September 2014 - Haskell-Cafe

Control.Monad.Error deprecated?
by Karl Voelker 15 Sep '14

15 Sep '14

Hello everyone, I just discovered that Control.Monad.Error [1] has recently become deprecated in favor of Control.Monad.Except [2], and I am curious about what motivated the change. Can anyone explain? Thanks, Karl 1: http://hackage.haskell.org/package/mtl-2.2.1/docs/Control-Monad-Error.html 2: http://hackage.haskell.org/package/mtl-2.2.1/docs/Control-Monad-Except.html

3 2

ANN: haskell-src-exts-1.16.0
by Roman Cheplyaka 13 Sep '14

13 Sep '14

I am pleased to announce the release of haskell-src-exts-1.16.0! Hackage: https://hackage.haskell.org/package/haskell-src-exts-1.16.0 GitHub: https://github.com/haskell-suite/haskell-src-exts This release brings many bugfixes, as well as the following language features: * DoRec * Closed type families * GADT records * ExplicitNamespaces * Type equality constraints * PolyKinds * DataKinds * Default associated types * Instance signatures * OVERLAP pragma * Parallel arrays * MINIMAL pragma * Default signatures * Safe Haskell * Binary literals * Qualified record puns * Type splices The full changelog is available at https://github.com/haskell-suite/haskell-src-exts/blob/1.16.0/CHANGELOG This release is brought to you by: Alan Zimmerman Erik de Castro Lopo klapaucius Lars Corbijn Leonid Onokhov Peter A. Jonsson phischu Roman Cheplyaka Stefan Wehr Stijn van Drongelen yuriy0 Special thanks to Peter Jonsson who has done an incredible job for this release and fixed many of the longstanding issues in haskell-src-exts. Roman

1 0

policy for new packages
by Corentin Dupont 13 Sep '14

13 Sep '14

Hi guys, In my software I developped some piece of code that I think are sufficiently generic to be useful for others (some datatypes and typeclasses). Should I put them in a separate package and annonce them? What is the best pratice here? Namely it's a datatype called Data.Todo allowing to keep track of actions left to do to complete a computation, and the other is a typeclass called Control.Shortcut designed for computations that can be run in parralel and shortcuted if necessary. So they are small (but useful) things, I'm wondering if creating one package for each is worth it, or another policy can apply. Cheers, Corentin

1 0

XML <--> Haskell Data Structure using HXT
by Han Joosten 11 Sep '14

11 Sep '14

I would like to write code to enable importing data to and fro from a Haskell data structure. It makes sense to have some XSD or so that specifies how the XML should be structured. The HDS is already in place. I have been looking into hte HXT library and I think it could be used for this job. However, I am not familiar (yet) with Arrows and the like. Also the use of picklers is new to me. It would help me a lot if someone has a neat example of how coversion between XML and some Haskell data structure can be implemented in het HXT-style. Do such examples exist? Where can I find them? For those interested: Here is the Haskell Data Structure: https://sourceforge.net/p/ampersand/code/HEAD/tree/trunk/src/lib/Database/D… Cheers & Thanks for reading! Han Joosten

3 3

understanding cloud Haskell serialization
by felipe zapata 11 Sep '14

11 Sep '14

Dear Cafe, I'm trying to understand the implementation of Cloud Haskell, specially the closures. According to the documentation, the serialization of a function uses a map from Label to values (RemoteTable), where Remotable is implemented as: newtype RemoteTable = RemoteTable (Map String Dynamic) Where Dynamic is given by, data Dynamic = Dynamic TypeRep GHC.Any The question is then, How does this Dynamic data type encodes functions? And how is related this Dynamic type to the decoder function? closure :: Static (ByteString -> a) -- Decoder -> ByteString -- Encoded closure environment -> Closure a Static provides two function, one function related to the label to lookup at the RemoteTable and one used for composition. But how is this related to Dynamic? Any guide will be appreciated. Felipe Z.

