Susan Potter, profile picture
Uploaded bySusan Potter
PDF, PPTX6,885 views

Why Haskell

The document discusses the programming language Haskell, highlighting its features like being pure functional, lazy, polymorphic, and strongly typed. It addresses claims about Haskell's performance and productivity, with an emphasis on its type system's verifiability and ease of testing. The document also touches on key considerations for choosing a programming language, including technical aspects and cultural implications.

Embed presentation

Download as PDF, PPTX
1 / 101
2 / 101
3 / 101
4 / 101
5 / 101
6 / 101
7 / 101
8 / 101
9 / 101
10 / 101
11 / 101
12 / 101
13 / 101
14 / 101
15 / 101
16 / 101
17 / 101
18 / 101
19 / 101
20 / 101
21 / 101
22 / 101
23 / 101
24 / 101
25 / 101
26 / 101
27 / 101
28 / 101
29 / 101
30 / 101
31 / 101
32 / 101
33 / 101
34 / 101
35 / 101
36 / 101
37 / 101
38 / 101
39 / 101
40 / 101
41 / 101
42 / 101
43 / 101
44 / 101
45 / 101
46 / 101
47 / 101
48 / 101
49 / 101
50 / 101
51 / 101
52 / 101
53 / 101
54 / 101
55 / 101
56 / 101
57 / 101
58 / 101
59 / 101
60 / 101
61 / 101
62 / 101
63 / 101
64 / 101
65 / 101
66 / 101
67 / 101
68 / 101
69 / 101
70 / 101
71 / 101
72 / 101
73 / 101
74 / 101
75 / 101
76 / 101
77 / 101
78 / 101
79 / 101
80 / 101
81 / 101
82 / 101
83 / 101
84 / 101
85 / 101
86 / 101
87 / 101
88 / 101
89 / 101
90 / 101
91 / 101
92 / 101
93 / 101
94 / 101
95 / 101
96 / 101
97 / 101
98 / 101
99 / 101
100 / 101
101 / 101
Why Haskell? Susan Potter March 2012
What is Haskell? Figure: "pure functional, lazy, polymorphic, statically and strongly typed with type inference . . . "
How can I drive this thing? Figure: Photo from "If programming languages were cars" blog post http://machinegestalt.posterous.com/if-programming-languages-were-cars
Can I drive Haskell without all this? Figure: No need to know Category Theory proofs, just some intuitions!
# finger $(whoami) Login: susan Name: Susan Potter Directory: /home/susan Shell: /bin/zsh Practicing since 1997-09-29 21:18 (GMT) on tty1 from :0 Too much unread mail on me@susanpotter.net Now working at Desk.com! Looking for smart developers!;) Plan: github: mbbx6spp twitter: @SusanPotter
Are we "doing it wrong"? Figure: Maybe! ;) http://absolutelymadness.tumblr.com/post/17567574522
Overview: Choosing a language Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
Overview: Choosing a language Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
Overview: Choosing a language Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
Overview: Choosing a language Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
Overview: Agenda My Claims / Hypotheses Laziness, Functional, Type System Toolkit & Runtime Library Ecosystem Pitfalls & Hurdles
My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
Haskell "lazy" by default Figure: Photo by Mark Fischer http://www.flickr.com/photos/tom_ruaat/4431626234/
Haskell "lazy" by default Figure: Photo by Mark Fischer http://www.flickr.com/photos/tom_ruaat/4431626234/ (jarring for mainstream programmers)
Example: doubleSum 3 (2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Example: doubleSum 3 (2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name (evaluates outer-most expressions first) doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
Laziness in Haskell is . . . CallByName + SharingOptimization + PossibleMinorOverhead
Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
Laziness: Also Pretty Suuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
Laziness: Also Pretty Suuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
Laziness: Also Pretty Suuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
Laziness: Also Pretty Suuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
Purity + Laziness=> Reasoning Equality (referential transparency) Can replace occurrences of LHS with RHS Higher Order Functions encourage exploitation of higher-level patterns Function Composition leads to greater reuse
Purity + Laziness=> Reasoning Equality (referential transparency) Can replace occurrences of LHS with RHS Higher Order Functions encourage exploitation of higher-level patterns Function Composition leads to greater reuse
Purity + Laziness=> Reasoning Equality (referential transparency) Can replace occurrences of LHS with RHS Higher Order Functions encourage exploitation of higher-level patterns Function Composition leads to greater reuse
mysum :: Num a => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
mysum :: Num a => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
mysum :: Num a => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
mysum :: Num a => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
mysum :: Num a => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
mysum :: Num a => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
mysum :: Num a => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
class Monoid m where mappend :: m -> m -> m mempty :: m instance Num a => Monoid a where mappend :: m -> m -> m mappend x y = (+) x y mempty :: m mempty = 0 instance Monoid Bool where mappend :: m -> m -> m mappend True _ = True mappend _ True = True mappend _ _ = False mempty :: m
