebase/update the Linux compatibility layer to the existing SteamOS and GhostBSD becomes the base for a gaming platform.

Manjaro imho is a crappy distro that claims to be Arch for beginners yet provides none of the benefits of Arch (kiss, close to upstream etc), and no the AUR is not a benefit of Arch because anyone can read the PKGBUILD files and replicate them on any distro.

Well, it's like you said above about the stable and rolling. They can either try to provide that stable environment in a KISS manner or keep it as close to upstream as possible. I think they do a good job at KISS -- easy to use GUI OS installer, GUI based update applications, easy to install proprietary drivers, easy to use kernel switcher and multiple kernels for realtime, low-latency, etc,.

For the best of both worlds, Manjaro "Silverblue" would be a nice OS to have. Install their latest stable ISO and all updates are provides as OSTree layers. That would be great -- OSTree, layers, read-only root, easy to revert system updates, access to the AUR and AUR layers so if we install mesa-git we can just omit the mesa-git layer when Hitman 2 starts crashing at random for a week and a half...I'm just saying that I could really make use of an OS like that...

Try to replicate Arch's Wine on Suse or Ubuntu. Oh yeah, Suse doesn't have some of the 32-bit or devel packages in their repositories that's needed to build their own software by their own instructions let alone what a PKGBUILD can provide and Ubuntu has libraries in different locations, uses different package names, and uses their own Ubuntu-specific Wine patches, all of which add extra layers of crud in the conversion process (look at issues surrounding Tk-Glitch's attempts at doing just that with Wine for multiple distributions).

[PackRat]

I bought a magazine in 2008 with hardy heron for $20 and have been useless whiny leech since then.

[Britoid]

I was basing that off their recent numbers where Arch/Manjaro are number 2 and 3 used by Steam Linux users (Ubuntu is 1 and LTS is 4 IIRC).

The problem with Debian/Ubuntu is release cycles and lots of out-of-date software in their repositories. Reading the Proton bug tracker and ProtonDB, a lot of issues are from people on Ubuntu/LTS and people on Arch/Manjaro are saying "what problems?" simply because we have newer Mesa versions, newer kernels, newer Vulkan libraries, etc that have had the underlying cause of that bug fixed. More bugs reported with actual logs, apitraces, etc seem to from the Arch/Manjaro crowd as well. I've done a couple and I know that a few other Arch/Manjaro Phoronix readers have too.

Arch makes no sense as a SteamOS base because they'd have to double up on what Manjaro has already done -- provide ease-of-use and noobiness to Arch.

Fedora would make a decent candidate, especially Silverblue, but it still has release cycles and those bring potential out-of-date scenarios.

Suse might work but Valve would have to put more work into it than most other mainstream distributions.

Based on what I'm seeing, Manjaro makes the most sense -- stable, yet rolling; new user and developer friendly; already a highly used Steam Linux distribution; it can directly piggyback off the Arch Handbook so documentation isn't that lacking; their community is friendly and helpful to new users; their community already posts bugs and is helping to make Proton better...and Manjaro just went Public so they can actually be bought and sold.

Manjaro checks off a lot of boxes Valve would want a newer SteamOS to have; and these days for Linux gamers, more boxes than Ubuntu or Debian. Fedora would be the #2 choice and Ubuntu #3. Obviously that's all my opinion...write an opinion or idea and some assholes around here take it like it's scripture and facts and run with it.

You can't have a stable (stable in the ABI sense) and rolling distro. The two are opposite release and support models. A binary compiled for a rolling distro may not work 6 months later without recompiling (which is why partial upgrades are not supported on Arch). That is what is meant by distro stability, and is an example of unstability. Distro stability actually has nothing to do with how often a system crashes or not.

You also need points at when to do major stack changes.

Out of date kernels/mesa is an Ubuntu/Debian-ism, Fedora uses a release model and still has up to date kernels and mesa. Fedora often has newer kernels/mesa than Manjaro as it's only a few days behind Arch.

Valve has the resources to do their own debian-based OS where they'd be in charge of mesa, kernel etc.

Manjaro imho is a crappy distro that claims to be Arch for beginners yet provides none of the benefits of Arch (kiss, close to upstream etc), and no the AUR is not a benefit of Arch because anyone can read the PKGBUILD files and replicate them on any distro.

[skeevy420]

Cloning Manjaro would make no sense. I don't see a reason for them to move off Debian.

It would make sense for some time in the future for SteamOS to be built with OStree though.

I was basing that off their recent numbers where Arch/Manjaro are number 2 and 3 used by Steam Linux users (Ubuntu is 1 and LTS is 4 IIRC).

The problem with Debian/Ubuntu is release cycles and lots of out-of-date software in their repositories. Reading the Proton bug tracker and ProtonDB, a lot of issues are from people on Ubuntu/LTS and people on Arch/Manjaro are saying "what problems?" simply because we have newer Mesa versions, newer kernels, newer Vulkan libraries, etc that have had the underlying cause of that bug fixed. More bugs reported with actual logs, apitraces, etc seem to from the Arch/Manjaro crowd as well. I've done a couple and I know that a few other Arch/Manjaro Phoronix readers have too.

Arch makes no sense as a SteamOS base because they'd have to double up on what Manjaro has already done -- provide ease-of-use and noobiness to Arch.

Fedora would make a decent candidate, especially Silverblue, but it still has release cycles and those bring potential out-of-date scenarios.

Suse might work but Valve would have to put more work into it than most other mainstream distributions.

Based on what I'm seeing, Manjaro makes the most sense -- stable, yet rolling; new user and developer friendly; already a highly used Steam Linux distribution; it can directly piggyback off the Arch Handbook so documentation isn't that lacking; their community is friendly and helpful to new users; their community already posts bugs and is helping to make Proton better...and Manjaro just went Public so they can actually be bought and sold.

Manjaro checks off a lot of boxes Valve would want a newer SteamOS to have; and these days for Linux gamers, more boxes than Ubuntu or Debian. Fedora would be the #2 choice and Ubuntu #3. Obviously that's all my opinion...write an opinion or idea and some assholes around here take it like it's scripture and facts and run with it.

[Britoid]

I'd imagine Valve either buying off, funding, or cloning Manjaro if they wanted to do a serious round 2 with Linux and SteamOS. No other Linux distribution makes sense..

Cloning Manjaro would make no sense. I don't see a reason for them to move off Debian.

It would make sense for some time in the future for SteamOS to be built with OStree though.