2 1

HaLVM Talk in NYC on Monday
by Adam Wick 11 Sep '14

11 Sep '14

Howdy - For those of you in the New York City area, I’ll be giving a talk this coming Monday, the 15th, on Galois’s use of the HaLVM — the Haskell Lightweight Virtual Machine — over the last many years. The talk ("Unikernels: Who, What, Where, When, Why") will be part of the Xen User Summit at the Lighthouse Executive Conference Center: http://events.linuxfoundation.org/events/xen-project-user-summit/program/sc… I believe you can get 50% off using the offer code “XenUser50off”. The talk will mostly about the general concept of unikernels like the HaLVM, and how we’ve used them well (and poorly!) on various projects, and so will be light on technical details. But if you’d like to chat with me about any aspect of the HaLVM, stop on by, and we can talk functional languages at a system’s conference. It’ll be great. - Adam

2 1

Cabal doesn't reset the search-path when invoking GHC: potential subtleties
by Phil Ruffwind 11 Sep '14

11 Sep '14

It seems that even when invoked via Cabal, GHC will look in the local directory *first* and then in the list of modules and dependencies supplied by Cabal. I was aware of this behavior subconsciously but I hadn't realized the implications until now. Normally, it's not a big problem, other than for the occasional situation where I forget to add those local modules to 'other-modules'. This would cause 'cabal sdist' to generate an incomplete package, even though 'cabal build && cabal install' works just fine! But there are more subtle "bugs": it breaks if the executable and library share the same directory, but the library has some hidden modules stashed away in a different directory! The conditions required to reproduce this is somewhat difficult to explain, so let me show a simple example: 1. Make a Cabal package: -- ./foo.cabal name: foo version: 0.0.0.0 build-type: Simple cabal-version: >=1.8 library exposed-modules: Exposed other-modules: Hidden build-depends: base hs-source-dirs: ., hidden executable main main-is: Main.hs build-depends: base, foo 2. Add a library with two modules at the given paths: -- ./Exposed.hs module Exposed where import Hidden -- ./hidden/Hidden.hs module Hidden where 3. Add an executable: -- ./Main.hs module Main where import Exposed main = return () 4. Now run `cabal build`. This would cause an error message akin to: > Could not find module Hidden The problem is that despite the dependency on the library `foo` being specified in Cabal, GHC still chose the *local* `Exposed` module over the `Exposed` module in the library `foo`. This means when it tries to recompile the `Exposed` module, it is unaware that `Hidden` is located in another directory. The error message is unhelpful because it doesn't point to the actual cause of the problem and led me astray thinking that something was wrong with my Cabal configuration. In the end, the problem has to do with GHC deciding to look up packages in locations outside Cabal's control. There are many other ways to trigger this problem -- the above was only one that I just happened to run into. For now, a workaround would be to use `-i` with all the executables to avoid this subtle issue in the future. RF

2 1

Categorical description for vector-space instance of AdditiveGroup (Maybe a)
by trevor cook 11 Sep '14

11 Sep '14

Hi Haskell Cafe, In Vector-space-0.8.7, there is the cited inconsistency of the definition of AdditiveGroup (Maybe a). I've posted the source, including comment, below for reference. Following the logic of the counterexample and Conal's (and or Andy Gill's) comment, the problem hinges on the equality, Nothing = Just (zeroV :: a). I'm curious about the catigorical description of this situation. Does this mean that the category of Additive Groups and Maybe Additive Groups are equivalent as categories? I'm not proposing that this is somehow a resolution to vector-space, but rather trying to gain some insight into equivalence, naturalitity, adjunction, etc. My thinking as to why equivalence is that if we have the instance below defined using: zeroV = Just zeroV and making appropriate changes following that new decision. Nothing ^+^ Nothing = zeroV a ^+^ Nothing = a Nothing ^+^ b = b negateV Nothing = zeroV negateV a = fmap negateV then we pretty much defer the group structure of (Maybe a) to be based totally on the structure of a. And since `a` and `Maybe a` are not isomorphic then "hey, I just learned about this new thing so maybe its that." Its a horrible argument, hopefully the cafe will show some better thinking for me. Regards, Trevor -- Maybe is handled like the Maybe-of-Sum monoidinstance AdditiveGroup a => AdditiveGroup (Maybe a) where zeroV = Nothing Nothing ^+^ b' = b' a' ^+^ Nothing = a' Just a' ^+^ Just b' = Just (a' ^+^ b') negateV = fmap negateV{-Alexey Khudyakov wrote: I looked through vector-space package and found lawless instance. Namely Maybe's AdditiveGroup instance It's group so following relation is expected to hold. Otherwise it's not a group. > x ^+^ negateV x == zeroV Here is counterexample: > let x = Just 2 in x ^+^ negateV x == zeroV False I think it's not possible to sensibly define group instance for Maybe a at all.I see that the problem here is in distinguishing 'Just zeroV' fromNothing. I could fix the Just + Just line to use Nothing instead of JustzeroV when a' ^+^ b' == zeroV, although doing so would require Eq a andhence lose some generality. Even so, the abstraction leak would probablyshow up elsewhere.Hm.-}

