Comments

• RyanT

  2008-09-16 15:16:33

  It's a sign of Mark Shuttleworth having his head firmly in the ground and realising that 90% of people don't use Free formats and therefore need them to an extent.
  
  Without that realisation I wouldn't have been able to take even one step into the Linux world because none of my files and formats would work, and I wouldn't have fancied spending several days solid re-encoding everything into OGG or even Dirac. It's through that I've been able to make my way in, and it's through that I can continue to step by step move over to Free formats. Sadly other parts of the web and other people don't have the same values, and so are still required to a degree. I have Fedora as a try out for much "cleaner" system I guess, and a tonne of things (web based) barely work.
  
  Before you can move everyone over, you need to give them the tools to be able to do so, and provide tools that can do what they currently want well enough. Sadly Flash still dominates, with half and half support so far, along with Java that's currently in the process of being open sourced but only at Java 6. Add into that the format thing, and you have a healthy dose of proprietary requirement for the kind of tasks 90% of people do.
  
  Most rehab clinics and smoking quitters realise that cold turkey is not necessarily the best way to go. Unless Ubuntu installs these by default on every version of its distro, and starts requiring them as dependencies or genuinely promotes them, rather than offering to those who needs, then I'll be worried.

• AlexH

  2008-09-16 15:26:58

  @Ryan - there is a sort of point there about codecs, but the Canonical shop has plenty of stuff for which there is free software for, e.g. DB/2, Unite, etc.
  
  Creating Ubuntu is pretty obviously not a profitable exercise. It would be sad to see Canonical trying to profit via proprietary software, though.

• RyanT

  2008-09-16 16:16:08

  It just seems like an extension of the previous deal they had on netbook remix.
  
  For Canonical and Ubuntu to succeed, it's entire business model needs to based around open source. It doesn't matter that Canonical might not make as much money as X company does from selling codecs, because they're a smaller company - they don't need as much to be profitable or to grow. Open Source software and the cloud are undoubtedly the models of the future - models where software isn't sold. or is sold for a lot less as to make most profits irrelevant, especially for a small company to be able to become a large one. For Canonical to be successful, it needs to stay and build its business around open source, even if right now it might not seem as profitable as the ordinary selling proprietary stuff model.
  
  I suggest you both read this:
  
  http://malstrom.50webs.com/disruptivestorm.htm
  
  It mostly relates to gaming, but can be applied here also.

• AlexH

  2008-09-16 17:44:55

  @Ryan,
  
  Let's look at this on a purely economic level. At the moment, Ubuntu (as a product) makes no significant revenue, but Canonical carry the costs of production. At a conservative estimate, those costs are $7 million per annum minimum.
  
  In order to generate profit - which, let's face it, is necessary - they have to sell goods and services based on Ubuntu which generate a margin in excess of that $7 million. So, if we assume a margin of ~10%, to simply break even they need to be a $70 million company.
  
  Reselling other people's proprietary software on their web shop, with the best will in the world, is not going to get them to that target.
  
  It's all very well to talk about disruptive business models, but those are the bare facts: they need to generate serious amounts of revenue to pay for Ubuntu.

• RyanT

  2008-09-16 17:59:09

  At the moment it doesn't make significant revenue. Half of that however, is not high enough adoption or visibility to generate the amount of customers needed to pay off the costs.
  
  Going forward however - emerging markets of netbooks where Linux is becoming more visible, of which Canonical is deliberately aiming for with a specialised distribution, along with the MID edition - and a growing server business, and along with this, recent retail availability with 60 day support, are all showing future potential.
  
  This is all what this is about - any company that thinks they're going to get that kind of money day one with a lesser known and adopted product, in emerging markets and an emerging business model, would be foolish, and Shuttleworth himself has acknowledged this. Open Source is a future prospect, and so is Ubuntu. That's the bare facts - for Canonical to become profitable, they have to realise this is a long term investment and project, not short term. If they think short term, they'll fall back on the model that'll die out in the coming years and they could go with it.
  
  Any disruptive business model, and any business model that involves any kind of competition with established vendors will not generate that kind of money on day one. It's a long term project, not short term.
  
  Could you tell me teh source of where you got the estimates?

• AlexH

  2008-09-16 18:12:17

  @Ryan,
  
  I think my one issue with Canonical, if I have one, is that by heavily subsidizing Ubuntu they're actually damaging the market - they've driven the price of distros down to zero, such that people can't sensibly make money from them. That's fine if it's sustainable, but when you're dropping ~$10 million a year on doing it, even with Shuttleworth's wealth there comes a point when enough is enough.
  
