OSCON07 Executive Briefing
Opening remarks from Tim:
- every open source project has mechansms for control as well as participation
- web2.0 projects drive participation levels beyond the wildest dreams of open source yet enable massive centralisation of power
- are we asking the wrong questions
How do we measure Freedom? use/build on and adapt / participate / “fork” / switch
Look at sendmail, it was the king but by being open it has now been superceded by other options, when did you last hear of anyone doing something with sendmail?
how do you maintain? when:
- running it requires 100s CPU and constant updates to petabytes of data
- when SaaS is based on proprietary algorithms and data
- when redistribution is no longer necessary for everyone to have access
- when improving the program may be less important than improving the shared data
Is free as much a component of the success of Open Source as everyone believes?
all the top opensource projects are P2P free file sharing progs, such as eMule (bet no-one from that team is here quips Tim)
Tim: whats the API platform built on ?
Dave: LAMP entirely and almOpen Sourcet all code in PHP, Facebook is committed to Open Source and thinks about it all the time, whenever they do something they think Open Source first and foremost.
Tim: Is all your source code available ?
Dave: certain parts of code are not fit for showing in public (it’s the usual story of dirty code not being something they would want people to see, I often wonder about this one, either it is but would still be useful to coders or it isn’t in which case there is another reason it is not being released)
Tim: What about the user data that people create on Facebook?
Dave: Facebook respects user ownership of data, they will allow people to delete data but not move it
(Some application statistics are given by Dave in relation to the new API but I missed them, however I’m sure Google will provide many answers)
Tim next question is based around how he gets friends requests and doesn’t even know who they are (an increasing problem for anyone using social network services, I get several unknown friend requests a day now spread around Twitter, Pownce, Facebook, LinkedIn), Dave says they allow many ways to define a relationship through a two directional confirmation of details (they do provide much more granularity in defining a relationship to Twitter or Pownce)
Tim: why should we (Open Source) care about Facebook?
Dave: because we’ve spent many years developing in Open Source, supporting it and they like to give back to it whenever they can, in essence – they care…and they will continue to release as much of their software as they possibly can unless it’s too baked in to the system, often software is too entwined in the overall platform to allow easy Open Sourcing but whenever you can you should, especially if it will help someone else with a similar task, as in do it when it makes sense..
Tim:Foxmarks piggy backed on Firefox to get usage and is now 30-40% of size of Del.icio.us, there are always examples of open sourcing up front working out…How high up does open source go in Facebook?
Dave: it flows in our veins all the way to the top and we have more cool stuff coming….
Let’s hope you do as well as Livejournal quips Tim as they are the gold standard in this space….
Next up is the Firefox Extension system:
Mike Shaver of Mozilla
Tim: How many extensions are there currently on the Firefox platform?
Mike: 2400 or so on main site, various levels of complexity (he mentioned a Dark matter answer but failed to state what it was)
Tim: Fireox can slow down at times and I’m left wondering if it is an extension or some problem with a new release
Mike: probably a bit of both
Matthew Gertner – AllPeers
Approx 400K downloads
51-100K active users perhaps
Garrett Camp – Stumbleupon
We enable people to find new sites that match their interests through leveraging community aspects of browsing.
about 2.9 million registered users with about a third active (I’m a registered non-active user and don’t have it installed anymore as I have no need to find more sites to browse)
Tim: IE has built an extension mechanism much like Mozilla so why Firefox?
Garrett: Because they (Firefox community) are more active and provide very good feedback
Tim: do people get the source when they get an extension?
Mike: Firefox provide souce code but do not provide documentation to the level of MSDN but are getting there..
Matt: why Firefox? – for us we were all using Linux and Macs so it made sense, personal use dictated the choice as they didn’t use IE
Tim: what did Ebay actually acquire (when purchasing Stumbleupon)
Matt: they want the technology, alludes to Ebay using their technology to help users find other useful things perhaps?
Tim: is allpeers looking for a similar exit?
Matt: well it’s a standard thing these days! , they started in IE as they were all Microsoft development experts, when they did the beta they were asked to make it more llike a browser so they needed to make it cross platform and at that time Firefox was way ahead of IE so logical choice then
Licensing in the Web2.0 era
Eben Moglen’s opening quote as he approached the stage during Tims introduction “I’ll talk about licensing but web2.0 era doesn’t exist – it’s a bunch of hooey created by self promoters”, this pretty much set the stage for the ensuing hubris. This session garnered a lot of attention due the personal attacks on Tim and passionate statements made by Eben who clearly had a bee in his bonnet, I decided to sit back and enjoy it mostly but captured some points..
