No Juju for you! Ubuntu's Not Invented Here syndrome

Today Brian emailed me to share his enthusiasm for the Ubuntu Juju project, developed by Canonical, the company that makes Ubuntu.

Brian is a good friend that has been advising us on all matters TurnKey practically since the project began. His advice and feedback is always well informed and insightful so even when I already have my own opinions on the matter, I still take the time to look into his suggestions carefully. Thanks Brian!

This time, Brian wrote in to share that he's been enjoying his (impressive) Juju experience and sent a few links for us to look at. He also asked:

Have you guys ever thought of creating Juju Charm's for all of the TurnKeyLinux apps?

The first thing I looked at was whether I could use Juju without using Ubuntu. Not really, and that's a major dealbreaker because TurnKey is based on Debian. It used to be based on Ubuntu but a few years after we started TurnKey it became increasingly clear that we made the wrong decision. Debian was superior on so many levels: community, security, stability, packaging quality and most importantly - the fundamental driving values. So we bit the bullet and moved over to Debian in 2012.

I figured a somewhat expanded version of my answer to Brian could start an interesting discussion so I'm posting it to the blog. In a nutshell, I'm trying to explain why I think many in the free software community are not terribly enthusiastic about building on top of Canonical's work and why Ubuntu seems to have lost so much ground as the world's favorite Linux distro.

In 2008, when Alon and I started TurnKey, Ubuntu was at its height. Here are the Google Trends for Ubuntu since:

Ouch. What happened? My response to Brian tells a small part of this story.

Brian, thanks for prompting me to take another look at Juju today. We are evaluating several directions for TurnKey 14, which we will be re-engineering to work as a collection of modular services built on top of Core rather than monolithic system images. We're going to try and avoid reinventing the wheel as much as possible by leveraging the best components.

Juju is an option but to be honest it's probably not the leading horse in the race, and sadly that has more to do with the track record of the company backing it then any technical fault. In the context of the free software community, getting the answers right at the technical level is almost never enough. Collaborating successfully with the broader ecosystem and winning over hearts and minds matters. A lot.

At this point, Juju doesn't seem to support Debian at all. Debian have even removed the Juju client from sid for some reason. Not sure what the story behind that is. Given the growing divergence between Ubuntu and Debian, we can't expect to be able to leverage the Juju Ubuntu charms without some serious forking.

More importantly, we don't want to back the wrong horse. Canonical have a bad case of not-invented-here syndrome and a tendency to not really listen to the community. They're like the Apple of the FLOSS world except that Shuttleworth is no Steve Jobs and I mean that both in a good way (not as much of an asshole) and a bad way (not as good a leader/visionary).

Brian responded by defending Canonical and explaining that from his perspective working with the world's largest service providers Canonical was making impressive in-roads, especially in the enterprise and cloud arena.

Brian is the expert here so I'm in no position to argue, and to be honest rereading the email I sent him it did come off as a bit more anti-Canonical / Ubuntu than I intended. But my main point wasn't that Canonical is a bad company or that Ubuntu sucks, just that what happens in Ubuntu stays in Ubuntu. Maybe that's great for Canonical in the Enterprise space, but it makes building on their work a shaky proposition.

Boldly going where no man wants to go after

Canonical has a special talent for either backing the wrong horse, or breeding it.  A few examples of Canonical's track record:

  • UEC vs OpenStack
  • Bazaar vs Git
  • Upstart vs systemd
  • Launchpad vs github
  • Unity vs gnome
  • mir vs wayland

Given this track record, a Canonical backed project is an unlikely winner in any race for widespread adoption. You'd think they would win some battles just by chance. What's going on? 

My pet theory is that it has to be a mix of reasons: They don't listen. They don't inspire. They don't make the best stuff. They don't have the best people. They don't have the most money or the best business.

They do good work, and provide nice solutions, but for some reason we never seem to see those solutions adopted outside of Ubuntu by the wider Linux community. If you aren't already in the Ubuntu camp it seems short-sighted to back their projects. 

I don't think that Canonical is bad at what it does. It's just that they're rarely the best and being mediocre (or even second best) isn't good enough when the tournament effect is at work. The winner takes home the pot (e.g., becomes the new standard) and Canonical isn't winning.

I'm not even sure they want to. I mean, does Apple want Firewire to become a standard? But Apple can afford to create its own standards. Can Canonical?

If companies were text editors, Canonical would be Emacs

Canonical is not a company driven by the Unix philosophy of doing one thing and doing it well. If companies were text editors, Canonical would be Emacs.

It's easy to lose count of the many different  directions they seem to be trying to go in at once: Ubuntu Desktop, Ubuntu Server, Ubuntu Cloud, Ubuntu Phone, Ubuntu Tablet, and Ubuntu TV. Oh my! I'm waiting for them to announce the Ubuntu gaming system and Ubuntu car. 

I'm impressed (and slightly fearful) by the way companies like Google have expanded their business, but Google waited until they were wildly profitable with their core product to do that. I'm no expert but having your fingers in so many pies when your company is still losing money a decade after its creation doesn't seem like sound business strategy.

And then there are the various community antagonizing fiascos that left me wondering how they didn't see it coming:

  • Sending Unity searches to Canonical (they've since fixed that)
  • Inserting Amazon product referrals into the desktop experience (they've since made it opt-in)

Sure they came to their senses, but as the old saying goes: an ounce of prevention is worth a pound of cure.

How much does Canonical really care about free software values?

Here's another thing that bugs me. It's unclear how much Shuttleworth/Canonical genuinely cares about the underlying values of free software. From the outside it looks like Canonical is firmly rooted in the "commercial open source" camp as opposed to the "free software" camp (what's the difference?). This is reflected in a tendency towards technical isolation and the design of solutions that encourage dependence on Canonical services.

The focus is on utility and convenience, not values. And to clarify what I mean by that - a value is a principle you would hold onto even if you get penalized for it by the marketplace. If you give lip service to a value but are willing to give it up to make more money that's not a value - that's marketing.

I'm not saying Canonical's focus on convenience and utility are bad. It's just not inspiring. And you need to be inspiring to lead.

Still, they do a lot of good work and have done much to popularize free software. We should congratulate them for that and be thankful that Shuttleworth decided to invest his millions to create the company. There's definitely a useful place for a company like Canonical in the ecosystem. Ubuntu provides a gentler introduction to the sometimes harsh world of free software. It's especially useful to the vast majority of "human beings" who aren't aware that free software has anything else to offer beyond the magic of getting stuff for free. Who knows, some of them may eventually pull back the curtain.

But it takes more then being useful to lead and Canonical's take on free software is just not very inspiring for developers and would-be contributors, many of whom, like myself, do care deeply about values. What you do is important, but why you do it is even more important.

Free software is more than a better way to develop software, and more than a way to get stuff for free. Free software is about freedom. The more technologically dependent our society becomes, the more free software values matter because technology is a double edged sword. It can be used to strengthen our freedoms, or take them away.

We need utility as a measuring stick, and the right values as our compass. It's not one or the other. We need both.

Which reminds me of a pearl of wisdom I came across that keeps reverberating in my head:

Develop people, not products.

Update: it seems some Ubuntu people took my post very personally. See the clarification of the intention behind this post. No harm intended.


Pol's picture

Your Google Trends graph is nice, but I'll add this one:

I'll let you take your own conclusions ;)

Liraz Siri's picture

So Ubuntu as a brand is almost as strong as Linux and both have been in gradual decline in a period corresponding with the rise of post PC computing (mobile, tablets and the cloud) which is largely based on Linux under the hood. Interesting!

