Random Hacks comments

"McCarthy's Ambiguous Operator" by person-b

Fri, 03 Jul 2009 18:32:36 -0400

This reminds me of Damian Conway’s Positronic::Variables CPAN Perl module. He wrote it for his (hilarious – highly recommended) talk on “Temporally Quaquaversal Virtual Nanomachine Programming In Multiple Topologically Connected Quantum-Relativistic Parallel Timespaces…Made Easy!” (http://blip.tv/file/1145545/). It actually did look into the future.

"Write a 32-line chat client using Ruby, AMQP & EventMachine (and a GUI using Shoes)" by TheCheshireCatalyst

Thu, 11 Jun 2009 13:59:21 -0400

Use

PSYC

And here

"How to make Data.Set a monad" by Lemming

Mon, 08 Jun 2009 05:42:44 -0400

Here an interactive link to Monad instance for Data.Set, again on Haskell-Cafe.

"How to make Data.Set a monad" by Lemming

Mon, 08 Jun 2009 05:40:24 -0400

The disadvantage of this approach is, that in a function you need a (Monad2 m a b) for each transition from a type ‘a’ to another type ‘b’. Wolfgang Jeltsch gave a solution on Haskell-Cafe which only needs a constraints like (Monad1 m a, Monad1 m b). It uses type families and existential quantification for packing and unpacking the Ord dictionary.

http://www.haskell.org/pipermail/haskell-cafe/2008-March/041084.html

"8 ways to report errors in Haskell" by andrew u frank

Sat, 06 Jun 2009 07:02:54 -0400

Using ghc 6.10 (with base 4.0.0.0) the module Control.Exception becomes easier to use and seem to be quite flexible.

The only limitation I can see is that the exception raised in pure code (this is possible) cannot be caught in pure code; catching exceptions is only possible in the IO monade. I wonder, if this is a serious restriction and how difficult it would be to ovecome. The same example with division by zero becomes (if properly formated):

import qualified Control.Exception as E

import Data.Typeable

data MyException = DivByZero | ThisEx | ThatEx String deriving (Show, Eq, Typeable)

instance E.Exception MyException

myDiv9 :: (Monad m) => Float -> Float -> m Float

myDiv9 a 0 = E.throw DivByZero

myDiv9 a 99 = E.throw (ThatEx “Fortran style ending”)

myDiv9 a b = return (a / b)

isDivByZero (DivByZero) = True

isDivByZero _ = False

example9 :: Float -> Float -> IO ()

example9 x y = E.catchJust (\e -> if isDivByZero e then Just (e) else Nothing)

(do q <- myDiv9 x y putStrLn (show q) E.throw ThisEx )

(\err -> do putStrLn (“my DivByZero ” ++ show (err:: MyException)) return () )

"8 ways to report errors in Haskell" by andrew u frank

Thu, 04 Jun 2009 11:44:07 -0400

moving from ghc 6.8.2 to 6.10 I cannot recompile the examples given here (especially 1 and 4). What has changed in base 4.0.0.0? The error I get is

Ambiguous type variable `e’ in the constraint: `Exception e’ arising from a use of `E.catch’

anybody can help?

"Jim Hefferon's Linear Algebra: A free textbook with fascinating applications" by Regnessem

Mon, 01 Jun 2009 17:00:40 -0400

“SVD is insanely useful” is true but in very limited sense. There’s plenty of richfull life out of the realm of ortonormal basis. Jim Hefferon really paves the road to move on beyond SVD!

"Map fusion: Making Haskell 225% faster" by Aaron McDaid

Sat, 30 May 2009 16:00:33 -0400

Jim, GHC can’t automatically know that such a rule is valid. For example, (filter f . filter g) is not equivalent to (filter (f.g))

Each author’s implementation of functions called treeMap could be quite different, and it would be difficult to improve GHC such that it could correctly know which implementations could be optimized.

"FTC Spam Archive" by Random Thoughts

Fri, 29 May 2009 01:17:53 -0400

That is interesting. They are even prosecuting spammers it says. GMail is really good at filtering spam, but there’s always a few that get buy. Spamhaus is typically used by the big e-mail providers. There are e-mail servers that get blocked because GMail, Yahoo, etc… query their database before accepting mail. One reason is for the mail client set up in Linux seems like spam if it is from a server hosted by Comcast. They expect all e-mail to be sent through Outlook or the Comcast e-mail site.

"Haskell: Queues without pointers" by Eric

Fri, 22 May 2009 08:13:07 -0400

Mike: Remember, we trying to build a purely functional queue, so we can’t modify any links once we’ve created them. So unless you do something fairly clever with lazy evaluation, you can’t set up the double links.