Ubuntu Open Week - Ask Mark - Mark Shuttleworth - Tue, Apr 24, 2007

see also Wednesday session.

TZ UTC+1

[18:08] <sabdfl> morning/afternoon/evening all
[18:08] <sabdfl> sorry for my tardiness, schedule-itis
[18:09] <sabdfl> ok! this is a Q&A session
[18:09] <sabdfl> so please fire away
[18:09] <sabdfl> will someone volunteer to act as a secretary and keep track of questions?

<JLP> QUESTION: Does Ubuntu have any plans to join forces with other Linux distributions, vendors and companies and together put some pressure on software companies and and offer them help so that they would port some of the most popular applications (like Photoshop, AutoCAD, ...) to Linux?

• We do already meet with ISV's, and encourage them to port to Linux. they always take a business view of the market opportunity. that's a factor of two things. the raw number of users, and the willingness of those users to pay for software. on the raw number, linux is doing very well. so hardware folks are starting to take notice. but on the ISV front, it's more challenging. many folks are on free software in part because they do not have to pay. there are some niche markets where that's not true - mostly places where Linux was a lower-cost (but not free) alternative to UNIX like workstations, graphics, CAD, movie animation and effects etc. but for raw consumers, i think the free software community should assume that we are going to have to build our own leaders in each of the major software categories. because, unless something changes and linux users start to be willing to pay for apps, the ISV's are unlikely to port

<adamant1988> QUESTION: What are Canonical and Ubuntu doing to ensure enterprise quality in the Ubuntu Linux distribution? Are there any plan to include more 'user friendly' modifications to gnome (ala SLED), interoperability work, etc.?

• two separate questions, i'll address them separately. on the quality front, i'm very excited about apport, and really want to credit pitti and the dev team with that infrastructure. we are working to make it something that can be deployed for stable releases as well as development releases. this will give us real stats as to places where apps break the most and allow us to improve the quality of packages after release on a systematic basis. i'm also really pleased at the community iso testing effort, and the hardware database. those are examples of places people can contribute to ensure that the whole platform works well and then we are working on per-package testing frameworks. we have to figure out how to make those work well in the whole debian universe too. on the user-friendly desktop mods front, we'll happily bring in work from other distros if it's a genuine improvement. upstream's willingness to consider the patches is of course the acid test, but we would bring something in that upstream was not yet ready for if we were really convinced ourselves. if you think there are items that really should be included, then (a) make sure they are packages byworking with -core-dev or -motu, and (b) participate in the ubuntu dev summit, in person or by voip, to motivate for their inclusion

<suzan> QUESTION: Mark, recently you told in an interview "the time for mass consumer sales of Linux on the desktop has not yet come". What do you think, when is the time for linux on the desktop? One year, two years, many more years? On what it will depend how quick the time comes?

• before mass consumer sales, we will see targeted niche markets. so, for example, right now workstations are a Linux-heavy market. i think there are some other specific markets that can be targeted in emerging markets, for example, price is a huge driver. so we see people in brazil, china, africa etc being genuinely interested in linux, not just in linux-before-pirated-windows. and in the west, people have said very loudly they want it, we just need to figure out how to reach JUST those folks, and not accidental windows users. sorry, windows users who accidentally buy linux

<kwah> QUESTION: Mark, what do you think about the latest MS 3$ initiative?

• clever move. http://www.manilastandardtoday.com/?page=business6_april24_2007. he gets it. free software is a far better platform for education than the $3 offer. - far more applications, means you can teach many more things with technology, not just word processing and spreadsheets. for example, databases, programming languages, art apps, music, astronomy, statistics... you name it - source code, means you can learn far more about how that technology works, important for super-star students who will rocket ahead of even their teachers - localisations, means you can teach in your n indigenous languages

• in short, education departments that really care about education have lots of reasons to choose free software rather than the cheap (and usage-limited) windows offer

<remix_tj> QUESTION: Mark, will exist in the future a certification relased by Canonical, like RedHat Certified Professional?

• yes! alon with training programs, and training materials

<mc44> QUESTION: How do you plan to make money out of launchpad? Is it generating revenue for you already?

• the first step will be to make it possible for people to subscribe and get the ability to annotate almost any data in LP, privately. so, for example, private bugs, private specs, private branches. then, we will also host other distributions that are ubuntu derivatives, so we handle all the building and archive management for them

<stuart_> QUESTION: Kubuntu is a great distro but (IMHO) lacks polish and features compared to Ubuntu. Are there any plans to bring in more paid Kubuntu developers now or in the future?