My take on this is that we're witnessing the increased centralization of technological power in an ever smaller priesthood. Linux use is now more widespread than ever, it's just that the people deciding to use it are not end-users but the priests that make the products.

I for one welcome our new technological overlords!

On second thought I take that back. I'm too afraid they'll turn into Morlocks. Eloi... delicious.

Jeff Spaleta's picture

It's really not a coincidence.... Android is the commodity OS in the post-desktop device form factors.
watches,tvs,settop boxes,tablets,phones... Android is there and it is linux.

Canonical missed a very critical window of opportunity in capturing interest in ARM OEMs. They did try.. they were part of linaro for a while..serious talk about Ubuntu on Arm back in the day... but Android landed...and basically blew them right out of the water very very quickly with a coherent OEM and developer story.

And its a little bit sad, because ubuntu is widely used in a commodity role all over the place. Hosters, ISPs, OEMs, startup web service companies... all make heavy use of Ubuntu to reduce their own costs, while charging customers for services or products. Credit where credit is due. Canonical's investment in Ubuntu has been a watershed commodity to unlock other people's entrepreneurial spirit and make it possible for people to launch web services super cheaply. But Canonical just can't make the key partnership relationship so they can get their share of the revenue being generated from that investment in Ubuntu. The revenue just isn't trickling back.

Android is itself a strong advertising and aggregate user data broker revenue stream for Google... making Android a commodity OS offering helps Google's revenue model. its more ads in front of more people. Its more search and usage data from more people that Google can sell.  Android is a loss leader model for Google. They put resources into it, and give it away and they really do get a multiplier benefit from the give away to bolster the core revenue model around users. And that model is not in servicing OEMs with support. There are only a handful of OEMs... their are millions of Android users. You don't build revenue around the few OEMs... hoping for big support contracts that will never actually happen. Squeezing companies with sysadmins savvy enough to choose Ubuntu in the first place.  No you scrape nickels and dimes from those users who don't really care about the OS and you profit.

But Canonical doesn't have anything close to a coherent advertising or user data broker revenue model at all.  So for every single dollar Canonical spends on Ubuntu is a dollar someone else is benefiting from...there's nearly no trickle back at all.  And the same goes with Juju and cloud right now to. What exactly is the revenue model for juju?  The real money with Ubuntu in the cloud is as a cloud provider. The value-add for support services are with companies like Rackspace and Amazon and Digital Ocean... the people end-users contract with to spin up vm instances or containers. Users don't contract with Canonical directly for the support. They pay cash money with those providers and build relationships with those providers. So when those providers offer value-add services on top of provisioning...users are going to go with them... not Canonical. And those providers know the cloud foundations much better than Canonical does. Canonical isn't going to compete for openstack service dollars against rackspace.. are you kidding me? Canonical's channel partners cut Canonical out of the customer relationship nearly entirely and it hurts Canonical's bottomline immensely.

You can't even realistically call Canonical's approach a loss leader model. It's just a loss model. Canonical simply does not have any revenue model right now that focuses on generating directly from end-users.  Its at best mid-sized organizational focused support. But only so many of those organizations are going to actually require services. Yes they'll win some contracts, but enough... never enough... and not the really lucrative contracts. They can't compete in government or fortune 500 or financial for mission critical systems where the security certifications that RHEL and SUSE can demonstrate can matter. Or meet the terribly complex purchasing requirements of the US fed government for example.  That's not to say Ubuntu won't be used inside those places... but Canonical's value-add services just can't be purchased in those organizations. So Ubuntu sneaks in at the edges displacing RHEL and SUSE a little bit but with no revenue upside for Canonical. Is that really a win for Canonical long term? 

And as cloud becomes more and more central to midsized organization...the cloud providers cut out more and more opportunity for Canonical to create those customer relationships and sell value-add service to end-users. What is Canonical's business model for monetizing the cloud and juju and maas and all that jazz? I can't fathom it.
Juju might be the greatest technology since the eletric nose hair trimmer...but if Canonical can't generate revenue from it... it just looks like another growing expense with no every Canonical led project.  And the blasted Canonical CLA still applies to juju. That CLA continues to be a chilling effect on external contribution for Canonical projets. Pretty sure that's the whole point of the CLA... to retard adoption of Canonical tech in competing distributions. It's just terrible.

And Canonical is now trying to conjule OEMs to buy engineering support for devices but OEMs are even worse than the cloud providers.  OEMs are much more cost sensitive as they are under immense pressure to control costs to field competitive hardware and produce new hardware on a yearly cadence. Canonical's approach to OEMs hasn't really worked out really well with Valve. Yes, Ubuntu is great to prototype on...but once you want to go product...and Canonical wants its arguably fair cut of the product revenue...people ditch Ubuntu and re-engineer on Debian.  Ubuntu loses a lot of its appeal as a piece of commodity tech once Canonical wants to squeeze you for revenue.

And the Uphone stuff is such a weird mash up of Android and Canonical in house projects. I really don't know if they have their head around it all internally inside the Canonical fenceline.  It will be interesting to see if the first phones are as good as say the first gen Android phones, or the FireFoxOS 1.0 phone. I don't expect it to be as good an experience as its been hyped, thats for sure. The real test will be if Canonical can do well enough to be get a chance at a 2nd gen phone.


Liraz Siri's picture

It's interesting that Android, "the Linux that crossed the chasm" as you put it, is so weakly associated with the "Linux" brand. I'm guessing that's because the target audience for Android is OEMs to whom the price and utility is the main (only?) selling point.

End users are not the target and those that do care can install the more open cyanogenmod if their phone supports it. I really like knowing that I (and others) have the option to roll my own version of Android, though I haven't taken advantage of it yet, except maybe occasionally to install an alt kernel that roots my phone.

Fred's picture

From Google re. these trend graphs: "A downward trending line means that a search term's popularity is decreasing. It doesn't mean that the absolute, or total, number of searches for that term is decreasing.".

Just because more and more netizens are searching for lolcatz, it doesn't mean searches for Ubuntu/Linux aren't also increasing (at a slower rate). One can't tell either way from the graph, due to lack of absolute numbers.

Liraz Siri's picture

Are you sure search trends are relative to the number of absolute searched? Because if that's true then we can expect any search term used more widely by technologically savvy early adopters to gradually decrease over time as greater swathes of mainstream society comes online.

It also begs the question relative to what. To searches on in English? Because if it's true just in the naive sense of absolute number of searches then we should see all English search terms gradually drop in Google Trends as more of the world's non-English population comes online (assuming they use Google).

Also, most of the planet is still not connected to the Internet and are presumably dirt poor. As more and more of them come online we should see the dominance of the items they are interested in reflected in the gradual drop of search terms that mostly interest the comparatively rich first 2 billion Internet users.

heldeman's picture

Unity - Mir is the future. To work good and the way Mark's vision go no other system would work just right. I will eat your hat if the phones with Ubuntu are not major players in the next four to five years. Ask Meizu, people are already trying to book these phones on pre-order with them.

Liraz Siri's picture

With regards to Mir, from what I've read up on the subject the same goals could have been better accomplished by collaborating with the rest of the community on Wayland. I'm pretty sure we'll eventually see a rehash of the upstart debacle with Shuttleworth backing down and realizing he doesn't really want to play alone in his sandbox.

Anyhow, I really loved the concept of the Ubuntu phone and would probably buy one myself if it became available, though I would prefer to run Debian on it rather than Ubuntu, mainly because I really like having security updates to all packages rather than a small subset of them.

