'''Day two Keynote Presentation'''
Chris Aniszczyk opened by telling a story of how fans of the movie "Castle in the Sky" brought it down by Tweeting all at the same time during a specific part of the movie.
The story helped put into perspective of just how massive the amount of tweets Twitter has to handle and how it must constantly change to adapt. Then he dived
straight into Open Source explaining Twitter's philosophy on how it approaches open source through the following 8 principles.
* use open
+ if it exists already use it don't reinvent the wheel
* assume open
+ assume code may be opened in the future even if no plans currently
+ helps keep code quality high and makes opening open source easier when time comes for it
* define secret sauce
+ don't open source core business values
* measure everything
+ measure to know what is happening
+ let all internal employees access to this data
* default to GitHub
+ build community around Github first
+ gather community feedback fail fast and carry on if no traction
+ if traction grows then move to a foundation
* default to Permissive
+ use permissive licenses
+ want everyone to use it by minimizing the barriers
* acquire and open
+ when hiring small companies for it's talent; open to ensure the software does not go to waste
* pay it forward
Then he talked more about scaling and how Twitter's overcame scaling issues by giving the JVM a shot. They liked the JVM because it has a huge open source community
with lots of people working together doing world class work. Overall, through the talk I gathered that Twitter wants to help make other developers lives easier through open source to help others overcome similar issues. For example, frontend technology is not a core value for Twitter they "opened" it by giving away "Bootstrap" and "Bower" to help other developers. I agree with the philosophy "acquire and open" because putting stuff out there to help others adds value and lets others make use of things in different ways.
'''Cloud Computing Infrastructure: Intro with Open Stack'''
Don Pinto gave an overview of how NoSQL is trending and how companies are using it today - it was very technical focused. He went over the benefits of CouchDB compared to a traditional relational store. Some of the benefits mentioned is that it can handle migrating data easily and because it has no schema it is flexible in comparison to a traditional relational store. He went over some of the use cases that make CouchDB a good choice and how companies
are utilizing it. Many companies are using it as a "high available cache" for storing search result queries, for powering popular HTML landing pages and storing adds that show the same thing over and over to millions of customers. So things that are expense to compute that must be computed over and over again are examples of good use case for CouchDB . Other use cases that CouchDB solves well are session stores, user profiles stores, rich content and meta-data stores. Apart from web mobile app can take advantage of CouchDB too. A mobile version called CouchBase Lite can be installed on your phone that can work in tandem with CouchBase. Open Source was mentioned with a link provided to its Github page at the end of the talk.
'''Concluding Remarks'''
Even though the two talks were very different, with different product and a different set of agendas they shared similiar approach to problems and problems to solve. Both are solving issues of scaling and both are utilizing and embracing help from the open source community.