class Monoid m where mappend :: m -> m -> m mempty :: m instance Num a => Monoid a where mappend :: m -> m -> m mappend x y = (+) x y mempty :: m mempty = 0 instance Monoid Bool where mappend :: m -> m -> m mappend True _ = True mappend _ True = True mappend _ _ = False mempty :: m
class Monoid m where mappend :: m -> m -> m mempty :: m instance Num a => Monoid a where mappend :: m -> m -> m mappend x y = (+) x y mempty :: m mempty = 0 instance Monoid Bool where mappend :: m -> m -> m mappend True _ = True mappend _ True = True mappend _ _ = False mempty :: m
Haskell type signatures can . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
Haskell type signatures can . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
Haskell type signatures can . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
Haskell type signatures can . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
Haskell type signatures can . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
Interfaces in OO . . . Figure: Class definitions are married to the interfaces they implement.
Interfaces in Haskell: Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
Interfaces in Haskell: Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
Interfaces in Haskell: Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
Interfaces in Haskell: Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
Interfaces in Haskell: Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
Interfaces in Haskell: Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
class (Eq a) => Ord a where compare :: a -> a -> Ordering compare x y | x == y = EQ | x <= y = LT | otherwise = GT (<), (>), (>=), (<=) :: a -> a -> Bool ... max, min :: a -> a -> a ...
Typically this just works . . . data SimpleShape = Square { size :: Double } | Circle { radius :: Double } deriving (Eq, Ord, Show) We explicitly use the default definitions . . . and when it doesn’t . . . instance Ord SimpleShape where ...
Are you awake? Figure: http://absolutelymadness.tumblr.com/post/18126913457
Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
Haskell Tooling: Runtime Reasonably performant between JVM 7 and C# Mono performance GC settings easily customized Numerous other runtime options
Haskell Tooling: Runtime Reasonably performant between JVM 7 and C# Mono performance GC settings easily customized Numerous other runtime options
Haskell Tooling: Runtime Reasonably performant between JVM 7 and C# Mono performance GC settings easily customized Numerous other runtime options
Haskell Tooling: Tools Testing tools QuickCheck, HUnit Documentation tools Haddock, Hoogle (lookup documentation) Build tools Cabal, cabal-dev, cabal-nirvana, see "next slide"
Haskell Tooling: Tools Testing tools QuickCheck, HUnit Documentation tools Haddock, Hoogle (lookup documentation) Build tools Cabal, cabal-dev, cabal-nirvana, see "next slide"
Haskell Tooling: Tools Testing tools QuickCheck, HUnit Documentation tools Haddock, Hoogle (lookup documentation) Build tools Cabal, cabal-dev, cabal-nirvana, see "next slide"
Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
Haskell Tooling: Don’ts for Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
Haskell Tooling: Don’ts for Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
Haskell Tooling: Don’ts for Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
Haskell Tooling: Don’ts for Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
Haskell Tooling: Suggestions Explicit language extensions Intentionally and explicitly enable per module Sandbox your builds with cabal-dev or similar Think in types and shapes and use Hoogle to lookup based on types and function "shapes"
Haskell Tooling: Suggestions Explicit language extensions Intentionally and explicitly enable per module Sandbox your builds with cabal-dev or similar Think in types and shapes and use Hoogle to lookup based on types and function "shapes"
Haskell Tooling: Suggestions Explicit language extensions Intentionally and explicitly enable per module Sandbox your builds with cabal-dev or similar Think in types and shapes and use Hoogle to lookup based on types and function "shapes"
Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
Questions? Figure: http://www.flickr.com/photos/42682395@N04/ @SusanPotter
Questions? Figure: http://www.flickr.com/photos/42682395@N04/ @SusanPotter
Bonus: References / Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
Bonus: References / Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
Bonus: References / Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
Bonus: References / Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
Bonus: References / Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
Bonus: Brief QuickCheck Example module Tests where import Test.QuickCheck (quickCheck) propReverseReverse :: [Char] -> Bool propReverseReverse s = (reverse . reverse) s == s excuse the weird syntax form, indenting didn’t show up ;( main = do {quickCheck propReverseReverse }

