February 2008

Video and Slides: Burak Emir’s Talk

Here are the video and slides from the talk on February 13, 2008. Burak’s homepage has a bunch of the papers he referred to at the end of his talk. Many thanks to our hosts at Twitter.


Creative Commons License
This work is licensed under a
Creative Commons Attribution-Share Alike 3.0 Unported License.Please link to this page directly when passing this on rather than the video or slides, as they may need to be moved in the future.

meetings
slides
video

Comments (0)

Permalink

Feb 13th Meeting Location: Twitter HQ

Many thanks to Alex Payne at Twitter for arranging to host this month’s meeting. Twitter’s address is:

164 South Park St
San Francisco, CA 94107

Alex says:

It’s a building with a dark green door. People can just come on in and walk to their right to a large conference room.

We’ll start at 7:30pm. You can find more details about the talk in the previous post.

Hope to see you Wednesday.

meetings

Comments (0)

Permalink

Feb 13, 2008: Burak Emir on “Extraction and Pure Scala”

The next meeting of BayFP is Wednesday, February 13th. We’re still ironing out the location details, but I wanted to send out the topic and dateThe meeting will be hosted at Twitter, 164 South Park St., San Francisco, CA. Burak Emir will be our speaker.

“Extraction and Pure Scala — from object-oriented pattern matching back to functional programming”

Extractions are a way to harness the conciseness of pattern matching known from functional programming with the versatility of object-oriented progamming. They are user-definable patterns which can match and deconstruct any data into subcomponents, without imposing constraints on the way data is defined. The flexibility comes at a price though, since user-defined code cannot be optimized as well as hardwired pattern matching logic.

In this talk, I will describe how extractions are used in Scala programs and how they relate to case classes, followed by just enough theory to see what kind of optimizations is done. These optimizations
have been proven correct (which does not happen too often in compiler construction) using a purely functional calculus that models Scala, as in fact extraction is a purely functional construct. At the end of this talk, I will make the connection to generic pattern matching and why I think it would be a good idea to start a purely functional dialect of Scala (Pure Scala).

Burak added unicode support to Scala, wrote Scala’s XML libraries, added XML expressions to the Scala compiler and wrote and maintained the Scala bugtracking system (before it was replaced with trac). He wrote his thesis on “Object-Oriented Pattern Matching” at the Programming Methods Lab at EPFL, under supervision of Martin Odersky, and now works for Google.

meetings

Comments (0)

Permalink