4 4

Haskell Weekly News: Issue 304
by Daniel Santa Cruz 10 Sep '14

10 Sep '14

Welcome to issue 305 of the HWN, an issue covering crowd-sourced bits of information about Haskell from around the web. This issue covers from August 31 to September 06, 2014 Quotes of the Week * srhb: segfaults sure are easier to come by doing FFI stuff Top Reddit Stories * Pandoc author working with industry on Markdown spec Domain: standardmarkdown.com, Score: 68, Comments: 7 Original: [1] http://goo.gl/JBte8k On Reddit: [2] http://goo.gl/LWT42a * ICFP 2014 Talks Domain: youtube.com, Score: 56, Comments: 0 Original: [3] http://goo.gl/2zRte4 On Reddit: [4] http://goo.gl/NvWBBa * ICFP 2014 Workshops (Videos) Domain: youtube.com, Score: 53, Comments: 6 Original: [5] http://goo.gl/gpfI8T On Reddit: [6] http://goo.gl/Bj5SwB * Let's Build a Browser Engine in Haskell Domain: hrothen.github.io, Score: 45, Comments: 18 Original: [7] http://goo.gl/FkOM49 On Reddit: [8] http://goo.gl/rXM38x * Open type families are not modular Domain: blog.ezyang.com, Score: 41, Comments: 52 Original: [9] http://goo.gl/XO1bVK On Reddit: [10] http://goo.gl/ROxxUA * IDE for Haste projects Domain: haskell-web.blogspot.com.es, Score: 36, Comments: 5 Original: [11] http://goo.gl/uvjZUC On Reddit: [12] http://goo.gl/3Ra4dt * Real world one-liner examples Domain: github.com, Score: 33, Comments: 7 Original: [13] http://goo.gl/E5wFiW On Reddit: [14] http://goo.gl/9gipjb * Planning Yesod 1.4 Domain: yesodweb.com, Score: 32, Comments: 8 Original: [15] http://goo.gl/7Kul02 On Reddit: [16] http://goo.gl/erCRmZ * Practical Machines: A simple tutorial for the Machines streaming library Domain: statusfailed.com, Score: 32, Comments: 24 Original: [17] http://goo.gl/bmRQot On Reddit: [18] http://goo.gl/0QSdOL * Anecdotes and Advice on Hiring Junior or Summer/Co-Op Students for Haskell Work Domain: self.haskell, Score: 32, Comments: 29 Original: [19] http://goo.gl/M7NOJO On Reddit: [20] http://goo.gl/M7NOJO * Write your compiler passes in Coq for great good! Domain: megacz.com, Score: 30, Comments: 13 Original: [21] http://goo.gl/l7MU6n On Reddit: [22] http://goo.gl/oSqi3Z * IntelliJ IDEA haskell-idea-plugin + Hasklig Domain: self.haskell, Score: 28, Comments: 3 Original: [23] http://goo.gl/cNWkLh On Reddit: [24] http://goo.gl/cNWkLh * Upcast, a declarative cloud infrastructure orchestration tool that leverages Nix Domain: github.com, Score: 27, Comments: 12 Original: [25] http://goo.gl/vPpXcn On Reddit: [26] http://goo.gl/MHj0Le * Ray Marching Distance Fields with Haskell and GLSL Domain: github.com, Score: 27, Comments: 13 Original: [27] http://goo.gl/HZSjFq On Reddit: [28] http://goo.gl/w0IZ0w * A Pretty-Printer that Says What It Means (or, Why Idris's Pretty-Printer is a Functor) Domain: davidchristiansen.dk, Score: 25, Comments: 14 Original: [29] http://goo.gl/vfdISz On Reddit: [30] http://goo.gl/mHU49x * The Ivory Language is an eDSL for safe systems programming. You can think of Ivory as a safer C, embedded in Haskell. Domain: ivorylang.org, Score: 23, Comments: 3 Original: [31] http://goo.gl/xJJ68c On Reddit: [32] http://goo.gl/RLljxg * Pre-published the slide "Writing NetBSD Sound Drivers in Haskell" for Haskell Symposium 2014. #icfp2014 Domain: slideshare.net, Score: 23, Comments: 4 Original: [33] http://goo.gl/gsENVq On Reddit: [34] http://goo.gl/mWH3zH * Haskell 2014 talk: Reflection without Remorse: Revealing a hidden sequence to speed up monadic reflection Domain: youtube.com, Score: 23, Comments: 8 Original: [35] http://goo.gl/Dffhnx On Reddit: [36] http://goo.gl/UFPUu3 * Wadler's Blog: Howard on Curry-Howard Domain: wadler.blogspot.com, Score: 21, Comments: 0 Original: [37] http://goo.gl/Fz90FG On Reddit: [38] http://goo.gl/P5kemv Top StackOverflow Questions * Infinite lazy bitmap votes: 11, answers: 2 Read on SO: [39] http://goo.gl/4zDwFX * Correct terminology for continuations votes: 11, answers: 4 Read on SO: [40] http://goo.gl/QHVRWI * How do I make MonadRandom a Functor? votes: 9, answers: 1 Read on SO: [41] http://goo.gl/vvD7bR * Why does Haskell's foldr NOT stackoverflow while the same Scala implementation does? votes: 9, answers: 4 Read on SO: [42] http://goo.gl/zdxgAM Until next time, [43]+Daniel Santa Cruz References 1. http://standardmarkdown.com/ 2. http://www.reddit.com/r/haskell/comments/2fgtuu/pandoc_author_working_with_… 3. https://www.youtube.com/playlist?list=PL4UWOFngo5DVBqifWX6ZlXJ7idxWaOP69 4. http://www.reddit.com/r/haskell/comments/2f8o3x/icfp_2014_talks/ 5. http://www.youtube.com/playlist?list=PL4UWOFngo5DXuUiMCNumrFhaDfMx54QgV 6. http://www.reddit.com/r/haskell/comments/2f60hd/icfp_2014_workshops_videos/ 7. http://hrothen.github.io/2014/09/05/lets-build-a-browser-engine-in-haskell/ 8. http://www.reddit.com/r/haskell/comments/2flosm/lets_build_a_browser_engine… 9. http://blog.ezyang.com/2014/09/open-type-families-are-not-modular/ 10. http://www.reddit.com/r/haskell/comments/2fi4zc/open_type_families_are_not_… 11. http://haskell-web.blogspot.com.es/2014/09/ide-for-haste-projects.html 12. http://www.reddit.com/r/haskell/comments/2fk91p/ide_for_haste_projects/ 13. https://github.com/sjoerdvisscher/blog/blob/master/2014/2014-09-03%20real%2… 14. http://www.reddit.com/r/haskell/comments/2feen3/real_world_oneliner_example… 15. http://www.yesodweb.com/blog/2014/09/planning-yesod-1-4 16. http://www.reddit.com/r/haskell/comments/2f4kym/planning_yesod_14/ 17. http://statusfailed.com/blog/2014/09/02/practical-machines-in-60-seconds.ht… 18. http://www.reddit.com/r/haskell/comments/2fao00/practical_machines_a_simple… 19. http://www.reddit.com/r/haskell/comments/2fj0v5/anecdotes_and_advice_on_hir… 20. http://www.reddit.com/r/haskell/comments/2fj0v5/anecdotes_and_advice_on_hir… 21. http://www.megacz.com/berkeley/coq-in-ghc/ 22. http://www.reddit.com/r/haskell/comments/2f8bu9/write_your_compiler_passes_… 23. http://www.reddit.com/r/haskell/comments/2fb1j5/intellij_idea_haskellideapl… 24. http://www.reddit.com/r/haskell/comments/2fb1j5/intellij_idea_haskellideapl… 25. https://github.com/zalora/upcast 26. http://www.reddit.com/r/haskell/comments/2f2meq/upcast_a_declarative_cloud_… 27. https://github.com/blitzcode/ray-marching-distance-fields 28. http://www.reddit.com/r/haskell/comments/2ffmys/ray_marching_distance_field… 29. http://www.davidchristiansen.dk/2014/09/06/pretty-printing-idris/ 30. http://www.reddit.com/r/haskell/comments/2fntec/a_prettyprinter_that_says_w… 31. http://ivorylang.org/ 32. http://www.reddit.com/r/haskell/comments/2f6p2h/the_ivory_language_is_an_ed… 33. http://www.slideshare.net/master_q/writing-netbsd-sound-drivers-in-haskell 34. http://www.reddit.com/r/haskell/comments/2fd02w/prepublished_the_slide_writ… 35. https://www.youtube.com/watch?v=_XoI65Rxmss&index=31&list=UUP9g4dLR7xt6KzCY… 36. http://www.reddit.com/r/haskell/comments/2fnj6p/haskell_2014_talk_reflectio… 37. http://wadler.blogspot.com/2014/08/howard-on-curry-howard.html 38. http://www.reddit.com/r/haskell/comments/2f9v8e/wadlers_blog_howard_on_curr… 39. http://stackoverflow.com/questions/25621889/infinite-lazy-bitmap 40. http://stackoverflow.com/questions/25648332/correct-terminology-for-continu… 41. http://stackoverflow.com/questions/25622571/how-do-i-make-monadrandom-a-fun… 42. http://stackoverflow.com/questions/25622959/why-does-haskells-foldr-not-sta… 43. https://plus.google.com/105107667630152149014/about

