Slashdot Mirror
Shuttleworth Says Canonical Is Not Cash-Flow Positive
eldavojohn writes "Mark Shuttleworth, the millionaire bankroller who keeps Ubuntu going strong, has revealed 'Canonical is not cash-flow positive' just as version 8.10 (Intrepid Ibex) of the popular Linux distribution is released today. In a call, he said he 'had no objection' in funding Canonical for another three to five years. He did say, however, that if they concentrated on the server edition of Ubuntu that they could be profitable in two years."
Red Hat itself has made it public that the desktop market is a very difficult one. Ubuntu has made very decent inroads to the desktop market for Linux, but it is true they need to put much more effort on the server side to become truly competitive. I think they have done some good work, but look forward to see what the community can provide in the next couple years. It's very hard to start competing in a market that is already spoken for by a few big players.
They are late to the party, and while I am glad for the strides they have made, Novell and Red Hat can eat them for lunch with other tie ins with their product line.
To me Linux has never been profitable in the Desktop-User side, but in the Servers Side. How can one make profit in the desktop world? Free software is mostly based on services not software license selling and it's not only libre but gratis (free as beer).
Linux (Ubuntu) has become really easy to use, and Linux users are mostly advanced users which can take care of themselves rather than paying for support, of for another service. And nowadays, most services are platform independent, IMHO.
Mr. Shuttleworth is truly praise-"worthy" (forgive the pun) because he's willing to put his money where his mouth is, and pay out of pocket to support his principles.
In the end, nothing is actually "free". While people can and do put in their time, without expecting to be compensated for their work on the various Linux distributions, or other open-source software, they do so because they have other jobs that support them financially. As the Linux desktop market expands, there will be a need for even more people to dedicate even more time to maintaining and perfecting the codebase... and this will require a positive cash flow into the industry. One way or the other we (the consumers of these wonderful products) are going to have to pay... and we shouldn't be apprehensive about it. I have no problem with paying let's say $50/year for Ubuntu, because it has worked great for me.
No, they don't want to concentrate on the server.
From the summary (emphasis mine):
A hypothetical does not a fact make.
A great many people think they are thinking when they are merely rearranging their prejudices. -- William James
This is a great idea. Supporting Debian will give them the support revenue, and eliminate all the development costs associated with maintaining their own derivative distro. They'd also be strengthening the Debian community, which is the underlying reason Ubuntu can exist in the first place. Ubuntu hasn't the resources to duplicate even a fraction of Debian's activity, so they serve both themselves better and the Debian community by simply supporting Debian stable and, if they *really* want, maintaining a custom patch set for whatever changes they may want (different process scheduler or whatnot).
I never understood why they needed or even wanted to create their own server distro when Debian stable is a rock solid, well known, highly regarded distro that they could profit from by supporting the existing users rather than trying to create a server user community of their own by convincing sysadmins (who are very hard to change by the way) to use their own, new, shiny distro that is untested and unproven, especially when compared to the likes of Debian stable.
Dumb move from Canonical, IMHO, and it smacks of the NIH (not invented here) mindset.
I hate printers.
I think you are missing the point here. With the GPL kind of open source software, there will always be a company who sees they can profit or at least not spend as much if they simply take available open source offerings and continue developing them instead of forking over some pile of cash to another company. In reality this is also what happened to the examples you mentioned, in a way - these companies did not suddenly create a new OSS product out of thin air, but started participating, bought other companies, employed developers etc. So while the basic idea of your thought has merit, in reality the situation will probably never really comeup where suddenly no company is behind the big OSS projects anymore...
"more current" in relation to Debian stable, maybe. In relation to the competition it is always subjective, given that RHEL/CentOS have 7+ year support lifetimes. I don't think anyone has done a "newness" and "correctness" metric for LTS vs. RHEL ... my guess is that they are about equal at GA.
This is hard to qualify statement, rpm is a super set of dpkg and it's hard to argue that yum is anything but a superset of apt-get (in terms of features, UI and speed). You could probably argue that Debian packaging is stricter than Fedora/RHEL/EPEL, mostly due to the above (which also means it's harder on the packager, but somewhat easier on the tools). Maybe you just mean that Debian/Ubuntu "offically support" apt-get dist-upgrade, whereas Fedora/RHEL/CentOS don't, yet, for various reasons ... which while valid is much less so in a real company setting, IMO.
I can only assume that you haven't used rpm/yum recently ... or that you have seen cases where bad external packages are imported into rpm case but not in the dpkg case (as the resulting dpkg hell is often much worse).
I can only assume this is some kind of weird joke, or maybe you are trolling. Ubuntu is infamous for kludging their kernels and not working upstream ... and personally if you are not running the distro. kernel on RHEL then you might as well set fire to your money instead.
ustr: Managed string API with ave. 44% overhead over strdup(), for 0-20B
I don't want to start an argument, but
Have you tried Red Hat, Fedora, CentOS lately?