On the other hand with competitors like Google the likelihood of Canonical/Ubuntu becoming a major player in the mobile space is next to zero. Remember that this is a space Microsoft is having trouble competing in. Long term I see even Apple shrinking to a small up-scale minority of the market just like it had in PCs until Steve Jobs came back.

So with regards to the Ubuntu Phone I think the best we can hope for is that this initiative gets enough adoption that Canonical will be able to sustain it as a niche option for power users. I'll support that.

Former Canonical Employee's picture

Phones with Ubuntu cannot be major players without a strong app ecosystem that rivals Apple or Google's which Canonical cannot offer. Much of the people who work on Ubuntu for phone are bottom of the barrel in every field. Look at the Designers none of them come from Apple, Google, RIM, Palm etc. Nobody has heard of them they are all freshmen designers.

Most of the people who work at Canonical today honestly do not have much of a career before Canonical and while they could go work at bigger companies they are definately not medium or top tier candidates in their areas of expertise.

Jorge Castro's picture

I have been asking (and hoping) for someone to help us get Juju in a usable state in Debian for a while now. If anyone is interested in helping us out, please contact our mailing list.

Jeff Spaleta's picture

but back in May, Shuttleworth was quoted in the press as saying Juju for CentOS was going to be released.
Let me refresh your memory:

Jorge, Is that CentOS support released yet.... nearly 3 months _after_ announcing it ? is there an official released version of Juju that runs on CentOS? I just pulled the juju repo from github... I don't see any reference to CentOS outside of a planning readme!!!!  A readme that says maybe Centos and maybe Windows.

What the hell man. Another press grabbing vaporware announcement from Mark at a conference?!?

And you don't see why people outside of the Canonical fenceline aren't interested in trying to spend time integrating Canonical tech like juju? And how Mark poisons interest with this sort of vaporware announcement? What little communication there actually is about projects comes from the non-engineering side of the Canonical ship and they keep screwing up the announcements and implying Canonical is doing the integration work, when they aren't actually doing the integration work.  He sets terribly unrealistic expectations on what Canonical is actually going to be able to deliver without external contributors.

And speaking of external contributors. I'm sure the blasted CLA that Canonical still required on JuJu doesn't help at all with Debian.  I'm pretty sure that CLA requirement really stops a lot of external DDs from spending anytime at all touching Canonical code and trying to integrate it into debian correctly.  Who wants to maintain code where integration patches you send to upstream won't be accepted just based on their technical merits. That CLA is still the death of wide collaboration. Drop the CLA and you'll get your debian integration help. Just asking for people to help without actually dropping the pointless legal barrier to them helping is just going to frustrate everyone.

Juju might be the best tech in the world. It seems useful from my toying with it... on CentOS.. using the patches globocom spun up that I read about in your blog post.  But being a good tech isn't enough. It has to be a well managed project to attract contributors. And right now.. non technical aspects of juju make it a hard project to contribute to or even integrate.

And I have to say the last tarball release... that I pulled from Launchpad is just utterly confusing... a tarball of different code hosting directories...with no readme in the top level at all to explain wtf is going on in the tarball. Its not inspiring... its not welcoming to anyone who is pulling the code and trying to integrate it on a different distribution than Ubuntu.  The lack of a breadcrumb readme with recipe build from source instructions in the top level of that tarball says "This is Ubuntu only... we aren't interested in distros integrating this... move along...nothing to see here"

It's like juju as a project does everything it can to make it as difficult as possible for people to actually try to use the source code and extend it. It is AMAZING that globocom bothered at all to work pass the project brokenness and spin up their small patchset and hack in support for CentOS.  Too bad it seems like their patches haven't ben mainlined yet. Why is that? The CLA getting in the way?


Liraz Siri's picture

I remember we had to sign the CLA once when we submitted a contribution to Ubuntu way back. If free software developers were vampires, the CLA would be a necklace of garlic cloves.

I suppose there might be some obscure legal justification for it, but it sends the wrong message: we're calling the shots, we're looking out for number one, if you insist on contributing, you'll need to play by our rules.

Gabriel's picture

I know this is an old post, but for those just stumbling across it, I just want to put in my 2 cents.

Windows support has landed ever since version 1.21 of Juju. CentOS 7 support has just landed in "master" branch and will be made available with the next version of Juju (the code was contributed by Cloudbase who have a great relationship with Canonical).

We had a working PoC for Windows for a while. Turning that PoC into production code takes longer then people realize. However, we have been using it for a while now (ever since last year actually) and have developed a number of charms for windows (including nova-hyperv that integrates into the ubuntu cloud deployment).


If you want to give it a shot, feel free to grab the boilerplate and powershell helpers and start hacking:


You can always ask for help on IRC or the juju Mailing List.

Supported versions are:

Windows 8, Windows 8.1, Windows server 2012 (and R2), Windows Hyper-V server 2012 (and R2), CentOS 7.

Debian Jessie is also comming soon.




brian's picture

Liraz and I certainly exchanged a couple emails about ubuntu.  I would like to respond to some of his post here though.

The above writeup makes Ubuntu sound like a bunch of greedy closed source ogres.    I learned from my folks to personally always walk in the other guys shoes before I judge them.

In regards to UEC vs OpenStack

I have personal experience with this.   When UEC first appeared it was based on Eucalyptus because Eucalyptus gave the closest approximation to the THEN and NOW largest IaaS Cloud provider - AWS EC2.

When OpenStack was first announced it was OpenSourced by the creator's Rackspace and NASA.    OpenStack was originally built on Ubuntu ... most of it still is!   

Its no coincidence that the latest OpenStack (Icehouse) was released the week after the latest Ubuntu Long Term Support (LTS) was released.   Icehouse was based on Ubuntu.

In regards to Github...

What many people don't realize is that Github has had some serious issues regarding Open Source for a long time.

GitHub's Terms of Service does not require public software projects hosted on GitHub to meet the Open Source Definition.

Reading the above I can see where several years ago there may  have been a good reason for some companies to not use github.    Canonical and many companies use LaunchPad.    Its just a choice but its not evil.   

Ubuntu uses Launchpad and Launchpad's code is OpenSource (github's is not).

In regards to Canonical's Upstart vs SystemD

When Debian decided to switch to SystemD, Ubuntu which is derived from Debian also announced the switch to SystemD a week later... not sure why this was an example of Canonical/Ubuntu "going their own way" ?

A complaint was made that Canonical doesn't listen to their users.  

I say that is horsepucky... and the statement is refuted by the fact that later in Liraz's writeup he notes:

  • Sending Unity searches to Canonical (they've since fixed that)
  • Inserting Amazon product referrals into the desktop experience (they've since made it opt-in)

Unity vs Gnome

Canonical aslo listened to their users and in the Ubuntu repositories today you can install Gnome3, MATE, XFCE, LXDE and more for a desktop alternative to Unity if you want to use them?   No one is locked into Unity... matter of fact if you install multiple desktops Ubuntu lets you choose which one you want to use for a session on its login screen.

A Statement was made:

"Here's another thing that bugs me. It's unclear how much Shuttleworth/Canonical genuinely cares about the underlying values of free software. From the outside it looks like Canonical is firmly rooted in the "commercial open source" camp as opposed to the "free software" camp"


Canonical makes their money through consulting NOT selling software.    Canonical's Ubuntu, Juju, Juju-GUI, Openstack, LXC work, etc are ALL Open Source.   

