March 2015 - Haskell-Cafe

[ANN] relational-record - relational-algebraic query building DSL
by Kei Hibino 16 May '15

16 May '15

Hello everybody, I am happy to announce relational-record library and its project page. relational-record is domain specific language for type-safe SQL query building, and database access API with compile time schema generators. Code is avaiable on github and hackage. https://github.com/khibino/haskell-relational-record/ http://hackage.haskell.org/package/relational-record/ Tutorial, examples and documentations are available from project page links. http://khibino.github.io/haskell-relational-record/ Supported features are below, * Type safe query building. * Composable query. * Outer join. * Aggregation. * Compile time compile code generation along with RDBMS schema. * Schema support of ProsgreSQL, SQLite3, MySQL, IBM DB2, MicroSoft SQL Server, and Oracle. Any comments, any questions about this library are welcome. -- Kei Hibino ex8k.hibino at gmail.com https://github.com/khibino/

6 12

Semantics of temporal data
by martin 14 Apr '15

14 Apr '15

Hello all, I wrote a Temporal datatype data Temporal a = Temporal { td :: a, -- "temporal default" tc :: [Change a] -- "temporal changes" } deriving (Eq, Show) data Change a = Chg { ct :: Time, -- "change time" cv :: a -- "change value" } deriving (Eq,Show) I got into semantic trouble when trying to write "join" for Temporal (Temporal a). If I have a Temporal Temporal, then each change's value is a Temporal. The question is: what significance does the temporal default have? In my implementation, when I look at this Chg t0 (Temporal default0 changes0) Chg t1 (Temporal default1 changes1) --<-- you are here Chg t2 (Temporal default2 changes2) I accept the changes1 which are >= t1 and less than t2. So far so good. But I also take the default1 and prepend it to the resuling list of changes with a change time of t1 (unless the first of the changes1 occurs also at t1). But by comparing results with a handwritten <*>, QuickCheck revealed that this is not in line with <*> = ap x >>= f = join $ fmap f x It appears to me that the defaults of the inner Temporal lose most of their meaning. The effective default is the change value of the last accepted change from the previous (t0) iteration. This would make some sense. If I have a summer schedule which lists the location of a single train over time, and it sais it is usually "nowhere", but at July 1st is shall be in Manchester. If I ask this schedule, where the train will be before July 1st it'll say "nowhere". Now if I prepend the summer schedule with a spring schedule, then there will be a last location of that train in the spring schedule (say Brighton), i.e. before the summer schedule becomes effective. Hence at the beginning of the summer, the train won't be "nowhere" but in Brighton. Does this make some sense? Please feel free to comment. BTW: I am really impressed that haskell pointed out this semantic issue to me. This could have gone unnoticed for months in other languages.

3 4

indexed writer monad
by Nicholls, Mark 10 Apr '15

10 Apr '15

Anyone? I can handle monads, but I have something (actually in F#) that feels like it should be something like a indexed writer monad (which F# probably wouldn't support). So I thought I'd do some research in Haskell. I know little or nothing about indexed monad (though I have built the indexed state monad in C#). So I would assume there would be an indexed monoid (that looks at bit like a tuple?)... e.g. (a,b) ++ (c,d) = (a,b,c,d) (a,b,c,d) ++ (e) = (a,b,c,d,e) ? There seems to be some stuff about "update monads", but it doesn't really look like a writer. I could do with playing around with an indexed writer, in order to get my head around what I'm doing....then try and capture what I'm doing...then try (and fail) to port it back. CONFIDENTIALITY NOTICE This e-mail (and any attached files) is confidential and protected by copyright (and other intellectual property rights). If you are not the intended recipient please e-mail the sender and then delete the email and any attached files immediately. Any further use or dissemination is prohibited. While MTV Networks Europe has taken steps to ensure that this email and any attachments are virus free, it is your responsibility to ensure that this message and any attachments are virus free and do not affect your systems / data. Communicating by email is not 100% secure and carries risks such as delay, data corruption, non-delivery, wrongful interception and unauthorised amendment. If you communicate with us by e-mail, you acknowledge and assume these risks, and you agree to take appropriate measures to minimise these risks when e-mailing us. MTV Networks International, MTV Networks UK & Ireland, Greenhouse, Nickelodeon Viacom Consumer Products, VBSi, Viacom Brand Solutions International, Be Viacom, Viacom International Media Networks and VIMN and Comedy Central are all trading names of MTV Networks Europe. MTV Networks Europe is a partnership between MTV Networks Europe Inc. and Viacom Networks Europe Inc. Address for service in Great Britain is 17-29 Hawley Crescent, London, NW1 8TT.

