I’m in Tampere, Finland, attending Openmind’06.
Chris Dibona from Google opened up by talking about how Google does things the open source way. He goes on to show some pictures from the various generations of their server rooms, concluding that PC architecture might not be the most reliable platform around. Hardware stinks, so you have to build a software layer that provides stability. Google runs Linux based servers and they use various OS libraries, languages, compilers, web toolkits and other things to achieve what they do today. They use due dilligence for Open Source, to see where a particular chunk of code comes from and if it raises any red flags, they will deal with it.
According to Dibona, Google wouldn’t exist without Open Source. It’s not about the price (Google has all the money), it’s about having control and ownership. Open Source enables you to control your own company and be adapting anf flexible to market conditions.
He shows examples of map mashups where people combine google maps to databases they might find useful. Google has no interest to provide information on where for example universities in Sweden are located or where to get best ice cream, but they are interested to solve the hard problems like mapping and let people innovate on top of that by bringing data in from various locations. Some discussion emerges on the trust issues of Web 2.0 providers. Dibona answers, that you have to deserve the trust to be an API provider. They are willing to fight US governments search results etc. to protect their users.
I asked about Youtube’s copyright problems and no surprise, they don’t really worry too much about copyright issues related to Youtube. I think their company was founded on taking advantage of the gray area anyway, by making money out of content made by others.
He asserts that web standards are important but there are cases where customer requirements get higher priority. The reason their front page is not completely standard-compliant is because that enables them to save terabytes of bandwidth and bandwidth is important if you want to offer a reliable service.
Brian Chan and Bryan Cheung from Liferay talked about Web 2.0, how we move from top-down software delivery (like platforms and other one-way of doing things) to an ecosystem of different approaches and solutions, just like a conversation. Open source is like a conversation while proprietary software was more like a broadcast.
They work on Liferay, a Java-based portal solution that currently has highest number of downloads per month from the Java alternatives. For them a portal is a cross-road, rather than the end (Rome) in itself: it adds value on top of the various roads (services). They themselves are also in the cross-roads: large enterprises still look for stable release cycles and robust services. Combining the chaotic dynamic development approach with a stable one is a challenge. Redhat is a good example of having community-supported (to drive innovation) and company-supported (to drive stability) versions of their solution.
Open Source enables people to try everything. It’s a kind of “Rice rocket” approach to application development: start from a cheap “car” and use lots of resources to tinker and modify it.
[UPDATE] What really struct me about Liferay is the ethical grounds by which they operate. Their philosophy is ROG, return of giving. They invest part of their revenue in charities. What also comes up with their case examples is that they want to make a positive impact in the living of others. With 150% increase in revenue every year since 2000, ROI is clearly present. What is better than ROI + giving? Maybe a theoretical economic model needs to be formulated for ROG to spread it among corporations that want to make a difference.