Package Management through Yum, or the Package Manager is easy to use, works fine and is much easier than loading individual packages through Rpm and divining dependencies on your own.
I assume you problems with Rpm are with the package installation program and not the file format itself.
The weirdest problem I have had lately was uninstalling Samba ripped Nautilus off a system, and my Desktop icons disappeared. Reinstalling Nautilus fixed the problem, and also re-loaded some tiny piece of Samba it thinks it needs.
You've totally missed the point of the open source model. Linux doesn't *need* a profitable parent company. Projects like PostgreSQL, FreeBSD, the Linux kernel itself and others prove that companies are not needed in order to create excellent software. Debian existed long before Ubuntu, and will live long after it, should Ubuntu die. If Ubuntu dies, you can be damn sure a community will spring up to take the slack up now that demand for an apt based distro that isn't 3 years behind has been proven and an appetite created.
While OSS certainly produces good software; in many ways it's a self licking ice cream cone - it can create a self sustaining community that finds it satisfies their needs but has trouble moving beyond that into the mainstream. They are often content in copying the features they find useful in closed source commercial products but see no need to really innovate.
As a result, many OSS projects remain somewhat quirky copies of existing commercial products; with just enough differences to prevent them from being more widely adopted. They simply are not better than the commercial products (other than being free) - so people simply stick with what they currently use or what is the "standard."
Companies, OTOH, provide direction and assess customer needs to drive features - which requires some degree of control and expertise beyond coding. Would OO have gotten to where it is today without Sun? Maybe, but it doesn't hurt to have cash and direction to spur development.
The other issue with communities is that people will eventually lose interest and move on; or decide they don't like the direction and fork development. In the former that eventually leads to orphaned projects with promise (GIMPShop anyone?) and in the latter confusion in the broader market over which one to use.
OSS development is great, but there are some fundamental issues that hinder wider adoption of it.
As for the impossibility of Linux profitability, Red Hat's financial statements show a consistent, increasing profit, quarter over quarter, for the last 2 years. Go troll elsewhere please.
Yes, they realized that the money is in consulting and services; Linux is merely the road in. Not a bad model; and one that many Windows consultants use as well.
I'm a consultant - I convert gibberish into cash-flow.
I'm a network engineer, like a lot of Slashdotters here. I focus on Ubuntu & LTSP in educational type environments.
I would *gladly* pay Canonical for upper-tier support, if it were affordable to me, the small-business. As of right now, Canonical support services offers server support (which includes LTSP servers) for $750/year, PER SERVER - and this is just 9-5, weekday only, 10 "cases/issues" maximum, support. This is pretty difficult for me, as one of my clients is a 7-site elementary school district, which have all migrated to Ubuntu and LTSP. That would be US $5,250 a year. It seems that you can't span the 10 support cases over different servers, which is one of the reasons why this support model is so unattractive to me.
It's amazing how much LTSP has developed over the past few years, but there are still tons of things that can be improved, with a little TLC and bugfixing. As it is now, I am very active in helping report and troubleshoot bugs - but again, I want support from Canonical because IANAP, and they employ people who work directly on LTSP in Ubuntu. I've heard straight from them that they just don't have enough time to work on it - and it's a shame, given the number of people with LTSP up and running. If the support model was a bit more flexible for us smaller tech businesses (usually the ones who push Linux in the first place), I think Canonical could be incredibly successful.
It is pitch black. You are likely to be eaten by a grue.
>You only get 18 months of support for those.
>Debian stable will always give you more than that.
First, you can't get official commercial grade support Debian for stable at all. Second, even if you could, the LTS in the average lasts longer than Debian stable usually does.
Not only are Debians unpredictable releases a disadvantage compared to Ubuntu LTS, but even the community grade support you _can_ get for a stable does not last long enough to compare with Canonicals LTS.
Ubuntu beats Debian on polish, predictablity, support quality and suport longevity. Debian should just accept the fact that they've become some kind of a headless living organ donor for Ubuntu, which, although important regarding that single aspect, has little to none purpose of existance on their own merit any more.
I think you've totally missed what's been driving Linux progress for the last few years. Money. Lots of it. Corporate money paying developers. Virtually every single successful open source project has large corporate backing of some sort, be it Apache, the kernel, Firefox, mysql, etc..
Without a profitable parent company, they can't afford to pay those developers, and thus paid development goes away, and then you're left with the snail pace of "in my spare time" development. You're also stuck with the "only doing what scratches my itch" development, and many of the finer fit and polish elements that have gone into Ubuntu and other projects would be hard to find.
Would these projects die? No, but they would greatly slow down, possibly to the point that the majority of users would give up waiting for them.
If you need web hosting, you could do worse than here