Nathan Howell

Palo Alto, CA

breaks.for.alienz.org

Age: 34

Oct
10
comment Is it possible to get the Kind of a Type Constructor in Haskell?
typeOf1 and Typeable1 (and friends) are still exported from Data.Typeable... they are usable in 7.8.
Sep
30
awarded Explainer
Jul
23
awarded Enlightened
Jul
23
awarded Nice Answer
Jul
23
revised Plan B, or what's the opposite of Maybe's >>=?
added 177 characters in body
Jul
23
revised Plan B, or what's the opposite of Maybe's >>=?
added 495 characters in body
Jul
23
answered Plan B, or what's the opposite of Maybe's >>=?
Jun
18
awarded Yearling
Jun
18
awarded Yearling
Mar
4
comment Using TChan with Timeout
@RomanCheplyaka good catch, I've never even noticed this was in the stm package.
Mar
4
revised Using TChan with Timeout
deleted 99 characters in body
Mar
4
revised Using TChan with Timeout
added 567 characters in body
Mar
4
answered Using TChan with Timeout
Nov
8
comment What is the cheapest way to compute md5 digest in Haskell?
Not sure what the fastest package is but I've found cryptohash to be fast enough and it covers a decent number of hashes hackage.haskell.org/package/cryptohash
Sep
12
comment ByteString to lazy Text and vice versa
@Yuras seems you're right, but it wouldn't have compiled anyways :^)
Sep
12
answered Type Inference with Reflection and DataKinds
Sep
12
comment Type level arithmetic using kindsignatures and typeoperators?
They're hoping to get some version of the type-nats solver in for GHC 7.8... the work is going on here github.com/ghc/ghc/tree/type-nats-simple
Sep
12
comment ByteString to lazy Text and vice versa
@Yuras: The Data.Text.Lazy.Encoding module deals with Data.ByteString.Lazy.ByteStrings not strict ones, the error message is misleading though.
Sep
12
answered ByteString to lazy Text and vice versa
Sep
10
comment Type Inference with Reflection and DataKinds
@Eric: I agree. I understand why this happens with GADTs or ExistentialQuantification when they're introducing new type variables into scope... this doesn't though and seems like it should work. Hopefully someone chimes in.
1 2 3 4 5