  In terms of the estimates, they're based on a discussion on LWN based on the statements Canonical have made about their staffing level (110 employees last count) - multiplied by average wages and costs. It's probably more than that - eg., they have central London offices etc.
  
  It's fair enough to say it's a long term investment, but you can only do that if you see a large revenue in the long term. I don't see where Canonical are going to get that from.
  
  And before you say "support", realise that providing support is an additional cost to what they're doing right now (unless they have many staff members wandering around doing basically nothing), and it's mostly a staffing fixed cost because you're not selling products or materials, you're selling time. To make that work, revenue would have to go up a lot - it's like putting fuel in a rocket, the more you put in the more you need to get it off the ground.

• RyanT

  2008-09-16 19:02:17

  There's a large revenue in the long term because, again, open source along with other computing areas like netbooks are emerging markets. Right now, they're fringe, they're small. Shuttleworth himself says:
  
  TG: How close are you to breaking even?
  
  MS Not close. It will require time and ongoing investment. We've positioned ourselves for what we see as the future of software - unlicensed software, people having access to the software that they want at the time that they want it. The service ecosystem around that software will fund it. And if we are the company that has best anticipated that future, then we will be best positioned to benefit from it.
  
  http://www.guardian.co.uk/technology/2008/may/22/internet.software
  
  It isn't just support as in "phone Canonical because this just exploded". It also extends to training, certification and perhaps more importantly, customisation. In small portable devices like netbooks and smarphones, customisation is key - the hardware isn't powerful enough for an ordinary monolithic, interdependent kernel like Windows, and nor does it really need it. Linux, and subsequently Ubuntu is a boon here - no need to write your own OS, but still have one tailored to your specific needs.
  
  This is where Linux, and subsequently Ubuntu's biggest opportunities are. They offer things Windows cannot offer to be nearly as competitive in this space, with Canonical still being able to provide a platform on that, and subsequently services to go with it. Yes, all these are addition cost, of course they are, it's the equivalent of saying 1 + 1 = 2. What matters is that:
  
  a) There's significant potential for growth, which there is.
  
  b) Said growth and potential outweighs any initial loss.
  
  As long as they're earning more charging for the service than it costs to provide service, then that's good. For a company that's making losses, any sign of profits is good, and should be leaped upon, not shunned because you can gain a more immediate gain. The costs you provide obviously do not contain money they're making that might offset that cost even the slightest. As long as they are seeing growing revenues, that's good. For the cost of support or service to outweigh cost of providing, they'd either have to completely jump the gun and as you said, have lots of people around the office doing nothing and burning through money. Few to no companies do that however - as long as the support and services are scaled to the demand, especially in emerging markets, and aren't priced so they are making a loss on the support and services they provide (which won't happen - that is entirely the point of the model after all), then it's very possible or even likely that they can reach a point of where revenues meet cost. This becomes an even more promising line on the consumer side - again, the retail copies and PC retailers where support is what you're paying for, not software. With Windows for example, after initial development, they technically don't have to charge much for the OS when it's sold - they make their money through the mark up on each copy sold compared to the little cost of making new copies that is in stark contrast to the continued production costs elsewhere which usually facilitates the cost. Using this as an advantage, replacing simply trying to subsidise development cost, they add in support for the extra value - swap one initial cost out for the other (if Ubuntu were proprietary, their OS development costs would be much higher remember).
  
  So they have several areas of growth and markets to to gain this revenue:
  
  Highly profitable upmarket consumers who demand support, usually on long term basis - currently small, but they are newer to the market than Red Hat or Novell
  
  Less profitable, but also less demanding and less cost - also currently small, but a fair amount of potential for growth, and even though they're less profitable, any amount of extra revenue and profit is good for a company just trying to earn it.
  
  Training - a double whammy here - businesses making the switch don't just need support, but their employees and bosses may need training too, and more obviously the IT peeps themselves.
  
  Customisation - Especially fruitful in mobile markets where computing resources are limited and so optimisation is needed both in hardware but also the software - subsequently meaning the OS and such. The likes of Linux and all things modular and open source are becoming more and more popular here precisely because of that. Offering customisation services in this area is a very good thing, and again, provides another avenue of growth and profitability. Even moreso with the range of devices and companies in this sector.
  
  They have several areas of growth, if not more. Again, this is long term. Canonical is a company based around offering services and support - they cannot just turn on a light switch and go proprietary. Their entire company and the entire point of open source takes emphasis away from selling the software - it would not be their strength. Not to mention their cost of providing support inherently replaces the extra cost that developing the entire OS themselves would - developing software to that extent is not their strength, and nor would it make much difference. It would be far more painful, and would sink it once for all.
  