More Related Content

PDF
Haskell for data science
byJohn Cant
 
PDF
Functional Algebra: Monoids Applied
bySusan Potter
 
PDF
High-Performance Haskell
byJohan Tibell
 
PDF
Functional Programming by Examples using Haskell
bygoncharenko
 
PDF
Beginning Haskell, Dive In, Its Not That Scary!
bypriort
 
PDF
Haskell in the Real World
byosfameron
 
PDF
Building a website in Haskell coming from Node.js
byNicolas Hery
 
PPTX
Scala Back to Basics: Type Classes
byTomer Gabel
 
Haskell for data science
byJohn Cant
 
Functional Algebra: Monoids Applied
bySusan Potter
 
High-Performance Haskell
byJohan Tibell
 
Functional Programming by Examples using Haskell
bygoncharenko
 
Beginning Haskell, Dive In, Its Not That Scary!
bypriort
 
Haskell in the Real World
byosfameron
 
Building a website in Haskell coming from Node.js
byNicolas Hery
 
Scala Back to Basics: Type Classes
byTomer Gabel
 

What's hot

PDF
Humble introduction to category theory in haskell
byJongsoo Lee
 
PDF
Pragmatic Real-World Scala (short version)
byJonas Bonér
 
PDF
Functional programming with haskell
byfaradjpour
 
PDF
Haskell
byRoberto Casadei
 
PDF
Pune Clojure Course Outline
byBaishampayan Ghose
 
PDF
First-Class Patterns
byJohn De Goes
 
PDF
Swift rocks! #1
byHackraft
 
PDF
Design Patterns - Compiler Case Study - Hands-on Examples
byGanesh Samarthyam
 
PDF
Quark: A Purely-Functional Scala DSL for Data Processing & Analytics
byJohn De Goes
 
PDF
Reasoning about laziness
byJohan Tibell
 
PDF
Swift Rocks #2: Going functional
byHackraft
 
PDF
Python Cheat Sheet
byMuthu Vinayagam
 
PDF
Hey! There's OCaml in my Rust!
byKel Cecil
 
PDF
"Немного о функциональном программирование в JavaScript" Алексей Коваленко
byFwdays
 
PPTX
Joy of scala
byMaxim Novak
 
PDF
Coding Guidelines - Crafting Clean Code
byGanesh Samarthyam
 
PDF
Demystifying functional programming with Scala
byDenis
 
PDF
Quicksort - a whistle-stop tour of the algorithm in five languages and four p...
byPhilip Schwarz
 
PPSX
What's New In C# 7
byPaulo Morgado
 
PDF
Procedural Programming: It’s Back? It Never Went Away
byKevlin Henney
 
Humble introduction to category theory in haskell
byJongsoo Lee
 
Pragmatic Real-World Scala (short version)
byJonas Bonér
 
Functional programming with haskell
byfaradjpour
 
Haskell
byRoberto Casadei
 
Pune Clojure Course Outline
byBaishampayan Ghose
 
First-Class Patterns
byJohn De Goes
 
Swift rocks! #1
byHackraft
 
Design Patterns - Compiler Case Study - Hands-on Examples
byGanesh Samarthyam
 
Quark: A Purely-Functional Scala DSL for Data Processing & Analytics
byJohn De Goes
 
Reasoning about laziness
byJohan Tibell
 
Swift Rocks #2: Going functional
byHackraft
 
Python Cheat Sheet
byMuthu Vinayagam
 
Hey! There's OCaml in my Rust!
byKel Cecil
 
"Немного о функциональном программирование в JavaScript" Алексей Коваленко
byFwdays
 
Joy of scala
byMaxim Novak
 
Coding Guidelines - Crafting Clean Code
byGanesh Samarthyam
 
Demystifying functional programming with Scala
byDenis
 
Quicksort - a whistle-stop tour of the algorithm in five languages and four p...
byPhilip Schwarz
 
What's New In C# 7
byPaulo Morgado
 
Procedural Programming: It’s Back? It Never Went Away
byKevlin Henney
 

Viewers also liked

PDF
Haskell vs. F# vs. Scala
bypt114
 
PDF
Haskell for the Real World
byBryan O'Sullivan
 
PDF
Haskell - Functional Programming
byGiovane Berny Possebon
 
PDF
Introduction to haskell
byLuca Molteni
 
PDF
Camomile : A Unicode library for OCaml
byYamagata Yoriyuki
 
PDF
Distributed Developer Workflows using Git
bySusan Potter
 
PDF
Dynamo: Not Just For Datastores
bySusan Potter
 
PDF
Writing Bullet-Proof Javascript: By Using CoffeeScript
bySusan Potter
 
PDF
Link Walking with Riak
bySusan Potter
 
PDF
A taste of Functional Programming
byJordan Open Source Association
 
PPTX
Using functional programming within an industrial product group: perspectives...
byAnil Madhavapeddy
 
