Randomly bumped into Nat last night on the streets of San Francisco – was good to see him, although I had to rush out. Attended the 401 Group networking session – was good to catch up with Cote and meet some new folk, but found the non-bloggable panel a little lame.
I’m back travelling again. Tonight Jayne and I head down to Christchurch to begin 3 weeks of looking after her sister’s kids. Tomorrow we hope to see the excellent Banff film festival showing, and hopefully a beer or two. Monday I’m back out to California for 4 days, to spend some time in the Menlo Park offices talking to a lot of people – if you’d like to meet up, drop me a mail, I’d love to sit down and talk. Looking forward to finally catching a Sillicon Valley OpenSolaris user group session, and with Ian presenting on Project Indiana, it’s sure to be an active night of discussion. I’ll also be at an open source reception sponsored by 451 group on Tuesday night, and possibly this if jet lag permits.
Pretty thrilled for Seth Wagoner that Interclue has now launched – a Firefox extension designed to make your life easier when browsing the web. It’s helpful to see text previews of the various links (unlike the various other competing extensions that only show an outdated screenshot), and I use this one quite a bit in my day to day browsing. Great to see this sort of stuff come out of the NZ IT, even if it means that we now can’t tease Seth anymore about when he is going to release. Now it’s out there, I can only imagine it will get better and better. Rock on Seth, nice work!
I’ve been thinking a lot about Project Indiana, especially in terms of how we might grow a community around the distribution. Fortunately, there are some awesome examples, and very much an opportunity to learn from.
It took just 3 years to create the phenomenon that Ubuntu has worked hard to achieve, with well over 600,000 registrations in their hardware database, and millions of users worldwide. Within that time, they’ve had 6 successful releases, top of distrowatch, 2.5 million posts on ubuntuforums.org, and really are one of the most awesome distributions. Most of the what follows isn’t exactly rocket science to the people who have been keeping a close eye on this community.
#1. Mark Shuttleworth is a spaceman
Anyone who has met Mark knows that he’s absolutely solid well thought out guy, and a perfect spokesman for free and open source software. Not only that, but the SABDFL has been to space. It has the cool factor, and people generally rally around cool people like Mark.#2. Smart hiring process
Mark (and Canonical) had smart hires. He concentrated on people that already had high profiles within the open source community, specific to the roles and technologies that would be leveraged. Who wouldn’t immediately follow and love the Jeff Waugh’s and the Benjamin Mako Hill’s of the world? Not only that, but he also gave each hire the opportunity to suggest other hires that should be interviewed. Not necessarily original, but smart, and the canonical line up were a whole bunch of smart motivated people.#3. Debian, Debian, Debian
Ubuntu started from a well engineered base distribution, Debian. With access to 16,000+ packages, an awesome infrastructure, support for multiple architectures and an organized community and user base, it was a winning combination. Ubuntu fortunately didn’t have the same social differences and the releases started rolling.#4. Time based releases
GNOME learned this relatively early on in its history, and we’ve never looked back since. With a 6 monthly time-based release following the GNOME upstream release schedule, it provided the best opportunity to automatically qualify for a whole range of user interface improvements, and focus efforts on a few of the harder issues for each release. With regular releases, developers could not only stay on the bleeding edge and concentrate on their project, but also provide the opportunity for new developers to join existing communities and ensure they were up to date with the tool chain and libraries.#5. Freedom
Very early on, Ubuntu had a strong commitment to freedom – the freedom to download, run, copy, distribute, study, share, change and improve. That freedom was complemented with their ‘humanity towards others’ definition, capturing the spirit of the free software world. Very early on, they implemented a code of conduct policy and encourage inclusiveness.#6. Freely available
An obvious benefit of having Mark behind the project, Canonical made a Linux distribution more freely available than ever before, regardless of location or bandwidth. shipit.ubuntu.com was born, sending millions of CD’s right around the world, allowing advocates to share spare copies to their friends and colleagues.#7. Soft porn