3 12

MinGHC for GHC 7.10.1 available
by Michael Snoyman 07 Apr '15

07 Apr '15

I've just uploaded a new release of MinGHC, including GHC 7.10.1 and cabal-install 1.22.2.0. This release can be downloaded from: https://s3.amazonaws.com/download.fpcomplete.com/minghc/minghc-7.10.1-i386.… In the process, I also needed to upload a cabal-install binary for Windows, which is available at: https://s3.amazonaws.com/download.fpcomplete.com/minghc/cabal-install-1.22.… I've tested this distribution, but only lightly. Feedback from others would be useful :)

2 1

Generalizing "unlift" functions with monad-control
by Michael Snoyman 05 Apr '15

05 Apr '15

I'm trying to extract an "unlift" function from monad-control, which would allow stripping off a layer of a transformer stack in some cases. It's easy to see that this works well for ReaderT, e.g.: {-# LANGUAGE RankNTypes #-} {-# LANGUAGE TypeFamilies #-} import Control.Monad.Trans.Control import Control.Monad.Trans.Reader newtype Unlift t = Unlift { unlift :: forall n b. Monad n => t n b -> n b } askRun :: Monad m => ReaderT r m (Unlift (ReaderT r)) askRun = liftWith (return . Unlift) The reason this works is that the `StT` associated type for `ReaderT` just returns the original type, i.e. `type instance StT (ReaderT r) m a = a`. In theory, we should be able to generalize `askRun` to any transformer for which that applies. However, I can't figure out any way to express that generalized type signature in a way that GHC accepts it. It seems like the following should do the trick: askRunG :: ( MonadTransControl t , Monad m , b ~ StT t b ) => t m (Unlift t) askRunG = liftWith (return . Unlift) However, I get the following error message when trying this: foo.hs:11:12: Occurs check: cannot construct the infinite type: b0 ~ StT t b0 The type variable ‘b0’ is ambiguous In the ambiguity check for the type signature for ‘askRunG’: askRunG :: forall (t :: (* -> *) -> * -> *) (m :: * -> *) b. (MonadTransControl t, Monad m, b ~ StT t b) => t m (Unlift t) To defer the ambiguity check to use sites, enable AllowAmbiguousTypes In the type signature for ‘askRunG’: askRunG :: (MonadTransControl t, Monad m, b ~ StT t b) => t m (Unlift t) Adding AllowAmbiguousTypes to the mix provides: foo.hs:17:30: Could not deduce (b1 ~ StT t b1) from the context (MonadTransControl t, Monad m, b ~ StT t b) bound by the type signature for askRunG :: (MonadTransControl t, Monad m, b ~ StT t b) => t m (Unlift t) at foo.hs:(12,12)-(16,25) ‘b1’ is a rigid type variable bound by the type forall (n1 :: * -> *) b2. Monad n1 => t n1 b2 -> n1 (StT t b2) at foo.hs:1:1 Expected type: Run t -> Unlift t Actual type: (forall (n :: * -> *) b. Monad n => t n b -> n b) -> Unlift t Relevant bindings include askRunG :: t m (Unlift t) (bound at foo.hs:17:1) In the second argument of ‘(.)’, namely ‘Unlift’ In the first argument of ‘liftWith’, namely ‘(return . Unlift)’ I've tested with both GHC 7.8.4 and 7.10.1. Any suggestions?