PDF
From Zero To Production (NixOS, Erlang) @ Erlang Factory SF 2016
bySusan Potter
 
PDF
Ocaml
byJackson dos Santos Olveira
 
PDF
From Zero to Application Delivery with NixOS
bySusan Potter
 
PDF
Ricon/West 2013: Adventures with Riak Pipe
bySusan Potter
 
PDF
Introduction to functional programming using Ocaml
bypramode_ce
 
PPT
Mirage: ML kernels in the cloud (ML Workshop 2010)
byAnil Madhavapeddy
 
PPT
Lisp Programming Languge
byYaser Jaradeh
 
PDF
Designing for Concurrency
bySusan Potter
 
KEY
An Introduction to Functional Programming using Haskell
byMichel Rijnders
 
Haskell vs. F# vs. Scala
bypt114
 
Haskell for the Real World
byBryan O'Sullivan
 
Haskell - Functional Programming
byGiovane Berny Possebon
 
Introduction to haskell
byLuca Molteni
 
Camomile : A Unicode library for OCaml
byYamagata Yoriyuki
 
Distributed Developer Workflows using Git
bySusan Potter
 
Dynamo: Not Just For Datastores
bySusan Potter
 
Writing Bullet-Proof Javascript: By Using CoffeeScript
bySusan Potter
 
Link Walking with Riak
bySusan Potter
 
A taste of Functional Programming
byJordan Open Source Association
 
Using functional programming within an industrial product group: perspectives...
byAnil Madhavapeddy
 
From Zero To Production (NixOS, Erlang) @ Erlang Factory SF 2016
bySusan Potter
 
Ocaml
byJackson dos Santos Olveira
 
From Zero to Application Delivery with NixOS
bySusan Potter
 
Ricon/West 2013: Adventures with Riak Pipe
bySusan Potter
 
Introduction to functional programming using Ocaml
bypramode_ce
 
Mirage: ML kernels in the cloud (ML Workshop 2010)
byAnil Madhavapeddy
 
Lisp Programming Languge
byYaser Jaradeh
 
Designing for Concurrency
bySusan Potter
 
An Introduction to Functional Programming using Haskell
byMichel Rijnders
 

Similar to Why Haskell

PDF
Clojure intro
byBasav Nagur
 
PDF
DEFUN 2008 - Real World Haskell
byBryan O'Sullivan
 
PDF
Functional programming using haskell notes iitk
bybenevolent001
 
PDF
Haskell - Being lazy with class
byTiago Babo
 
PDF
[FLOLAC'14][scm] Functional Programming Using Haskell
byFunctional Thursday
 
PDF
02. haskell motivation
bySebastian Rettig
 
PPT
Functional Programming - Past, Present and Future
byPushkar Kulkarni
 
PPT
Functional Programming Past Present Future
byIndicThreads
 
PDF
Why Haskell Matters
byromanandreg
 
PDF
Introduction to Functional Languages
bysuthi
 
PDF
04. haskell handling
bySebastian Rettig
 
PDF
The other side of functional programming: Haskell for Erlang people
byBryan O'Sullivan
 
PDF
05. haskell streaming io
bySebastian Rettig
 
PDF
Real World Haskell: Lecture 1
byBryan O'Sullivan
 
PDF
Real World Haskell: Lecture 6
byBryan O'Sullivan
 
PPTX
Functional programming seminar (haskell)
byBikram Thapa
 
PPTX
Good functional programming is good programming
bykenbot
 
KEY
Functional programming in clojure
byJuan-Manuel Gimeno
 
PPTX
Столпы функционального программирования для адептов ООП, Николай Мозговой
bySigma Software
 
PPTX
VCE Unit 01 (2).pptx
byskilljiolms
 
Clojure intro
byBasav Nagur
 
DEFUN 2008 - Real World Haskell
byBryan O'Sullivan
 
Functional programming using haskell notes iitk
bybenevolent001
 
Haskell - Being lazy with class
byTiago Babo
 
[FLOLAC'14][scm] Functional Programming Using Haskell
byFunctional Thursday
 
02. haskell motivation
bySebastian Rettig
 
Functional Programming - Past, Present and Future
byPushkar Kulkarni
 
Functional Programming Past Present Future
byIndicThreads
 
Why Haskell Matters
byromanandreg
 
Introduction to Functional Languages
bysuthi
 
04. haskell handling
bySebastian Rettig
 
The other side of functional programming: Haskell for Erlang people
byBryan O'Sullivan
 