Canonical does NOT charge for their Ubuntu Server or Cloud images and they provide the source for them.

How about Red Hat Enterprise Linux (RHEL)

  • What does RedHat charge for RHEL ?  yes ... thats how they made $1.5Billion last year.
  • Can you get the source for RHEL6/7 ? ... No.

How about SUSE Enterprise Linux? 

Note for RHEL and SUSE Enterprise those prices are PER Server PER year.

Some people argue about the amount of FOSS contributions RedHat makes vs Canonical.

I always try to put that in perspective.    RedHat is a company with $1.5B in profits.   Canonical is not as commercialized and best guesses (its a private company are that they break-even at best).

Redhat has over 5000 employees (not volunteers but paid employees)

Canonical has about 650 employees (not volunteers but paid employees).

So Redhat is almost 10X the number of employees.     I would suggest that is a logical explanation for why RedHat is such a contributor to code base.

Finally the last item on the list MIR vs Wayland.

I followed this closely over the years.  Canonical/Ubuntu originally were very involved with Wayland.   However, as Wayland's development dragged on and on it was becoming a critical blocker to some companies that were trying to advance their SW architectures.  

Canonical  decided to take control of their own destiny in this regard and build MIR.     Their requirements were different from most Linux distro's because Canonical had been targeting Ubuntu for:

  • desktop
  • tablet
  • phone
  • TV
  • automobile
  • Cloud

Canonical designed MIR to satisfy the above needs.   MIR is open source!   I fail to see why this is a problem given that every Distro at some time or another goes their own way with major software components.

Also... Mir will still let you run xorg apps.

Does anyone complain that there are so many Distro's with so many different Package Managers ...

  • apt-get
  • pacman
  • RPM
  • etc 

No... they use what works on their Distro.

Does anyone complain there are multiple Linux Security approaches?  No..

  • AppArmor
  • SeLinux
  • Smack 

So to wrap up... I don't pay much attention to all the harsh ubuntu articles I read just as I don't pay any attention to people that hit on say .. Arch for being to complicated for example.  What works for 1 person may not be another person's cup of tea.

I know Ubuntu is derived from Debian and I like Debian myself and use it at times.

According to DistroWach today over 70 Distro's are derived from Ubuntu

So in a sense they are also derived indirectly from Debian.  

I didn't know an initial inquiry as to whether Juju & Juju "Charms" would be a future possibility for TKL's applications would devolve into such an extended discussion about "Ubuntu vs the World" but its always good to get things out in the open.  


Liraz Siri's picture

Hi Brian, thanks for starting the discussion and for weighing in on it. I have to run out now for Friday dinner with the family so I won't be able to adequately respond to all of your comments in detail but I just want to point out again that my intention isn't to bash Ubuntu / Canonical but to "devolve" our personal email exchange into an open discussion. That can be a good thing. There's a reason they call it constructive criticism. Sadly I admit attempts to illuminate often produce more heat than light.

Note that my perspective is multi faceted. While I'm as tempted to oversimplify as much as the next guy, I make an effort to avoid false dichotomies (e.g., black and white, left and right, good and evil, right and wrong), and try to pay attention to the nuance and complexity. I know that doesn't prevent me from often being more wrong than right but I do try to seek contradicting opinions and information that will help balance out my prejudices.

Remember that none of us can fit the world into our heads. There's only room for a rough sketch. Hopefully the sketch hints at the subtleties rather than turning into a gross caricature.

For now I'll just address my main contention with your comment. I don't believe it is fair to characterize my view of Canonical / Ubuntu as "greedy ogres". That's a straw man:

Rereading the email I sent him it did come off as a bit more anti-Canonical / Ubuntu than I intended. But my main point wasn't that Canonical is a bad company or that Ubuntu sucks, just that what happens in Ubuntu stays in Ubuntu. Maybe that's great for Canonical in the Enterprise space, but it makes building on their work a shaky proposition.


I'm not saying Canonical's focus on convenience and utility are bad. It's just not inspiring. And you need to be inspiring to lead.

Still, they do a lot of good work and have done much to popularize free software. We should congratulate them for that and be thankful that Shuttleworth decided to invest his millions to create the company. There's definitely a useful place for a company like Canonical in the ecosystem. Ubuntu provides a gentler introduction to the sometimes harsh world of free software. It's especially useful to the vast majority of "human beings" who aren't aware that free software has anything else to offer beyond the magic of getting stuff for free. Who knows, some of them may eventually pull back the curtain.

Liraz Siri's picture

Brian, thanks for highlighting all the various ways Canonical and Ubuntu contribute to the ecosystem. We should congratulate them for that and we can show our support by giving them feedback, and something to think about. Diverse opinions can be useful in the same way diverse software is useful. You can pick and choose. You can mix and match.

I know there are some people out there that just hate Ubuntu for no good reason and perhaps you responded to that, but that was not my intention at all. I think we're all on the same side here. We're free software buddies, and buddies can give each other advice even when they might be wrong and even when that advice isn't something the other side wants to hear. Maybe they need to. Maybe it will do them some good. Or maybe they will just shrug it off for now but it'll stay in the back of their minds for a while and then click into place when the circumstances are right.

Who knows?

Should we self censor and not express ourselves when we think we might be able to add something to the discussion for fear of someone taking offense?

There's always room for improvement. Nobody's perfect. Criticizing Ubuntu / Canonical on certain things doesn't cancel out all the good they've done over the years or turn them into greedy ogres. Most of us have room in our heads for a more nuanced view.

Steve's picture

Source RPMs for RHEL are published here:

Trent Lloyd's picture

While some parts of your article have merit, your shotgun list of Canonical backing or breeding the wrong pony is at best ill informed, and at worst intentionally misleading.

Canonical was out there with a release of all all of these projects/platforms before their "ultimate" winner was released, and their solutions were generally good (not necessarily the best, but they were good.).  Yes they have backed or bred the wrong pony in a number of cases, but it has nothing to do with "not invented here".


UEC was released in Ubuntu 10.04, Openstack did not yet exist - it was formed in July 2010 and released it's first code towards the end of the year.  Ubuntu picked up on Openstack very early in in 2011 with a preview in 11.04 and a full release in 11.10.  They got in to the cloud platform early, then saw Openstack come together and changed directions very early, and earlier than many.  That seems to be paying off for them now, their Openstack-based cloud operations are a big part of their business and they are very present in the Openstack community in general.  (Amusingly Juju is also a big part of that).

Bazaar ('bzr') was initially released in March 2005, a month before git was released.  The 'baz' bazaar project which was it's predecessor (also canonical backed) started in 2004.  They were similar tools in many respects, canonical tried to fill that void early - git won out but fact is Canonical sunk it's teeth in monetarily to an open source solution early on.  

Upstart pre-dates systemd by *4 years* and was one of the earliest init system replacements that gained adoption, largely through Ubuntu but also in Enterprise Linux 6.  Debian's adoption of upstart or systemd was also a heated 50/50 split.  Now that Debian have adopted systemd (and most people are), Ubuntu have also changed.

Launchpad vs github is not even a fair comparison, they do completely different things.  Launchpad is a much more complex bug tracker, designed for efficiently tracking issues across multiple releases (e.g. in Ubuntu) and also synchronising issues across multiple projects -- on top of this it supports most of the Ubuntu development tools (blueprints and other collaboration), package building for Ubuntu, translation, and a number of other things.  

Yes, there is some perhaps argument for claims of NIH in the Unity/Gnome and Mir/Wayland space, however, Ubuntu has very clear goals to achieve in the unified mobile space which dictates their Mir direction, and the GNOME project is still not really in a stellar state even after GNOME 3.