1 0

Re: [Haskell-cafe] api-tools questions
by Karsten Gebbert 10 Sep '14

10 Sep '14

Chris, thank you for the pointer, that is really helpful indeed! I see that you are using Haskell types for some fields, rather than types defined in your API. When I, for instance, try to use Map as field type, with String as both key and value type, I get an error: myrec :: MyRec // a map = record myfield :: Map String String Exception when trying to run compile-time code: Syntax error at line 345, column 20 (@8596): [(AlexPn 8596 345 20,TypeIden "String"),(AlexPn 8603 345 27,TypeIden "String")] I suspect this is only the case with types that are parameterized over others, as a type synonym seems to make this possible anyways. Is there a better way to deal with this than the one I mentioned? Thanks for your help, karsten Chris Dornan <chris(a)chrisdornan.com> writes: > You might also find the keystore[0] package useful as it makes use of > This technique in several places. See, for example the definition of > Pattern in the `Data.Keystore.Types.Schema`[1] and > `Data.Keystore.Types`[2]. > > Chris > > [0] http://hackage.haskell.org/package/keystore > [1] > https://github.com/cdornan/keystore/blob/master/src/Data/KeyStore/Types/Sch > ema.hs#L179 > [2] > https://github.com/cdornan/keystore/blob/master/src/Data/KeyStore/Types.hs# > L68 > [3] http://hackage.haskell.org/package/keystore-0.5.0.4 > > > On 09/09/2014 17:20, "Karsten Gebbert" <k(a)ioctl.it> wrote: > >>Excellent, thanks a lot Adam. I'll open a PR when I'm through with this >>project :) >> >>Adam Gundry <adam(a)well-typed.com> writes: >> >>> Hi Karsten, >>> >>> This isn't very well documented, but there is a (hidden) feature of >>> api-tools that should do what you want. If you say something like >>> >>> mt :: MyTime = basic string with inj_MyTime, prj_MyTime >>> >>> then you can define your own type MyTime and give conversion functions >>> inj_MyTime and prj_MyTime to convert back and forth from a >>> newtype-wrapped Text. (In fact, you could probably use UTCTime as >>> MyTime...). More precisely, api-tools will generate something like >>> >>> newtype REP__MyTime = REP__MyTime Text >>> >>> and you will need to implement >>> >>> inj_MyTime :: REP__MyTime -> ParserWithErrs MyTime >>> prj_MyTime :: MyTime -> REP__MyTime >>> >>> I hope this helps, and further questions or documentation contributions >>> are very welcome! >>> >>> Cheers, >>> >>> Adam >>> >>> >>> On 09/09/14 14:02, Karsten Gebbert wrote: >>>> Hi All, >>>> >>>> I have a question concerning the api-tools package. >>>> >>>> Its not clear from the tests, sources or the tutorial how I can use >>>> other date/time formats that the default `utc` type can handle. I'm >>>> trying to wrap a JSON API that is not under my control, so I have to >>>> adhere to whatever I get back from it. Could anybody with experience >>>> with the package point me to some example, relevant bits in the sources >>>> or a tip how to do it? >>>> >>>> I'm planning to create some more documentation around the package to >>>> contribute back once I figured out a few more details, because I think >>>> its quite a useful abstraction when dealing with (foreign) APIs IMO. >>>> >>>> Thanks already for any hints, >>>> >>>> karsten >>> >>> >>> -- >>> Adam Gundry, Haskell Consultant >>> Well-Typed LLP, http://www.well-typed.com/ >>_______________________________________________ >>Haskell-Cafe mailing list >>Haskell-Cafe(a)haskell.org >>http://www.haskell.org/mailman/listinfo/haskell-cafe

1 0