05. haskell streaming io
bySebastian Rettig
 
Real World Haskell: Lecture 1
byBryan O'Sullivan
 
Real World Haskell: Lecture 6
byBryan O'Sullivan
 
Functional programming seminar (haskell)
byBikram Thapa
 
Good functional programming is good programming
bykenbot
 
Functional programming in clojure
byJuan-Manuel Gimeno
 
Столпы функционального программирования для адептов ООП, Николай Мозговой
bySigma Software
 
VCE Unit 01 (2).pptx
byskilljiolms
 

More from Susan Potter

PDF
Thinking in Properties
bySusan Potter
 
PDF
Champaign-Urbana Javascript Meetup Talk (Jan 2020)
bySusan Potter
 
PDF
From Zero to Haskell: Lessons Learned
bySusan Potter
 
PDF
Dynamically scaling a political news and activism hub (up to 5x the traffic i...
bySusan Potter
 
PDF
Functional Operations (Functional Programming at Comcast Labs Connect)
bySusan Potter
 
PDF
Scalaz By Example (An IO Taster) -- PDXScala Meetup Jan 2014
bySusan Potter
 
PDF
Twitter4R OAuth
bySusan Potter
 
PDF
Deploying distributed software services to the cloud without breaking a sweat
bySusan Potter
 
Thinking in Properties
bySusan Potter
 
Champaign-Urbana Javascript Meetup Talk (Jan 2020)
bySusan Potter
 
From Zero to Haskell: Lessons Learned
bySusan Potter
 
Dynamically scaling a political news and activism hub (up to 5x the traffic i...
bySusan Potter
 
Functional Operations (Functional Programming at Comcast Labs Connect)
bySusan Potter
 
Scalaz By Example (An IO Taster) -- PDXScala Meetup Jan 2014
bySusan Potter
 
Twitter4R OAuth
bySusan Potter
 
Deploying distributed software services to the cloud without breaking a sweat
bySusan Potter
 

Recently uploaded

PPTX
UiPath Autonomous Agents | Building and Orchestrating Agents End-to-End
byUiPathCommunity
 
PDF
Day 5 - Red Team + Blue Team in the Cloud - 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
PPT
software-security-intro in information security.ppt
byismaeelsahib032
 
PDF
Our Digital Tribe_ Cultivating Connection and Growth in Our Slack Community 🌿...
bysanjeetmishra30
 
PDF
The year in review - MarvelClient in 2025
bypanagenda
 
PDF
Cybersecurity: Safeguarding Digital Assets
byYasir Naveed Riaz
 
PPTX
communication-skills-with-technology tools
byJaleto Sunkemo
 
PDF
Dev Dives: AI that builds with you - UiPath Autopilot for effortless RPA & AP...
byUiPathCommunity
 
PDF
Six Shifts For 2026 (And The Next Six Years)
byDavid Armano
 
PDF
Making Sense of Raster: From Bit Depth to Better Workflows
bySafe Software
 
PPTX
AI in Cybersecurity: Digital Defense by Yasir Naveed Riaz
byYasir Naveed Riaz
 
PDF
DevFest El Jadida 2025 - Product Thinking
byElmehdi AMLOU
 
PDF
Lab 1.5 - Sharing Secrets with GPG ~ 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
PPTX
Conversational Agents – Building Intelligent Assistants [Virtual Hands-on Wor...
byUiPathCommunity
 
PDF
Unlocking the Power of Salesforce Architecture: Frameworks for Effective Solu...
byvarsha30tiwari
 
PDF
The major tech developments for 2026 by Pluralsight, a research and training ...
byChris Skinner
 
PDF
TrustArc Webinar - Looking Ahead: The 2026 Privacy Landscape
byTrustArc
 
PDF
Eredità digitale sugli smartphone: cosa resta di noi nei dispositivi mobili
bySpeck&Tech
 
PDF
Day 3 - Data and Application Security - 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
PPTX
AI's Impact on Cybersecurity - Challenges and Opportunities
byYasir Naveed Riaz
 
UiPath Autonomous Agents | Building and Orchestrating Agents End-to-End
byUiPathCommunity
 
Day 5 - Red Team + Blue Team in the Cloud - 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
software-security-intro in information security.ppt
byismaeelsahib032
 
Our Digital Tribe_ Cultivating Connection and Growth in Our Slack Community 🌿...
bysanjeetmishra30
 
The year in review - MarvelClient in 2025
bypanagenda
 
Cybersecurity: Safeguarding Digital Assets
byYasir Naveed Riaz
 
communication-skills-with-technology tools
byJaleto Sunkemo
 
Dev Dives: AI that builds with you - UiPath Autopilot for effortless RPA & AP...
byUiPathCommunity
 