Are there places they could collaborate better? Sure.  Did they spend precious time and money on developing open source projects, that were good solutions but ultimately did not win in the market? Yep.  Have they changed over to the "winning solution" in many cases? You bet.

Perhaps Juju is not the right solution for Turnkey Linux, but what does its availability in Debian have to do with that? Absolutely nothing.  If it was the best tool for the job, you could easily contribute or foster it's management in Debian so that you can make use of it.  If it's not the best tool for the job, then move on to the next one.  The jury is still out on systems management and service orchestration though, the clear winner in that space is not yet here.  


Canonical and Mark Shuttleworth have sunk a *lot* of money and energy into the linux and free software community.  Yes, he is trying to make a profitable company -- that is hard in the free software space.  Mark has been putting his money where his mouth is for 10 years with no signs of stopping.  They have done a lot of good work, and had some good success.

Personally, it makes me sad seeing so many projects heavily invested into falling to the wayside as you have rightly pointed out.  I hope that they find success in Ubuntu Cloud professional services (this seems to be their current primary business direction), otherwise there is a good chance we may see the end of Canonical and Ubuntu one day soon.  That thought makes me sad, and it should make you sad too.

Liraz Siri's picture

You're right. As pure specimens of "not invented here" syndrome, some of these projects are not the best examples.

I actually changed the title of the post a few times before publishing it and settled on the one I did not because it perfectly represented everything I was trying to say but because it was the least worst.

It's true that Canonical have in fact proven that they are not so deeply entrenched in the Not Invented Here mentality that they can't recognize their mistakes. They have, at least on a couple of important occasions (e.g., UEC and systemd).

The critical thing in my mind is that Canonical has a tendency (intentionally or not) of alienating would-be stakeholders. IMHO, that reduces their impact and limits their success. The CLA is an example. The message it sends: you are under my roof and you will play by my rules.

If Canonical was better at winning over hearts and minds then the horses they breed would becoming the winning horse more often because more people from a broader range of interests and companies would get engaged and invest themselves in helping Canonical make their horse the best.

Also, if they were better at collaborating with the community they would get more mileage out of participating with existing projects (e.g., Wayland) to further their goals for Ubuntu. So they wouldn't feel compelled to go it alone as then have to backtrack later and lose their investment when everybody else decides not to follow.

I think that's what's going to happen with Mir as well. They'll work on it, and it will get better but Wayland will get better faster and they will eventually ditch Mir for Wayland just like they're ditching upstart for systemd. Most of the energy that goes into developing Mir will then go down the drain.

If Ubuntu/Canonical was better at playing nicely with others I think they would be better off for it as would the entire ecosystem.

I don't criticize Canonical because I'm against the company or don't like Ubuntu. Quite the contrary. I'd like to see them stick around and prosper. I criticize as a friend. I may be wrong of course and everyone is entitled to their own opinion.

Liraz Siri's picture

I've received some private feedback that seems to miscontrue my intention with regards to this blog post. The intention is not to harm Ubuntu or Canonical. I wish them well. I don't view them as foes but as friends. I believe it's ok for friends not to agree on everything. It's ok to express an opinion and seek contrary points of view. If you're open minded, getting feedback from people who don't agree with you is a great way to expand and evolve your point of view. Friends can help each other do that.

I care for Ubuntu more than I care for other commercial Linux distros

I've been involved with Ubuntu and Canonical on many levels since almost the very beginning. Alon is an official Ubuntu member and has gone to UDS. Over the years both of us have had good relationships with their people. I've spoken with quite a few high-level people on the phone and met a couple in person. They were all very nice people. I even exchanged a few emails with Shuttleworth. All this time I've been ruminating and forming various pet theories as to why it is they do what they do and how the world reacts to that. I try not to get too attached to my pet theories, lest they become dogmas. I accept that I might often be more wrong than right, but that doesn't dissuade me from voicing an opinion.

I'm not seeking to single out Canonical as somehow worse than the other commercial Linux distros. Quite the contrary. I don't really care about the other distros. Not because I think they're bad either, I'm just not engaged and don't have a useful opinion to share one way or another.

Standing up for free software values

I don't have a right to expect anything will come out of a blog rant on this tiny corner of the Internet but if I could wave a wand and choose what kind of influence it would have I would choose Canonical to be a bit more collaborative with the rest of the community and put more of an effort into leading and inspiring by putting more of an emphasis on free software values.

By that I don't mean just the licensing, which they are respectful of. I mean promoting free software as a moral good. Not necessarily a moral absolute, but more than a better way to develop software. A better way to develop software is useful, and free software certainly delivers that, I just think it was a historic mistake to try to cover up the moral aspects with the utilitarian aspects to appease business interests. What's good for society can also be good for business. It doesn't have to be a choice.

20 years later "open source" has gained wide acceptance by the technical elite, and is silently powering the world's increasingly diverse array of computer systems, while "free software" values have not.

We are centralizing too much power in the hands of too few

Society is increasingly dependent on computer systems from which we are increasingly alienated (e.g., cloud SaaS, shrink wrapped technological appliances like tablets and mobile phones). Most of these systems are actually powered by free software under the hood but in a way that adheres to the letter rather than the spirit of the licensing. This centralizes more and more power in the hands of fewer and fewer people and organizations. I find that alarming because any student of history knows that power tends to corrupt and absolute power tends to corrupt absolutely.

Last year we discovered that abuses of this centralized power are already running wild. This is a threat to our freedoms. You could be a targetIt's not just terrorists they're after. Terrorism is a hot button they push to try and manipulate us through fear. Terrorism is a problem, especially where I live. But rampant abuses of power are a much bigger problem. We witnessed the terrifying consequences in the last century. That wasn't too long ago. Human nature hasn't changed.

The price of liberty is eternal vigilance.

This is as true today as it was in Thomas Jefferson's time. Maybe even more so given that technology is accelerating and pushing down the costs. Totalitarian states in the 21st century have technological tools of oppression that the authoritarians of the last century could have only dreamed of. Moore's law applies as much to mass warrantless surveillance as it does to genetics research. Technology is a tool and a tool in itself has no values. It's up to us to take a stand by developing free software, strong crypto and distributed systems to decentralize the distribution of power.

The values and spirit of free software have never been more relevant

Looking back, Stallman turned out to be right with respect to his argument with the people who backed the creation of the Open Source Initiative and led to the rebranding of free software as open source. The pragmatism was short sighted. 20 years later the post-PC shift in computing is eroding much of the power the open source licenses have to promote freedom. As a neutral utility open source has never been stronger but as a tool for promoting freedom "free" software is slowly becoming obsolete. With the body dying we need to strengthen the spirit. The licensing was always a means to an ends. The underlying values have not tarnished. If anything they are even more relevant than they were when the free software movement began.

Technology is transforming society at an ever accelerating rate and the technologists have a critical role to play in helping create free, open, distributed alternatives to massively centralized, tightly controlled opaque computing systems.

The path of least resistance may have a nasty surprise at its end. Focusing exclusively on utility and convenience is naive and dangerous. The free software movement has a role to play in balancing the scale by both educating society and providing the tools to prevent and fight oppression.

If you seek to help, join the open source community and fight to keep the spirit of the press alive and the Internet free.

- Edward Snowden

What does this have to do with Ubuntu?