• yes, the kubuntu team will expand, but i think kubuntu will always be more independent of canonical, which is in many was a good thing. somethings happen first in ubuntu, because that's where we focus our resources for new releases, other things happen first in kubuntu, because the decisions there are more bottom-up

<McKinney> sabdfl: QUESTION: in last time there are discussions that ubuntu should not be released every 6 month. you shoult give the developers more time. what do you think about it?

• i think the current system, of 6-month releases with LTS releases every 2-3 years, will work well. there are no discussions or plans to change that at the moment

<kwah> QUESTION: What do you think about use of proprietary software in education? Shouldn't it be just a _privilege_ for software developers to provide SW for education without having any direct profits?

• no. there's nothing wrong with the profit motive, unless society lets it run amok. regulation is supposed to ensure that every business has checks and balances. competition does the same. i think it's naive to disregard the value of self-motivated investment. i have no problem with proprietary software - microsoft has every reason to charge me to use software they wrote. but i think the free software community can produce better software. and i choose to use that instead

<pointwood> QUESTION: Any rough numbers on how many *buntu machines there are running in the world currently?

• 4-16 million, my best guess including derivatives is around 10 million.

<mallux> QUESTION: Did you have anything to do with Michael Dell running Feisty on his home laptop?

• no - i was as surprised as everyone else. a good surprise, though

<tsmithe> QUESTION: How are we doing on the bug 1 front?

• quite well I think. i'm seeing a lot of people who are NOT tech specialists talking about linux. and who knows, perhaps ubuntu will shortly be available from your favourite source of PC's!. how do YOU think we're doing on the Bug #1 front?

<Clem92> QUESTION: ShipIt is currently not available für Xubuntu... will that also come for Xubuntu in the next time?

• i don't think so. xubuntu, as i understand it, is a more specialist environment, so we probably would not fund cd shipments for it. it's hard to justify on either philanthropic or business grounds. though it would be nice at least if you could order xubuntu

<zul> QUESTION: where do you see Linux virtualization 3 years down the road?

• i'd really like to see a free software solution that is comprehensive, stable and fast. i'm pressed with the diversity of options. xen, of course, and kvm and then container-style things like vServer and others. *im*pressed, i should say. now we need those to mature. i hope ubuntu will be a good place for people to test those, as well as a good place for people to run things like vmware if they want

<Belutz> QUESTION: if a loco are creating a nation wide event, is it possible that we ask canonical for sponsor (aside from conference kits)?

• ask of course! we haven't sponsored anything like that before, but it is possible. so feel free to make a case for it

<ryuujin_> QUESTION: why is launchpad a proprietary software?

• several reasons. first, it's worth pointing out that we fund a huge amount of GPL software development. so LP is not non-free because we don't know any better in general. we've thought about it very carefully. the major reason is that LP is explicitly a short-term, WRONG solution to the problem. the problem is a lack of information flow between projects. the right solution, from an engineering perspective, is a federated, distributed, standards-based approach. where data from bugzilla flows to LP, and into debbugs, and into roundup, and into sourceforge. then everyone uses their preferred tools, and the data just migrates as needed. but, we couldn't wait for that to happen, so we wrote a tool for ubuntu that new how to link to other tools. now, if there were MULTIPLE tools like that, it would divide the eyeballs interested in agregating this information. so, imaging you have a bug, reported in ubuntu, debian, upstream and gentoo. with LP, someone from each community just has to annotate it once, saying "our bug tracker knows about that issue, and we are tracking it as #324342". if there were multiple LP's, people would have to do that work multiple times. the result would be a mess. you would not have 1 bug number for 1 problem. no more Bug #1! so that's the major issue. we have released bits of code that we thought WOULD be useful to other communities. either infrastructure stuff (Zope, SQLObject, translation bits etc). or app stuff. we do have a plan to get to a point where we can GPL it. but that involves a lot of work, and i think it will take time, i don't want anyone to expect it to be free in 2-3 years even. but we're working to that roadmap, and will hopefully get there

<joebaker> QUESTION: Do you have any opinions you'd like to share about the progress of GPL v3.0?

• only that i think it's important work, and a good transparent process, and I really hope they come to a result that the community embraces wholeheartedly, even Linus. from the drafts, we're not there yet. i do trust both rms and eben moglen. when v3 is out, we'll take a decision about the code we have copyright in. ultimately, though ubuntu is an aggregated work of hundreds of items under many licences. so that will remain true

<YokoZar> There is a strong chance that Wine will come out of beta sometime during Gutsy or Gutsy+1's development. In the past several distributions (Corel, for instance) have included supported versions of Wine officially, to make other applications easily ported to their platform. This would allow Ubuntu to support the installation of Win32 software, both free and proprietary, via something like Add Applications or Click and R

• interesting. i admire the WINE effort, hugely. i think it's important. but i don't think that the future of free software lies in running non-free software! we should definitely discuss it at the UDS, so please come along, and remind me to invite the WINE guys too

<\sh> QUESTION: Mark, good evening What is the best way to convince companies (who are providing DC services) to switch from regular RHEL or SLES/Novell Business Server to Ubuntu Server, especially when you think about Oracles turn towards it's own Linux solution?

