Map, Filter, Reduce - In the Small and in the Cloud

Very interesting talk. Good coverage of the topic without getting too detailed in any particular area, pace and progression were fine, explanations of abstract concepts were effective. Further examples with implementations other than Hadoop, such as CouchDB, would be nice additions to the talk.

Very good explanation of concepts, though I didn't exactly come away with a clear picture in my head of _why_ this technique excels at its use case. It would probably have helped me to see a generally familiar practical use case with a common approach, with its performance numbers (not a live demo) contrasted with the performance numbers of a map-filter-reduce approach. Maybe a CouchDB vs. MySQL comparison for a practical data transaction sequence would fit this need.

Ultimately, I came away from this talk feeling like "these are good ideas", but not feeling like I could then convince the next person of _why_ they were good.

Excellent intro and overview of very new and complicated topics. Nice way of boiling down some complex technologies. From theory to a demo with Hadoop. Slides could use some improvement.