Google will be the biggest secret service force in the future leading us to store things ourselves (cheap storage will mean we will all store our own data rather than getting it free from Google)
on the 10 year timephase all this web2.0 is thermal noise
we need to get back to the concept of Freedom, Google is right that you dont need to ask someone to run a program
you should be allowed to make changes (to Open source code) and keep them to yourself, making people expose them is like the CIA asking to see your filing cabinet….
google do not infringe any of our rights but in the long term they might (not sure i understand why still)
what will end it all ? – the discovery that advertising revenue business is not worth all the money that is being pumped in to it
more ability to create software now (more processing etc) in the room than there was in entire IBM on 70s
90% time worrying about 2% of distribution
net net is worrying about principles rather than vapouring about Open Source we wouldnt be here now, still got patent problems, lots of work to do now that should have been done 10 years ago instead of talking about web2.0
Google Moglen on Google video to get an hour long video of his ideas
Facebook is what the kids are doing this week, doesnt care they’ll be doing something else next year
Moglen “we carried your water for 10 years while you made money”
all this user data in the hands of companies that can be made to give it out to goverments whenever they like, it’s a huge problem that isn’t being challenged now.
another dig at the people that left the free software movement early to go make money…
a good analogy that companies have privacy officers and investors (rapist or something) and who do you think will win when it comes to a matter of money over privacy??
long discourse on GPLv3 and the opportunity it took or didn’t
gplv3 gives us a 10 year timeframe to sort this stuff out
Eben’s closing quote after a clearly unsettled Tim suggested they stop…
“i invite you to a conversation you have been ignoring for the last 10 years…”
I enjoyed this session, not because it was confrontational or that Tim was on the end of some personal attacks but because there was vigour and passion in it, it made me really think about what was being said on stage, it was a new viewpoint I hadn’t considered before, it wasn’t rhetoric on a service I was familiar with or had already used, it was a call to action. I don’t know much of the history to this, I’ve met Tim a few times and have been to Foo before, I think he has done a great job of focusing people around a theme and it doesn’t bother me that he’s made money doing it (it seems Eben is bothered by it) but I can also see Eben’s point of view.
Having thought it over it’s clear there are bigger issues at stake than open APIs and friend requests that are not being addressed in the general arena, having said that the issues are being discussed in what is still a small group, most of the planet don’t have a clue about open source or web2.0 and most likely don’t care either…
After the break – Tim quips that “the future of web2.0 is dust under our feet” in reference to the last session and then introduces another slight change to the plan, instead of seperate talks the next speakers will all get on stage toghether.
Brad Fitzpatrick of SixApart, Doug Cutting of Yahoo! and Simon Peyton-Jones of Microsoft join him on stage for another panel session
Tim: Is Yahoo! getting in to Open Source because Google is?
no comment basically
disc seeks are the problem now and are relatively slower than before – tape based technologies are more relavant today?
Doug: explains the history of how the name became Hadoop, it was a small yellow toy belonging to his son (following some humour on how his other childs toy was called NeighNeigh (sp?)…)
Tim: EC2/S3 are great training wheels for people building systems that could turn big but how hard is it to move off these platforms
Doug: very easy if your approach is similar to what Amazon do now (I assume he means that by using structured APIs in the same fashion you could swap out the backends that EC2/S3 currently provide)
There is some discussion on how S3 could have used Webdav to make things more standard given the widespread adoption already out there, the fact is they haven’t though.
Next up Roger Magoulas does the trends review through book sales review this year titled “Trends in Open Source Adoption” this session should be familiar to anyone that’s been to an O’Reilly conference ( I wish he would update the picture on the opening slide!!) and provides some factual insights in to what’s hot or not in programming and open source adoption (I tried to take pictures of the slides like earlier but failed for reasons I wont go in to here hence have no data – Ruby still hot I think though !)
In a change to the published programme Cory Doctorow speaks next in a session about privacy titled Privacy Isn’t Dead — Let’s not Kill it (I last saw Cory talk at Etech 2005 I think and am interested to see what his take is)
The essence of his talk is about how we make decisions about technology that puts it more or less in control of our lives but ultimately its up to us to decide
One quote that stood out (enough to be twittered in unison with Nat) “I call that the Urinary tract infection business model, where the internet that once came in a rush is now a painful dribble”
Post lunch – what happens when a young open source company grows up? The session quotes Mark Twain in it’s description
“When I was a boy of 14, my father was so ignorant I could hardly stand to have the old man around. But when I got to be 21, I was astonished at how much the old man had learned in seven years.”
It’s a lot easier to be a sassy teenager than a responsible adult.
Tim wants these guys to ‘fight’ over the idea that all software should be free but not sure there is an argument to be had here as they look uncomfortable with the idea, (I find it disappointing in the extreme that a mock fight is being setup to which the contenders clearly don’t feel comfortable with, shouldn’t this have been cleared up before they took the stage,especially as the website states “We’ll get these guys going head to head” – lame)
Vmware taken the Open Source lesson to heart without being Open Source…. (this is the only thing I wrote during the session and looking back now I’m not even sure why)
the result is an uninspiring session that fails to produce anything of note sadly.