Six Shifts For 2026 (And The Next Six Years)
byDavid Armano
 
Making Sense of Raster: From Bit Depth to Better Workflows
bySafe Software
 
AI in Cybersecurity: Digital Defense by Yasir Naveed Riaz
byYasir Naveed Riaz
 
DevFest El Jadida 2025 - Product Thinking
byElmehdi AMLOU
 
Lab 1.5 - Sharing Secrets with GPG ~ 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
Conversational Agents – Building Intelligent Assistants [Virtual Hands-on Wor...
byUiPathCommunity
 
Unlocking the Power of Salesforce Architecture: Frameworks for Effective Solu...
byvarsha30tiwari
 
The major tech developments for 2026 by Pluralsight, a research and training ...
byChris Skinner
 
TrustArc Webinar - Looking Ahead: The 2026 Privacy Landscape
byTrustArc
 
Eredità digitale sugli smartphone: cosa resta di noi nei dispositivi mobili
bySpeck&Tech
 
Day 3 - Data and Application Security - 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
AI's Impact on Cybersecurity - Challenges and Opportunities
byYasir Naveed Riaz
 

Why Haskell

  • 1.
    Why Haskell? Susan Potter March 2012
  • 2.
    What is Haskell? Figure: "pure functional, lazy, polymorphic, statically and strongly typed with type inference . . . "
  • 3.
    How can Idrive this thing? Figure: Photo from "If programming languages were cars" blog post http://machinegestalt.posterous.com/if-programming-languages-were-cars
  • 4.
    Can I driveHaskell without all this? Figure: No need to know Category Theory proofs, just some intuitions!
  • 5.
    # finger $(whoami) Login:susan Name: Susan Potter Directory: /home/susan Shell: /bin/zsh Practicing since 1997-09-29 21:18 (GMT) on tty1 from :0 Too much unread mail on me@susanpotter.net Now working at Desk.com! Looking for smart developers!;) Plan: github: mbbx6spp twitter: @SusanPotter
  • 6.
    Are we "doingit wrong"? Figure: Maybe! ;) http://absolutelymadness.tumblr.com/post/17567574522
  • 7.
    Overview: Choosing alanguage Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
  • 8.
    Overview: Choosing alanguage Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
  • 9.
    Overview: Choosing alanguage Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
  • 10.
    Overview: Choosing alanguage Many considerations political, human, technical Runtime performance, reliability, configurability Knowledge culture, mindshare, resources, signal to noise ratio Tooling development, build/release, configuration, deployment
  • 11.
    Overview: Agenda My Claims / Hypotheses Laziness, Functional, Type System Toolkit & Runtime Library Ecosystem Pitfalls & Hurdles
  • 12.
    My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
  • 13.
    My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
  • 14.
    My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
  • 15.
    My Claims Performance is reasonable on par with Java and C# Mono; 2-3x CPU times of C (approx); BUT always doubt benchmarks http://shootout.alioth.debian.org/u64q/haskell.php Productivity with long-term benefits after initial steep learning curve Haskell types offer stronger verifiability strong and meaningful checks applied Pure functional code is easier to test probably not controversial
  • 16.
    Haskell "lazy" bydefault Figure: Photo by Mark Fischer http://www.flickr.com/photos/tom_ruaat/4431626234/
  • 17.
    Haskell "lazy" bydefault Figure: Photo by Mark Fischer http://www.flickr.com/photos/tom_ruaat/4431626234/ (jarring for mainstream programmers)
  • 18.
    Example: doubleSum 3(2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 19.
    Example: doubleSum 3(2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 20.
    Example: doubleSum 3(2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 21.
    Example: doubleSum 3(2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 22.
    Example: doubleSum 3(2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 23.
    Example: doubleSum 3(2 + 3) Call by value doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 24.
    Example: doubleSum 3(2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 25.
    Example: doubleSum 3(2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 26.
    Example: doubleSum 3(2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 27.
    Example: doubleSum 3(2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 28.
    Example: doubleSum 3(2 + 3) Call by value (evaluates inner-most expressions first) doubleSum 3 (2 + 3) → doubleSum 3 5 → 2 * (3 + 5) → 2 * 8 → 16 Call by name (evaluates outer-most expressions first) doubleSum 3 (2 + 3) → (λ x -> 2 * (3 + x)) (2 + 3)
  • 29.
    Laziness in Haskellis . . . CallByName + SharingOptimization + PossibleMinorOverhead
  • 30.
    Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
  • 31.
    Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
  • 32.
    Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
  • 33.
    Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
  • 34.