Ubuntu is one of the best known examples of free software. I don't expect Shuttleworth to oust MBA Jane and install a free software zealot as Canonical's CEO but I think his company could a lot more good in the world and would ultimately be more successful if it stood for a higher purpose than useful software at a lower total cost of ownership. I know that kind of sense of purpose is in short supply amongst profit seeking corporations but I personally believe that making money should be a means to an ends rather than an ends to means.

We need more missionaries not mercenaries.

Former Canonical Employee's picture

I left Canonical this year and have seen many of my colleagues leave over the last three years all from different teams. The reason I left Canonical is it was really difficult to have a say in the direction of the area I supposedly managed.

Mark is a great guy and so was boss but at the end of the day there was a terrible amount of micro management and Mark always has his way on every detail.

A handful of the Juju team has already defected to companies like HP and Red Hat because the honest truth is they didn't see any future in the product they were working on. I think the only reason many people stick around is because the job is stable and Mark doesn't give up on things even if he is losing at it and Canonical pays a good salary.

Canonical could do so much better if a new CEO was appointed versus Jane who is really just a figurehead and if Mark would loosen up and let a CEO run the company like a real business with a profitability goal.

For me I couldn't continue working at a company that kept losing the race. I think Ubuntu will do fine on mobiles but I think the goal of competing with Android is grandious and unattainable.

Liraz Siri's picture

I think putting more of an emphasis on values could help Canonical with hiring as well, though it would require them to believe the values to the core and that would probably require a serious reorg of the company to expel people who don't really buy into this value system. You can't just let the marketing team communicate a message that goes against your true values because if mediocre utility is uninspiring, hypocrisy is a downright turn off.

Either that or you go the other way and just decide you're a business like any other and then you have to focus on only the things you can be the best at while making a healthy profit. In any case, standing for making useful, but second/third-best software at a lower total cost of ownership doesn't inspire the best people to come and work for you and it's very hard to compete in high-speed cut-throat fields like mobile if you can't get the best people. Especially when you're already the underdog and networking effects are working against you.

The people issue is the most critical in deciding what areas your company can compete with.

A relatively small company like Canonical can't compete with goliaths like Google or Apple for the best people. The best people are like volunteers. They can work anywhere and receive competitive compensation. Once money is off the table as motivation, a higher sense of purpose and job satisfaction kick in. Whatever you care about. If you care about winning, then you want to work for a winning company. If you care about making the world a better place, then you want to work for a company that makes a big impact in that way. If you care about becoming the best at what you do then you want to work for a company that really takes you out of your comfort zone and gives you the resources to develop yourself. Probably for the best people it's a combination of factors.

I suspect this is the reason companies like Google for example particular will not only outbid smaller companies for the best people they also put in enormous efforts to make their employees happy. They pay a lot of attention to the culture and put a lot of resources into tracking and quantifying happiness. It's really hard to compete with that.

If you just try to pour money on the problem your overhead is going to be higher and you're going to have a selection bias for mercenaries rather than missionaries. Also, if you're really smart then presumably you already know that deciding where you want to work based on who bids the highest is probably not a good idea. According to research beyond $75K a year (better than average in most parts of the US), additional money won't make you any happier. Once you realize that money itself becomes less of an issue. You need to be paid enough to feel secure and to take money off the table.

j openshare's picture

Agree, ubuntu seems unbalanced these days. But what do you think about debian. The OS is great and supports ARM stuff. Just too few install guides and tutorials. Other wise works everywhere from Raspberry to cheaplaptops..

Skandhas_Deleted's picture

I mean, does Apple want Firewire to become a standard?

Seriously? It's been a standard for 20 years:

1394 Trade Association


Liraz Siri's picture

There are multiple ways in which you can use the word standard. The one I meant: used or accepted as normal. It's one of the definitions that pops up if you Google "define:standard".

If you don't buy a computer from Apple, it usually doesn't come with Firewire. They're both standardized but only USB is a standard.

Carl Kiendl's picture

By your definition, SAS is not a standard. IPMI is not a standard. POSIX is not a standard.

The definition you are citing also belongs to the use of "standard" as an adjective, not to standard as a noun - "a standard connector" vs. "the USB standard".
One of Google's definitions for standard as a noun is "an idea or thing used as a measure, norm, or model in comparative evaluations". I am pretty sure IEEE 1394 is used as a norm in comparative evaluations of manufactured connectors.

Even beyond that, your entire argument hinges on artificially limiting the ecosystem looked at. Sure, if you limit your examples to only desktop computers purchased by end users not manufactured by Apple, then yes, it's unlikely they'll encounter FireWire.
But if I limit the sample set to "end users who have bought consumer laptops", then, suddenly, 7200 RPM hard drives are really, really uncommon. If I limit the sample set to "end users who have bought tablets", then hardware keyboards are definitely "not standard".

If I limit the sample set to "end users who have bought LED-LCD displays below 100 €", then suddenly, HDMI is no longer a standard.

I know this was all not fundamental to the point your were trying to make, but a chain is only as strong as its weakest link, and the FireWire argument is weak through and through.

Even in a hypothetical scenario where one would accept your definition and declare IEEE 1394 not a standard, it would not be the same: Apple didn't stubbornly create FireWire at a time where USB was already strongly entrenched, just to be different - quoth Wikipedia:

"Apple's development began in the late 1980s, later presented to the IEEE,[6] and was completed in January 1995."

"A group of seven companies began the development of USB in 1994 [...] Few USB devices made it to the market until USB 1.1 was released in August 1998, fixing problems identified in 1.0, mostly related to using hubs. USB 1.1 was the earliest revision that was widely adopted."

Apple tried to solve a problem at a time no solution existed. It just lost out to the cheaper-to-produce solution. Canonical is looking at all the existing solutions and saying "fuck it, we're rolling our own".

Apple and Canonical did fundamentally different things, so the comparison (and clinging to it) weakens your argument.

The LaunchPad vs. github comparison is also questionable - github is a rather technical place, whereas LaunchPad is geared towards end users and newbies. github focuses on technical development, on code improvement - LaunchPad on community- or product development.
Does that mean github isn't more successful? No. But it does mean that maybe they're not necessarily competing in the first place, at least not as much as you think.

Just yesterday, an open letter to the Linux world went viral, decrying systemd as a tool of centralization and control, fundamentally against the values of free software.
The author might argue not supporting systemd was the right move by Canonical.

I also can't help but notice that TKLBAM was written specifically for TKL ("TKG/L"?), and that it and the entire application development process were only opened up very recently, making you more open than Canonical today,'s not always been that way, has it?


I'm not trying to say you're wrong, I'm just trying to point out that the lines are by far not as clear as you're making them out to be, and that several of the "Canonical vs. the world" examples you picked are at least up for debate.

Overall, yes, Canonical definitely has an issue with distancing themselves from the community.
But if you look at the software landscape: Red Hat, Ubuntu, Android, Mac OS, iOS - the most successful POSIXes have always been the ones a bit distanced from their origins, a bit tighter in the grasp of one company.

Point being: What you and I interpret as "NIH syndrome", others may very well interpret as being a necessary step in being successful with a Linux distribution.

Ask yourself: How many tablets can you find in brick and mortar stores with

  • iOS preinstalled
  • Android preinstalled
  • Ubuntu preinstalled
  • Debian preinstalled


Why is this important? Because of Linux mythology. Because of the oft-cited "Year of the Linux desktop". For as long as I can think, year after year after year, Linux brethren all over the globe have hoped, cried and prayed for Microsoft's domination to be broken, and for Linux to take over as the dominant OS.