(The remaining sessions find me writing little down either as interest wanes, not just from me but I find myself exchanging looks with those at the table with me as laboured themes are gone over again, maybe my imagination but even Tim seems to be struggling to keep the momentum going, the desire to keep on the open source theme seems stretched and strained even, I find myself wondering and hoping that this is not a sign of the says to come)
A chat with MySQL CEO Marten Mickos about the impending IPO titled as “The Path to IPO”
During the chat Marten asks the crowd “why does change happen so slowly?” he wants things to happen quicker, I see what he means but as referred to earlier it will take even longer for the mass market to adopt (perhaps this is what he is referring to, if so then the answer is that the need just isn’t there imho)
Managing Linus Torvalds and Other Small Challenges is an impressive sounding session mentioning “game plans for world domination” no less, it’s an interesting insight in to the challenges of managing a diverse development group that has garnered global interest and therefore diverse opinions.
Why Free Software Values Work for Business a chat with founder Mark Shuttleworth around how Ubuntu has become the fastest growing Linux distribution by clinging fiercely to the free software principles.
Ubuntu – launchpad – web2.0 approach to Open Source …? upping the collaboration, good marketing of a well designed product with a good and vibrant community behind it
Next up is a showcase session with three companies selected:
Ohloh.net – service that allows developers in Open Source to compare what ‘stacks’ they are using and therefore build a community view of the best set of tools for a particular mode of development
OpenADS – open source ad server that has built a network of thousands of publishers that rival the big ad networks
OpenCV – listed but either didn’t show or I didn’t register them on stage
Open Source Hardware – A panel session moderated by Phillip Torrone, Senior Editor, MAKE Magazine with :
Christy Canida, Community and Marketing Manager, Instructables – a vibrant community of people building things (guns, ferris wheels) from K’Nex [edit: Christy left a comment that points out they do a lot more than K’Nex, I apologise as I should have checked but as I have a son mad with the stuff I focused on the K’Nex bit!]
Brad Carlile, Sun Microsystems – open source chips? myself and others at my table couldn’t help but wonder why – it’s not like just anyone has access to a fab plant to make one?
Limor Fried, Adafruit Industries – suppliers of kits and parts for original, open source electronics projects, parts are made specifically easier to solder and use
Andrew Huang, Vice President Hardware, Chumby Industries – it really is going to be released soon apparently – wifi internet browsing device (I really can’t believe this is onstage again tbh)
I’d have preferred to see 3D printing in this session as this has more power to disrupt than making bigger things out of toys, one off devices out of circuit boards or fluffy wireless lcd displays that can show you content from the web, the end of the exec briefing seems to be forced and running out of steam, as are my wrists and patience.
Why Congress Needs a Revision Control System – Karl Fogel from QuestionCopyright.org
The notion behind this is very interesting albeit only tangentially linked to Open Source (there are non open source version tracking products) but the ability to put effective version control on acts of legislation as they get amended and passed through government is a valuable one, i think the UK site TheyWorkForYou has some form of this in place. It makes me wonder about how well this could work in a corporate environment with regard to policies and mission statements. The point being to be able to accurately track changes in any important documentation, not just that for code
Open Source Documentation
I’ve yet to meet a good coder that likes to write documentation, if freelancers didn’t have to provide it as part of a contract it would be worse than it is now I’m sure, Andy Oram talks about some methods he is trying in order to get better documentation in place, he shows us some open source software that adds polling to documentation to get an idea of how accurate/good it is, I didn’t really follow along with it
We finish with a showcase devoted to OpenID with David Recordon and Simon Willison
After a quick overview of what OpenID is they run through some stats, the most interesting to me is that at this point they have approx 120million registered IDs (of which 60 million or so are from AOL) but as more companies adopt (Plaxo for example) the graph is looking nicely exponential for them at this point. There is some debate both internally and externally about how relevant OpenID is vs Liberty Alliance or Cardspace, I find these arguments moot now since mass adoption by web users to simplify their online identities and logins, essentially providing single sign on for the web will be a deciding factor, will banking use it for online banking? perhaps not. There is little interest from operators due to it not being seen as secure enough at this point (I had a chat with David and Simon about this later and the final outcome is that there will probably have to be some hybrid solution if the customer is going to come first in all of this) I’ve written before about how in my ideal world my OpenID is based on my MSISDN which is something I personally see as a unique identifier to me and something I almost always have with me, I’d hate to see the two competing although two identities is preferable to the 50+ i have now of course….. (including 3 OpenIDs only one of which I am using thanks to the excellent Sxipper) I’d really like to have something like [MSISDN].openid.com work for me with access from the mobile being automatically authenticated but I digress.
The Exec briefing was informative a the start but ultimately fizzled out for me in the end with too much showcase and not enough engaging debate.