    Laziness: Buyer Beware filter (λ x → x < 6) [1..] (never terminates) takeWhile (λ x → x < 6) [1..] (does terminate) dropWhile (λ x → x >= 6) [1..] (does not terminate) Need to understand implications of laziness on functions Laziness with I/O implications lazy I/O with handles is problematic, but iteratee idiom solves most of these. There are a number of libraries available to help with this: enumerator, pipes, . . .
  • 35.
    Laziness: Also PrettySuuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
  • 36.
    Laziness: Also PrettySuuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
  • 37.
    Laziness: Also PrettySuuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
  • 38.
    Laziness: Also PrettySuuweeeet! Infinite sequences/lists made possible Recursive functions become practical Recursive types become simple Much more as well ...
  • 39.
    Purity + Laziness=>Reasoning Equality (referential transparency) Can replace occurrences of LHS with RHS Higher Order Functions encourage exploitation of higher-level patterns Function Composition leads to greater reuse
  • 40.
    Purity + Laziness=>Reasoning Equality (referential transparency) Can replace occurrences of LHS with RHS Higher Order Functions encourage exploitation of higher-level patterns Function Composition leads to greater reuse
  • 41.
    Purity + Laziness=>Reasoning Equality (referential transparency) Can replace occurrences of LHS with RHS Higher Order Functions encourage exploitation of higher-level patterns Function Composition leads to greater reuse
  • 42.
    mysum :: Numa => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
  • 43.
    mysum :: Numa => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
  • 44.
    mysum :: Numa => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
  • 45.
    mysum :: Numa => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
  • 46.
    mysum :: Numa => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
  • 47.
    mysum :: Numa => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
  • 48.
    mysum :: Numa => [a] -> a mysum xs = foldr (+) 0 xs myproduct :: Num a => [a] -> a myproduct xs = foldr (*) 1 xs myany :: (a -> Bool) -> [a] -> Bool myany pred xs = foldr ( x b -> b || pred x) False xs myall :: (a -> Bool) -> [a] -> Bool myall pred xs = foldr ( x b -> b && pred x) True xs
  • 49.
    class Monoid mwhere mappend :: m -> m -> m mempty :: m instance Num a => Monoid a where mappend :: m -> m -> m mappend x y = (+) x y mempty :: m mempty = 0 instance Monoid Bool where mappend :: m -> m -> m mappend True _ = True mappend _ True = True mappend _ _ = False mempty :: m
  • 50.
    class Monoid mwhere mappend :: m -> m -> m mempty :: m instance Num a => Monoid a where mappend :: m -> m -> m mappend x y = (+) x y mempty :: m mempty = 0 instance Monoid Bool where mappend :: m -> m -> m mappend True _ = True mappend _ True = True mappend _ _ = False mempty :: m
  • 51.
    class Monoid mwhere mappend :: m -> m -> m mempty :: m instance Num a => Monoid a where mappend :: m -> m -> m mappend x y = (+) x y mempty :: m mempty = 0 instance Monoid Bool where mappend :: m -> m -> m mappend True _ = True mappend _ True = True mappend _ _ = False mempty :: m
  • 52.
    Haskell type signaturescan . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
  • 53.
    Haskell type signaturescan . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
  • 54.
    Haskell type signaturescan . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
  • 55.
    Haskell type signaturescan . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
  • 56.
    Haskell type signaturescan . . . express side effects e.g. String -> IO Int declare computational strategies e.g. Num a => [a] -> Sum a impose constraints e.g. Num a => a -> a question value availability e.g. String -> Maybe Int verify client-server protocol dialogs? an exercise for reader ;)
  • 57.
    Interfaces in OO. . . Figure: Class definitions are married to the interfaces they implement.
  • 58.
    Interfaces in Haskell:Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
  • 59.
    Interfaces in Haskell:Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
  • 60.
    Interfaces in Haskell:Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
  • 61.
    Interfaces in Haskell:Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
  • 62.
    Interfaces in Haskell:Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
  • 63.
    Interfaces in Haskell:Typeclasses. . . Decouple type definition from interface Allow upstream implementations Extend thirdparty libraries easily Redefine implementations upstream No meaningless "any type" functions Very flexible
  • 64.
    class (Eq a)=> Ord a where compare :: a -> a -> Ordering compare x y | x == y = EQ | x <= y = LT | otherwise = GT (<), (>), (>=), (<=) :: a -> a -> Bool ... max, min :: a -> a -> a ...
  • 65.
    Typically this justworks . . . data SimpleShape = Square { size :: Double } | Circle { radius :: Double } deriving (Eq, Ord, Show) We explicitly use the default definitions . . . and when it doesn’t . . . instance Ord SimpleShape where ...
  • 66.
    Are you awake? Figure: http://absolutelymadness.tumblr.com/post/18126913457
  • 67.
    Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
  • 68.
    Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