Computing is shifting to mobile. And Android dominates over MS-OSs.

My point is: The reason you are condemning Canonical for its lack openness is because you see it as a lack of freedom. But the only time the Linux community has actually achieved what it wanted on a grand scale -for a Linux OS to overtake a Microsoft OS- is also the first time a Linux distribution's development was almost entirely led by a single company, with little regard for the community's needs.

What you have outlined is idealistically right. But history has proven that it's strategically wrong. If open communities like Fedora's or Debian's could rule, it would have happened ten years ago. Instead, Fedora just got swallowed by Red Hat, and Debian's work is being used by others for their success, not for Debian's.

You said:

"a value is a principle you would hold onto even if you get penalized for it by the marketplace."

In order for FOSS to bring the wide-spread freedom you desire, FOSS itself must become more wide-spread. If you always get penalized, you are getting nowwhere.

It is exactly the Debian vs. Ubuntu relationship that exemplifies that: Debian is nice as a technological basis. But newcomers and end users went with Ubuntu. Why? Because the rejection of proprietary 3D drivers and the "obsolete on release" stability model are not what the people want.

You cannot free people against their will. the United States regularly try that and fail. In order to spread FOSS, you need adoption by the people at large.

And the only ones who achieved that so far are the ones with a one-company-less-community model.

To use an American idiom: You can't have your cake and eat it, too.


tl;dr: What canonical does is, from a FOSS perspective, idealistically wrong. But it is strategically right to achieve the overarching goal of spreading free software to the masses.

Liraz Siri's picture

First Carl, I'd like to thank you for posting the most in-depth analysis / response to my post yet! I enjoyed reading it. You've raised some good points. Sadly I'm running out soon so I won't have time to respond to all of them.

To clarify the tl;dr version of the point I was trying to make with the blog post is that Canonical / Ubuntu, and the free software community would benefit by:

  1. Focusing only on the things they can be best at. Canonical vs the world is not viable.
  2. Learning to play better with others rather than try to establish an edge for Ubuntu. In game theory terms they are playing the defect card in the prisoner's dilemma and other players have been giving them tit for tat by pretty much ignoring their work and making much of it irrelevant.
  3. Focusing on free software values so that Ubuntu stands for a higher purpose than the uninspiring de-facto goal of making useful solutions at a lower total cost of ownership.

Regarding the pragmatism of focusing on freedom

Yes, you can't force people to be free. They need to care about their freedoms. That's OK. True, Linux didn't break out into the desktop in a big way because the networking effects were too strongly in favour of Microsoft's dominance.

But it also didn't lose, it was biding its time.

The tiny niche of users that were using and developing Linux turned out to be much more influential than the average joe or grandmother that all the naysayers kept pointing out would never use Linux.

So today Linux and free software is almost everywhere. It powers most of the cloud, phones, tablets, televisions, my eInk Kindle, all 3 robots I have in my home etc.

Average joe and your grandmother are using Linux, they just don't know it.

Linux didn't need to win on the desktop. It was just there when the next shift in computing started happening and that was enough. Microsoft, the giant unstoppable behemoth is getting more and more irrelevant with each passing day.

If you look further back you can make an analogy with the scientific revolution. Science started spreading in the elites within the elites and it faced an uphill battle when it's findings clashed with religious doctrine. It took centuries for it to break out and even today the average person doesn't really get science. But that matters only up to a point, because the science literate segment of the population has much more of an influence over society than a count of hands would suggest.

Regarding your comments on TKLBAM

TKLBAM is poorly named. It is actually not specific to TurnKey and we added support for Debian and Ubuntu in the last version. It could be ported to other distributions and I would support that except that I don't really care about doing that myself so it would have to get picked up by someone else who does.

It's true that it is still used primarily with TurnKey and I accept that we are partly to blame for that. Poor naming convention and not enough of an effort into evangelizing that other distributions should use it, though I doubt that would have been enough. We're hoping to eventually get TKLBAM (and all other TurnKey specific components) into Debian but I think we'll need to sponsor a Debian Developer to do that since we're already stretched very thing as it is.

Also note that before developing TKLBAM I thoroughly researched all other available solutions and none of them could provide the ease of use I wanted in a backup solution. So I rolled up my sleeves and got to work, but if you look at the design you'll notice it leverages Duplicity for the backend storage. Reinventing the wheel serves no one.

Regarding your comments on the TurnKey development process

It's been open for a couple of years now but you're right in a sense because initially this wasn't the case. Instead of self-contained build system we had a loose confederation of scripts and build machines tied together by the programming equivalent of duct tape and stick-it notes.

In other words it was an ad-hoc process not software we could have distributed. I suppose you can also open up a process by documenting it thoroughly enough so that other people can reproduce it but it was such a mess we figured it would be less trouble to just scrap it and redesign the build process properly from scratch.

We did that at enormous cost. It took about a year or so of development. It was certainly not the path of least resistance and there was no short term payoff to motivate us. But we realized as long as we didn't do that we were isolating and alienating TurnKey from the free software community. Stop-gaps measures such as TKLPatch turned outsiders into second class citizens and there was no future in that.

Regarding what I meant by Standard

Semantics is tricky. Arguing over what a word means usually generates more heat than light. Language is inherently imprecise and ambiguous. Words are empty boxes where meaning is filled in by convention and a long process of evolving consensus.

That makes using language to make a point more of an art form than a science. When you need to make an argument that needs that kind of precision it's better not to use words at all but a more precise and less subjective form of expression like mathematics or code.

I realize there is some ambiguity in my use of standard so let me try and clarify what I meant a bit, which may be different from what I said or all the different ways in which what I said may be interpreted, for good or ill.

Let me start first with what I didn't mean. By standard I didn't mean ubiquitous or standardized. I meant a technology on which a range of diverse (possibly competing) interests have decided to cooperate towards their common rather than individual good.

As an example, FireWire is standardized but outside of the Apple tightly integrated ecosystem it is rarely used. If you don't buy a your computer from Apple or you don't buy a peripheral made for use with computer hardware made exclusively by Apple you're much less likely to run into it. FireWire may not have started life as a way to lock people into the Apple ecosystem, but that's how it works today. If you've already invested in all of these computer peripherals that work with FireWire, you're more likely to get your next computer from Apple, and vice versa. I don't think I'm going out on a limb by presuming that this is just the way Apple wants it. That's good for Apple, but it's probably bad for Apple's customers and for the industry as a whole.

Imagine if everybody decided to play the 'defect' instead of 'collaborate' card in the endless games of hi-tech prisoner's dilemma? That would serve no one. Alliances of competing players that decide to put aside their differences and collaborate within a framework that serves the common good won the PC wars and are winning the post-PC technological wars. There's a reason collaboration, even amongst potential competitors eventually evolves into the winning strategy. This is as true in biological evolution as it is in technological evolution. Apple's culture of overbearing control and selfishness dooms it to technology irrelevance in the long term. I didn't have to wait for the newer generation of Android phones to blow Apple out of the water to know that. It's just the inherent folly of the screw the world I'm going at it alone mentality.

Being inspired by Apple to make products with a really sleek UX is a good thing. But Shuttleworth seems to be 'inspired' by Apple in more then that regard and the whole does not play well with other type thinking that underlies many of their decisions is poisoning Ubuntu. Apple has a lot of money to burn until they run into trouble. Shuttleworth, with all of his millions, is like a homeless guy on the street in comparison.

Jeremy Davis's picture

I really enjoyed reading these posts. Great analysis and counter points Carl, and great follow up Liraz!