• DC services? you mean large-scale hosted servers?

  <\sh> sabdfl, DC == Datacenter :) it would be great if folks running huge data centers adopted ubuntu. salesforce would be awesome. ebay, amazon etc. i think they would have a great experience, save money but still get great support. are you asking becuase you know such a company? ok, i'll assume so

  <\sh> sabdfl, in general...most companies are using RHEL or SLES because they are oracle certified...now oracle comes up with its own linux solution...how to convince our bosses that ubuntu is better ah, right. well, the oracle test is one we have to pass, and we will do so eventually. unfortunately, oracle's strategy means they will likely not look at the hard data showing rapid adoption of ubuntu in the wild. because they want to see adoption of Unbreakable Linux (UBL). i tihnk they could have done UBL around Ubuntu and had a much better result already. but they decided to do it as a RHAT clone. we just have to be patient and keep growing elsewhere. if you know a company that has a large data center, its worth asking them to find a niche area where they can test ubuntu. usually they will do this in low-risk areas. like web farms, or redundant compute farms. and when they have a good experience there, they will often expand their use. and put pressure on the ISV's they deal with to certify the platform. a foot in the door is all it takes. so try to find a low-risk area where they can try it

<imbrandon> ( just a small note , you can poke me later about \sh , our company runs one of the largest PCI DC's in the USA and we're phasing out all our CentOS in favor of Ubuntu )

• would they consider being part of a case study?

  <imbrandon> definately

<[doctor]> QUESTION: Mark, what do you think about interaction at a level of the governments (Russia & East Europe) for distribution edubuntu. Any plans?

• well, Georgia has already done a large program like that. i think Armenia is looking at the same. i do expect others to follow. and we will work with any government that wants consulting and customisation help

<dinda> QUESTION: In a recent interview you spoke about using a Web 2.0 to harness the power of user data/input. Do you see a place for that within the Ubuntu Community? For example an Ubuntu Community Space that serves as a front end to that data mining process.

• the ubuntu forums and ubuntu wiki are already an amazing resource for users. it's incredible, the knowledge that is encapsulated there. we are also seeing amazing work in the ubuntu question-and-answer tracker. which formalises community support for the platform. and builds a knowledgebase. https://answers.launchpad.net/ubuntu. there is a superb team building there. i'm open to other suggestions too!

<Clem92> QUESTION: What OS do you use on your PCs at home and at work? Ubuntu, Kubuntu, Debian, etc?

• i use ubuntu on my laptop, kubuntu on my desktop, i also have a MacOS PPC machine at home. and i've seen vista running in a virtual machine, but never for very long

<adamant1988> QUESTION: I think we all recognize Canonical ltd.'s importance to the Ubuntu project's continued growth and prosperity, however Canonical remains a mysterious figure to Ubuntu users because it is a private corporation. I'm sure many users of Ubuntu are concerned with Canonical's financials (profit, etc.) and would like to invest (read: buy stock) in the company. Are there any plans to 'go public', if not why?

• no plans to go public at this stage, though it's nice to have lots of VC's expressing interest. i don't want to take a short-term view, and the markets ultimately force you into that. it's rare to find public companies that don't have that pressure, or even VC-funded private companies, and right now we have the luxury of having plenty of funding and a long-term mandate to change the economics of the software industry in a profound, philanthropic and commercial way. we are slowly building up a revenue stream in canonical, all of which gets reinvested in ubuntu. my dream is to build something unique: a genuinely free platform, that PAYS FOR ITSELF. that has never existed before, but i believe it can exist, and our goal is to find out. at the same time, canonical and ubuntu are separate entities. other companies are starting to invest in ubuntu too. sometimes via canonical (contracts for development, ports etc). and sometimes directly (hiring developers and getting them to spend time on ubuntu or kubuntu). ubuntu is bigger than canonical, and that's a very good thing. it would be nice to have more non-canonical people in the CC and techboard. nobody can buy their way onto those, but as more companies devote full-time engineers to ubuntu, it's more likely that they will be able to build the necessary experience and credibility to hold a seat their on their own merits.

<erstazi> QUESTION: where do you see Ubuntu in 10 or 20 years?

• that's a very long time, my crystal ball gets all cloudy. right now, the goal is (a) be the best free software desktop possible, and (b) work towards being sustainable without licence fees. that's all i can see!

[19:02] <sabdfl> thanks everybody, great questions!

[19:02] <PriceChild> Thankyou everyone else for all your questions!