  • 69.
    Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
  • 70.
    Haskell Tooling: Libraries Quite a few Practical libraries Often freely available Permissive OSS licenses
  • 71.
    Haskell Tooling: Runtime Reasonably performant between JVM 7 and C# Mono performance GC settings easily customized Numerous other runtime options
  • 72.
    Haskell Tooling: Runtime Reasonably performant between JVM 7 and C# Mono performance GC settings easily customized Numerous other runtime options
  • 73.
    Haskell Tooling: Runtime Reasonably performant between JVM 7 and C# Mono performance GC settings easily customized Numerous other runtime options
  • 74.
    Haskell Tooling: Tools Testing tools QuickCheck, HUnit Documentation tools Haddock, Hoogle (lookup documentation) Build tools Cabal, cabal-dev, cabal-nirvana, see "next slide"
  • 75.
    Haskell Tooling: Tools Testing tools QuickCheck, HUnit Documentation tools Haddock, Hoogle (lookup documentation) Build tools Cabal, cabal-dev, cabal-nirvana, see "next slide"
  • 76.
    Haskell Tooling: Tools Testing tools QuickCheck, HUnit Documentation tools Haddock, Hoogle (lookup documentation) Build tools Cabal, cabal-dev, cabal-nirvana, see "next slide"
  • 77.
    Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
  • 78.
    Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
  • 79.
    Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
  • 80.
    Haskell Tooling: Dependency Management Hackage database of freely available Haskell libraries Cabal great to get started, BUT . . . cabal-dev & similar provides sandboxing, list RVM with gemsets; more important for statically typed environments cabal-nirvana think compatible distribution snapshot of Hackage DB
  • 81.
    Haskell Tooling: Don’tsfor Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
  • 82.
    Haskell Tooling: Don’tsfor Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
  • 83.
    Haskell Tooling: Don’tsfor Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
  • 84.
    Haskell Tooling: Don’tsfor Newbies Use GHC (not HUGS) Hugs written for educational purposes not industrial usage Forget what you know (imperative/OO) relearn programming in a functional-style return is a function name it does not mean return in the C/Java/C# way class does not mean OO-class think decoupled interface with optional default impelementations and a lot more power
  • 85.
    Haskell Tooling: Suggestions Explicit language extensions Intentionally and explicitly enable per module Sandbox your builds with cabal-dev or similar Think in types and shapes and use Hoogle to lookup based on types and function "shapes"
  • 86.
    Haskell Tooling: Suggestions Explicit language extensions Intentionally and explicitly enable per module Sandbox your builds with cabal-dev or similar Think in types and shapes and use Hoogle to lookup based on types and function "shapes"
  • 87.
    Haskell Tooling: Suggestions Explicit language extensions Intentionally and explicitly enable per module Sandbox your builds with cabal-dev or similar Think in types and shapes and use Hoogle to lookup based on types and function "shapes"
  • 88.
    Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
  • 89.
    Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
  • 90.
    Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
  • 91.
    Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
  • 92.
    Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
  • 93.
    Oh, the possibilities! Parallel / Concurrency Options threads, dataflow, par, seq "Cloud" Haskell A kind of Erlang/OTP clone in Haskell Data Parallel Haskell GHC extensions to support nested data parallelism accounting, "Nepal" Haskell’s Foreign Function Interface (FFI) Interface with native code from Haskell GPU Programming in Haskell Obsidian, Nikola, GpuGen, numerous papers on this too Much more. . . Research meeting industrial application
  • 94.
    Questions? Figure: http://www.flickr.com/photos/42682395@N04/ @SusanPotter
  • 95.
    Questions? Figure: http://www.flickr.com/photos/42682395@N04/ @SusanPotter
  • 96.
    Bonus: References /Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
  • 97.
    Bonus: References /Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
  • 98.
    Bonus: References /Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
  • 99.
    Bonus: References /Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
  • 100.
    Bonus: References /Resources Channel 9 Lectures (Erik Meijer) http://channel9.msdn.com/Shows/Going+Deep/ Lecture-Series-Erik-Meijer-Functional-Programming-Fundamentals-Chapter-1 Learn You A Haskell http://learnyouahaskell.com Haskell Reddit http://www.reddit.com/r/haskell/ Haskell Cafe http://www.haskell.org/mailman/listinfo/haskell-cafe Real World Haskell http://book.realworldhaskell.org/
  • 101.
    Bonus: Brief QuickCheckExample module Tests where import Test.QuickCheck (quickCheck) propReverseReverse :: [Char] -> Bool propReverseReverse s = (reverse . reverse) s == s excuse the weird syntax form, indenting didn’t show up ;( main = do {quickCheck propReverseReverse }