While it proved to back fire on the team, Mark’s idea of showing the ‘humanity towards others’ and a natural look by a series of controversial nudie backgrounds wasn’t necessarily a bad idea. It created a buzz, and people downloading it to see for themselves. It turned out that the initial target audience was a little more wider reaching and mature, but perhaps a shrewd publicity stunt?#8. Derivatible distributions
Ubuntu encouraged and provided technology for derivatible distributions – both install and LiveCD. Out of that spawned the likes of kubuntu, edubuntu and xubuntu among others.#9. Face to face
With the reality of a 6 monthly release cycle, stuff had to happen fast. The open development meet-ups were a strong part of that cycle, allowing the teams to set goals for the next releases, along with proper specifications and boundaries for that work. Everyone was invited to attend and participate, with Canonical sponsoring travel for those developers who had core knowledge to get the important goals achieved.#10. Best of breed
Not taking away from the awesome job everyone did, but for a long while Ubuntu was considered purely an integration team – very little development was done. It was a distribution with a collection of the best of breed open source technology, polished to perfection. They stuck to their ‘just works’ blue sky dream and focused on the right set of priorities – from power management to wireless support. They were the epitome of a typical web 2.0 company with their hardware database, with every submission helping to grow their potential user base and determine their priorities.
Clearly OpenSolaris needs to stand on its own – every community is different, and it would be a mistake to generalize. However, I do believe there are overlaps in terms of having a core set of goals – whether that’s being free and open, focusing on things like hardware and peripheral support, application availability or help and support. We have much to learn.
Phil hooked us up with some visiting Californians, Peter and Hans from simplyhired. Unfortunately they got the wrong cafe for our planned breakfast so missed out on meeting Steve, Jennifer and Mauricio, but we caught up with them a little later for a coffee and some gelato. Great to have a bit of a geek chat with Peter, who was previous co-founder and CTO of AtWeb later acquired by Netscape/AOL, and give them some insights into the NZ IT scene. Awesome guys, and hopefully I can visit them when I’m next over in Mountain View.
I’m proud to be a gold member of an airline that does this. Awesome – now just need to convince them to bring back wireless.
Given that I didn’t have a session pass, I’m slowly starting to catch up on the feeds of the general sessions. Is it just me, or is Stephen O’Grady starting to do voice overs? – I’m almost sure it was his voice at the end of John Gage’s initial introduction to the conference.
JavaOne was a pretty stressful week all in all – not only was it my first experience first hand with the enormous conference that it is, it was also an opportunity to interview for a program manager position for Project Indiana. I’ll spare you the interview details; but most of them involved both sides getting really excited by the thought of what could be achieved when the concept turns into reality, and how we might go about it. Ian presented some of his thoughts during his CommunityOne talk on Monday, though didn’t specifically mention ‘Project Indiana’ – it all sounded sane; it all sounded very real.
Queue some unexpected press. Midway through the week, the pace quickened up thanks to Stephen Shankland’s Sun hopes for Linux-like Solaris article, and the opensolaris-discuss rumour mill kicked into full gear (rightly so; the communication was not ideal and I don’t think anyone involved would dispute that). Marc chipped in with some great detail, and slowly the pieces started to take shape (along with some very reasoned follow-up reactions thanks to John and Bryan, which was awesome to see).
I’m excited about the project and I think it has an important future, so much so that I’m planning on moving to marketing and leaving my desktop engineering roots to take on the product management role for the distribution. Yes, I can still feel my heart beat, so hell hasn’t frozen over yet! Ian posted his thoughts on the concept, and that’s about as much as we’ve covered during the discussion at JavaOne. We all have different ideas of how this might work etc.. but nothing is set in stone other than the very real desire to get this done. I’m very much looking forward to seeing how it all pans out.
After Stephen’s post last week about whether it was appropriate to post jobs on OpenSolaris, this one is probably suitable for the new alias – Sara has an interesting and cool new position for an intern working in the US (not actually just Austin) on marketing initiatives for OpenSolaris. Rock!
