Scala Vienna User Group

Name: Functional Programming Study Group #4
Start: 2014-05-22T16:00:00Z
Location: openForce Information Technology GesmbH

Functional Programming Study Group #4

May 22, 2014 · 6:00 PM
openForce Information Technology GesmbH

This will be the fourth meeting of our functional programming study group.

We are going to discuss chapter 7 "Purely Functional Parallelism". It's only one chapter this time as chapter 7 is quite long. Please note: we are currently using revision 13 of the book.

Christian has volunteered to get the discussion started and to lead us through the chapter.

What do I have to do in order to join?

All you have to do is read the text (and RSVP). But you will learn/benefit more in my experience if you also try to do some of the exercises. If you put your solutions somewhere publicly accessible (e.g. GitHub) or bring them to the meeting, we can discuss and compare different approaches

What does presenting a chapter involve?

There is no need to prepare slides or anything like that. But you will need to read the text more closely to be able to summarise the individual sections of your chapter for us and lead and encourage the discussion. It will be your call to separate the less important aspects from the more important ones, as we usually have only about 45 mins per chapter.

Join or login to comment.

Hit enter to add your reply

Alexander D.

The book »Learn You a Haskell for Great Good!« was recommended by Axel and Yago at the after-party ;-) http://learnyouahaskell.com

0 · May 23, 2014
- Axel G.
  
  all the others are thinking "that must have been one hell of an after-party" :)
  
  0 · May 23, 2014
- Yago A.
  
  You know things are getting wild when you start talking about Haskell :D
  
  1 · May 23, 2014
Bernhard H.

Hi, just following the meet-up again: Here's what I meant with "the computation graph can be compressed/cut due to the introduction of fork":
https://gist.github.com/bhuemer/657f59312165d63feaf6

If stuff is strictly evaluated, we might as well apply functions to it directly, without deferring it to a thread (the compiler knows how to optimise compile-time constant additions, etc. as well). If one wouldn't want that behaviour, you could always use fork() or delay(). I've also implemented the memoize we spoke about.

3 · May 23, 2014
- Bernhard H.
  
  .. and a version of fork that doesn't deadlock without actors is now part of it as well. As I said using something like promises and back-off if the future hasn't finished yet.
  
  0 · May 23, 2014
Dominik G.

Solutions: https://github.com/TheDom/functional-programming-in-scala

0 · May 22, 2014
- View all 3 replies
- Peter B.
  
  https://github.com/peb...
  
  0 · May 22, 2014
- Alexander D.
  
  https://github.com/Ale...
  
  0 · May 22, 2014
- Peter N.
  
  https://github.com/pno...
  
  0 · May 22, 2014
Peter B.

https://plus.google.com/hangouts/_/gzlcjrzyt2knxgdoritntw4unqa

0 · May 22, 2014
Aaron C.

Hi Everyone,
I would like to do a combined user group BBQ on the 4th of July at Cobenzl. Pay 10-15 EUR for all you can eat/drink. I was also talking with Clemens my co-organizer for the Freelancer meetup and he wants to do an additional one in June (because you can't have too many BBQs) at Donauinsel. If the team is interested, I will get some dates and we can add the event.

2 · May 17, 2014
A former member

Hi guys, I was wondering if it would be possible to know in advance the dates of the next meetups.
In this way it would be way easier to organize different appointments/events/duties. Some meetups, for example, follows a pattern (like "every 2nd wed of the month")

I just don't want miss any of them :(

0 · May 14, 2014
- Peter B.
  
  Hi Alberto, I think its a bit difficult to give exact dates more than four weeks in advance. But the general "rule" is that we try to have a meeting every month (usually the last week of the month, this time it's the penultimate because next week is a holiday)
  
  0 · May 15, 2014
Peter S.

I have finished reading this chapter yesterday, but not sure if I can present in comprehend-able form to others.

Generally speaking - I find FPs view on multi-threading standard - let's build a algebra (data structure) representing the computation & execution and then just run it.

I think even if we don't have presenter we can discuss different aspects of api design. And I think the book can be particularly useful in this area.

0 · May 15, 2014
- Peter B.
  
  Hi Peter, don't worry, I think Christian has volunteered to lead the discussion. Cu next week!
  
  0 · May 15, 2014
A former member

I only just yesterday started with the intro of chapter 7. If no one has been found so far for presenting it, I can try to make extra effort to not only read & understand it, but also extract reproducible information.

I typically don't find multithreading difficult and I'm furthermore interested in how this is solved in FP.
Manager speak: I'll see it as challenge :)

0 · May 15, 2014
- Peter B.
  
  Thx Christian!
  
  0 · May 15, 2014
A former member

Phew, back from serious internet spaceships, now I have to keep up with the last two chapters I haven't fully done as well :P Let's see how that goes.

0 · May 8, 2014