  It is better to incur the short term losses for long term gain. All the markets they are serving are emerging, and are therefore currently small. This is changing however, and Canonical is placed to take advantage of this long term opportunity, not the short term. Changing that would do far more harm than good.

• RyanT

  2008-09-16 19:05:33

  I meant to put "less profitable but also less demanding and less cost *PC users and such*".

• Victor Soliz

  2008-09-16 19:29:46

  So, canonical is selling codecs, what's the relation between that and MS getting money when you get ubuntu?

• AlexH

  2008-09-16 19:32:05

  @Ryan: it's all very well to say the netbook market will grow; it almost certainly will. But Canonical are not players in that market: although they supply an OS to Dell, and hopefully they're charging for that, it's not going to make an awful lot of money. Netbooks are a low-value product, and the margins are equally low. You have to sell them by the bucket-load to make anything, and if you charge too much someone else undercuts you.
  
  As I said before, yes, it's quite possible to sustain short-term losses for long term gains. But, here's the problem:
  
  

  Not to mention their cost of providing support inherently replaces the extra cost that developing the entire OS themselves would

  
  
  It doesn't matter if it's cheaper to create Ubuntu than roll their own from scratch, the point is that they have a substantial fixed cost and they're not really providing services at this point. In order to provide services means taking on more fixed cost, which in turn increases the turnover target.

• Roy Schestowitz

  2008-09-16 19:41:38

  The question here is not about business models. It's an observation suggesting that Canonical puts a bundle on some (even all) PCs that contains Microsoft codecs. This was discussed here.

• landofbind

  2008-09-17 04:11:26

  Mr. Roy Schestowitz, Canonical doesn't sell codecs, it allow others to sell them. In this case Fluendo. Fluendo has been doing that for a very long time.
  
  http://www.fluendo.com/products.php
  
  It includes Windows Media, MPEG2 and MPEG4 and they read only.
  
  Your article is full "errors" as usual. Canonical isn't increasing the cost of anything. Ubuntu is still free. And they aren't forcing anyone to buy these codecs, nor did they forbid anyone from downloading the free ones.
  
  

  Just to clarify, we have written literally hundreds of posts denouncing software patents and we continue to do so.

  
  
  This is a lie. It's not "we". It's you. And you have written thousand articles promoting patented software and hardware. It's an daily occurrence. I call you on that several times. And you still do it. You're an hypocrite.
  
  Your "articles" are poorly researched (euphemism for not researched at all), full errors and deliberate lies.
  
  Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.

• Michael

  2008-09-17 04:12:44

  Does South Africa have software patents? Why aren't they just using the free codecs from mplayer or other free projects? Or paying for someone to make them usable at least.
  
  Presumably there is nothing legally questionable about any of those outside of software-patents-are-legal countries such as USA.
  
  i.e. why are the laws of a few countries being applied 'de-facto' by *foreign* software vendors to the rest of the world?
  
  And if the answer is 'they need to consider markets where s/w patents are legal', then why not do both and not lumber everyone with this unnecessary cost.

• AlexH

  2008-09-17 07:41:15

  @Michael:
  
  Canonical are based in the tax-haven of the Isle of Man, not South Africa.
  
  Also, if you're a business, it doesn't matter where you're based: if you sell in a particular jurisdiction, then its laws apply to you to some greater or lesser extent.

• Roy Schestowitz

  2008-09-17 08:58:37

  AlexH,
  
  I don't think that Dell should sell all of its Ubuntu PCs with those proprietary (and costly) addons. There should be a choice. Software patents are not valid here.

• AlexH

  2008-09-17 10:34:01

  @Roy: I'm not trying to suggest that; in fact, the opposite.
  
  I don't think these proprietary do-das actually help the financial bottom line, and in some cases are of questionable value anyway (e.g., watching DVDs on a netbook isn't going to be a common use case given they don't have, y'know, a DVD drive).
  
  I actually think that this model - no-cost OS, and pay-for add-ons - damages the free software market, and isn't going to help Canonical succeed. Hence my points about business model: you could understand it (but not condone it) if it was going to help them survive. With their fixed costs, it's fiddling on the margins, though.
  
  Software patents are, sadly, demonstrably valid here though. The UK is better than the EU in some ways, and measurably better than the US, but it's not true to say all software patents are invalid. Again, don't take this as a statement of support or defence, it's not.

• Roy Schestowitz

  2008-09-17 10:43:38

  I was going to do a quick post about the issue, especially in light of this, but I don't have the time. As for the UK and software patents, it is indeed complicated and I wrote about it many times before. I think that such issues can be addressed through debate though. Mandriva, for example, has just responded to my post, saying that the Mono dependency had been removed from OOo in Mandriva 2009 (KDE).