3 7

Is it possible for cabal sandboxes to use $HOME/.cabal/lib?
by Aldo Davide 04 Apr '15

04 Apr '15

Hi all, There are some packages that I use again and again and I would like to install them once (but not in a system-wide location) and then have cabal sandboxes use that install, instead of installing them in every sandbox, which is time consuming. Is there a way do that? For example, I noticed that cabal, when operating inside a sandbox, can still find the packages installed system-wide, but not the ones in $HOME/.cabal/lib. Can I change that? Thanks

5 8

Parallel Profiling
by Dominic Steinitz 02 Apr '15

02 Apr '15

Does anyone know of any tools for analysing parallel program performance? I am trying to use threadscope but it keeps crashing with my 100M log file and ghc-events-analyze is not going to help as I have many hundreds of threads all carrying out the same computation. I think I’d like a library that would allow me to construct my own analyses rather than display them via GTK. There is ghc-events but that seems to be just for parsing the logs and I couldn’t find anything that used it in the way I would like to (apart from threadscope and ghc-events-analyze of course). Thanks Dominic Steinitz dominic(a)steinitz.org http://idontgetoutmuch.wordpress.com

2 2

Enforcing data structures invariant in the type system
by Nicola Gigante 31 Mar '15

31 Mar '15

Hi all. I’m very interested in what can be done with Haskell’s type system regarding the enforcing of algorithms and data structures invariants. For example, I remember I saw a paper about an implementation of a red-black tree in Haskell where the types guaranteed the invariant about the alternation of rad and black nodes. I would like to learn those techniques, in particular the use of language features that enable this kind of things and then how to do them in practice. In the past tried hacking something with GADTs and Data Kinds but I've always got stuck with GHC errors that I could not understand, so I think I’m missing something. Where could I start from? Are there simple walk-through examples and/or fundational material? For example is it possible to write a sorting function that guarantees in the types that the output is sorted? I don’t need this specific example but you got the point. Thank you very much, Nicola

5 9

Data.EnumMap ?
by Johannes Waldmann 31 Mar '15

31 Mar '15

Dear Cafe, in containers, we have Data.IntMap/IntSet . Because of the underlying implementation (patricia trees), keys/elements indeed have to be Int (we need their bits). This is a conflict with type-safety: I try to avoid Int, and use some newtype T = T Int instead. E.g., I want to distinguish several different indices for the same structure, like row and column numbers in (sparse) matrices, or vertices in bipartite graphs. So, I made these trivial wrappers EnumMap and EnumSet: https://github.com/jwaldmann/satchmo-solver/tree/master/src/Data Is this reasonable (do from/toEnum really have zero cost)? Would this be a useful addition to containers? - J.W.

4 5

Hackage trustee proposal: Curating the Hackage package collection
by Adam Bergmark 31 Mar '15

31 Mar '15

Dear Haskell Community, For some time Hackage has contained a user group called "Trustees", http://hackage.haskell.org/packages/trustees/ . Description: The role of trustees is to help to curate the whole package collection. Trustees have a limited ability to edit package information, for the entire package database (as opposed to package maintainers who have full control over individual packages). Trustees can edit .cabal files, edit other package metadata and upload documentation but they cannot upload new package versions." In short, making sure that packages keep building and filling the gap between unreachable maintainers and package take-overs. Up until now we have been very careful with changes since we haven't had a defined process. Spurred by SPJ and others we have been working on a proposal for how we should operate. You can find the proposal here: https://gist.github.com/bergmark/76cafefb300546e9b90e We would now like your feedback! Some specific things from the proposal that we'd like your opinion on: * Section 1: No opt-out for restricting bounds * Section 2: Opt-out rather than opt-in procedure for loosening version constraints * Section 2: Do you care whether you are notified before or after a version constraint is loosened? * Section 3: The time frame for publishing simple source changes * Section 3: What exactly should constitute a "simple source change" We also have a github repository where YOU can file issues about broken packages, you can start doing this right now! https://github.com/haskell-infra/hackage-trustees/ Please share this with as many people as possible. We are looking forward to hear your thoughts! Sincerely, Adam Bergmark

6 8