From my perspective Apple have had some great innovation, most of which IMO has been more about style though, rather than tech. From where I sit their tech side of things has been more of a way to 'lock in' customers (at least that's what it looks like from where I sit - hence why I've never been a big fan). I think that this is demonstrated by their move towards using 'standard' hardware components but packaged in an Apple way (i.e. physical appearance/UI/etc).

Apple under Jobs didn't appear to be too concerned about giving people what they wanted, but he had a clear vision of what he thought people needed (and it seems that many consumers thought he was right, or at least close enough). More importantly for the success of Apple; he was able to translate that vision to the rest of the company so his vision was realised. Whilst that heralded the high point of Apple's success, I think that the danger of that has been shown (now he's gone...). I think that it was probably inevitable anyway, but IMO certainly happened faster once he was gone...

Ubuntu/Canonical OTOH seem to have made some steps to create a 'unique' style and no doubt Shuttleworth has a vision of what he thinks people need too, but he doesn't seem to have the ability to translate that into the products; not cleanly anyway. I agree with Liraz that if Ubuntu/Canonical put more back into Debian (especially regarding the technical, nuts & bolts stuff) rather than 'going it alone' then perhaps they would have more success relative to effort.

Even though I'm not a big fan of Unity, I think that 'going it alone' with something like that (i.e. end user UI) is not a bad plan. Whereas with regard to underlaying technology that end users rarely care about (at least in the consumer space - as pointed out by Liraz) IMO it seems crazy to funnel effort to "go it alone' that has limited (or potentially negative) impact.

From my perspective as a long term Windows 'power' user (prior to my involvement with TurnKey) I found Ubuntu a great starting point, however I was already interested in tweaking things and fixing/working-around bugs. Ubuntu gave me a nice way to get a new experience but forced me to stretch myself a little to make it work the way I thought it should. The friendliness and accessability of the Ubuntu community was a huge bump in the right direction! After a while though I got sick of having to fix stuff every six months (or more often sometimes) when I upgraded. Unity then pushed me over the edge and away from Ubuntu. I still persevered with Ubuntu based distros, but in the end have landed firmly in the Debian camp (partially due to my involvement with TurnKey, but also partially due to my frustrations with stuff that was broken in Ubuntu).

In fairness, now that I am using Debian stable, I'm not updating anywhere near as much. Also installing additional software (when not available in the Debian repos, or software that is under heavy development and hence really out of date in Debian) can be a real pain, all in all I am much happier using Debian. I really hope that Debian continue with the LTS and it becomes a regular thing. If so, then I strongly suspect that Ubuntu may well find themselves as the poor cousin in the server space. There may well be some large companies that continue with Ubuntu as you can purchase support (a la RHEL) but as (in my experience) Debian requires less maintenance and there are plenty of Debian pros around who can provide paid support.

As for the desktop space (which is becoming more and more irrelevant anyway) I think that Linux Mint have a good model in their 'Debian Edition'. The idea of using Debian testing as a basis for a 'rolling release' distro, with their own repo as well (which I understand is pinned with higher priority) to resolve the core buggy apps (from Debian testing) and provide their own consumer experience. I think that is a good middle ground and allows them to focus on adding value both up and downstream with minimal effort... I think that Ubuntu wold be better to move in that direction. I recall reading some time ago that they were planning to shift all but LTS to rolling release but not sure if that is still the plan...

PhillMoore's picture

I have read articles recently about companies moving over to Ubuntu also due to their LTS is better than Debian.  Just an fyi.

I know turnkey moved over to Debian already though.  Anyway, I'll be diving into the OpenStack/JuJu/MaaS as well soon.  The first thing I thought of when thinking about setting everything up was Turnkey, lol.

Liraz Siri's picture

  • Debian supports a much greater range of packages than Ubuntu with security updates.

  • Debian is more likely to still be around in 5 years than Canonical as it is not dependent on success in a competitive marketplace.

  • Debian officially doesn't have 5 year support like Ubuntu's LTS releases, though they have started experimenting with that for Squeeze.

    If that gets enough support and gets adopted officially by the Debian community for future releases it will be a huge win for Debian and less of a reason to use Ubuntu, especially if all Debian releases are supported for 5 years. Ubuntu only does LTS releases every so often.

I know it would be a hard sell due to the tragedy of the commons and all but I wish companies would wisen up and pool resources to help sponsor and support Debian. They could get better, higher quality support without having to bet on the success or failure of any single commercial Linux distribution.

Better to work together with the missionaries than with the mercenaries.

Deb's picture

Forget about the long discussions.

It has come a point where many new users are confused about Debian and Ubuntu. Just read any Linxu mag and or podcast and you hear more about Ubuntu than Debian.

For worst, many website/organisation who needs/wanted to looks hip or genuinely trying to support Linux have no clue about the difference between Debian and Ubuntu.

Eg, I have seen a university wanting to support their Linux users for their vpn and hardcoded the word Ubuntu into their script. So now only Ubuntu works and Debian and it's other derivative are not supported.

Ubuntu in trying to make a business out of Linux had successfully created a big mess out of the entire situation where it does not benefit anyone. Many Debian developers was too nice or naive, let's the Debian and Ubuntu name to come too closed to create this confusion.


Many new users will just get confused or poor support (Debian and it's derivative not supported!) and just gives up and goes back to Windows and Linux developers back to Macs.

What left is hobbist or those who needs to make a living creating/supporting/writing about Linux.

To be frank, I know of many people who is so sick of hearing the word Ubuntu that they switched away from Debian (to Arch or openSUSE) to avoid this constant nagging.

Well done.

Liraz Siri's picture

Thanks for weighing in. I can understand how some people would be confused and frustrated by the situation with Ubuntu/Debian but abandonding Debian altogether is throwing the baby out with the bathwater.

Not that I'm against other Linux distributions. The free software ecosystem is complex and rich. There are pros and cons to different distributions going in different directions, using different brands.

Like with natural ecosystems I don't think anybody out there fully understands all of the possible interactions. Without understanding how do you figure out if something is really good or bad in a universal sense. We can try to spot patterns and spin our own little theories (like I've done myself) but I wouldn't get too attached to them or worked up about how this all plays out. Most likely labeling something as good or bad is at best a crude approximation.

Divtech's picture

No more news about Ubuntu in this category/

Visit my own blog Divtech

Heldeman's picture

Myself and many of my friends is using Ubuntu. Not one of us is a expert or can code at all. We moved from Windows to Ubuntu and some tried other systems. All came back to Ubuntu. The thing is, here you mention Windows and 100% know it. Mention the name Ubuntu and about 20% to 30% have heard of it. Mention Debian, Arch or SUSE and maybe 1% will know it is a operating system, if so many. Ubuntu is growing here and a computer chain store now will sell you a Ubuntu pre installed system. Here all know of Apple but they have very overpriced computers and hardly sell any computers except to a few corporates. I do not know all these computer terms but with Ubuntu you just plug and play. We  think it is better than Windows.


Jeremy Davis's picture

I prefer Ubuntu to Windows too. But I found Ubuntu was a little like a bike with training wheels. It allowed me to get comfortable with Linux and helped me to switch my thinking (I grew up on Windows and many things that are specific Windows things I thought were computer things...).

But when I got much more comfortable with Linux I found that Ubuntu didn't serve me quite so well. I find that Debian is much more secure and stable and so now that is what I run exclusively. Ubuntu LTS isn't too far away from Debian (stable) really although I still prefer Debian. At the end of that day each to their own! :)


Add new comment