- Haskell-Cafe - Haskell.org

Job opening: further Haskell Developers at Anchor
by Andrew Cowie 08 Jul '14

08 Jul '14

Openings for additional Haskell developers... About Anchor ------------ Anchor Systems, based in Sydney Australia, has successfully been providing managed hosting and infrastructure to clients around the world for almost 15 years. The critical difference in the service we offer our customers often boils down to providing operations expertise and infrastructure configuration balancing the short term needs to get a new installation running with the long term imperatives of security, scalability, and flexibility. The engineering department is building the analytics capabilities, internal systems, cloud infrastructure, deployment tooling, and operations practices that will be needed to take the company to the next level. We have every intention of changing the hosting industry to make developing applications — and operating them at scale — easier for people facing massive growth. Haskell is the working language for all internal development, and we're exploring it for web facing code too. The benefits of type safety, resilience when refactoring, and stability over time added to the power of functional programming has already paid dividends. We have a significant programme of work ahead, and we're looking to add further capability as we face these challenges. Interested in helping? Requirements ------------ You'll need to have demonstrable experience programming in Haskell, familiarity with open source development practices, and the ability to tolerate other people refactoring your code. Exposure to the harsh realities of IT operations and experience of the thrill of Linux systems administration will stand you in good stead. Knowledge of the history, progress, and problems in configuration management will be helpful. A history of contributing to libraries on Hackage would be relevant. More relevant would be a track record of working well in the open (regardless of language or project). Good documentation is as important as good code (a type signature is not sufficient), and command of written English is necessary both to collaborate with your peers and to describe and promote your results. A Bachelor's in science, engineering, or xenobiology will be well thought of. Your specific field of study certainly doesn't matter, but how you use computing to solve problems in your chosen field does. A degree in engineering will be a distinct asset. An undergraduate degree in computer science will not be held against you. Applying -------- If you're interested, send a PDF copy of your CV to <jobs(a)anchor.com.au>, along with the URL to your GitHub account. We also invite you to include a pointer to something that shows you in a good light: a mailing list thread where you argued a contentious position, a piece of particularly thorny code in which you found an elegant solution to a problem, or even an essay or article you've written. We don't do international relocations (sorry everyone who wants to work remotely from Hawaii; I want to work remotely from Hawaii too), but we readily sponsor work visas for new hires should they need one. So whether you've just moved to Australia or have been here for generations, give us a shout next time you're in Sydney. AfC Sydney -- Andrew Frederick Cowie Head of Engineering Anchor Systems 201 Elizabeth Street Level 11, Sydney NSW 2000, Australia http://www.anchor.com.au | afcowie(a)anchor.com.au | +61 2 8296 5133

1 0

strange behavior of GHC 7.8.2
by Kazu Yamamoto 08 Jul '14

08 Jul '14

Hi cafe, I noticed that strange behavior of GHC 7.8.2. Consider the following example which requires the "ghc" package: ---- {-# LANGUAGE GeneralizedNewtypeDeriving #-} module A where import GHC import MonadUtils import Control.Monad.Trans.RWS.Lazy (RWST(..)) newtype M a = M (RWST () () () IO a) deriving (Functor,Applicative,Monad,MonadIO) ---- The "ghc" library depends on "transformers" 0.3.0.0. If "transformers" *0.4.1.0* is NOT installed, GHCi can handle the code above well: ---- % ghci -package ghc A.hs ... Loading package transformers-0.3.0.0 ... linking ... done. ... Ok, modules loaded: A. [*A] > ---- However, if "transformers" *0.4.1.0* is installed, an error happens: ---- % cabal install transformers % ghci -package ghc A.hs ... Loading package transformers-0.3.0.0 ... linking ... done. ... [1 of 1] Compiling A ( A.hs, interpreted ) A.hs:10:49: No instance for (MonadIO (RWST () () () IO)) arising from the 'deriving' clause of a data type declaration Possible fix: use a standalone 'deriving instance' declaration, so you can specify the instance context yourself When deriving the instance for (MonadIO M) Failed, modules loaded: none. [Prelude] > ---- A you can see, "transformers" 0.3.0.0 is certainly linked. How can I interpret this behavior? If this is a bug of GHC 7.8.2, I will file this to GHC's trac. P.S. We noticed this because "doctest" of ghc-mod fails only for GHC 7.8.2. --Kazu

3 5

withFile variant which takes a ByteString?
by Bardur Arantsson 08 Jul '14

08 Jul '14

Hi all, Is there a variant of withFile which takes a ByteString? (I'm aware that I could use the lower-level System.Posix FD-based functions, but if someone already has a package out there, I'd rather not.) Regards,

3 2

Checking minimum package bounds
by Omari Norman 07 Jul '14

07 Jul '14

Often a topic of discussion is making sure that the upper bounds on one's packages are up to date, or even whether upper bounds are a good idea at all. I don't see much discussion on lower bounds. Lower bounds that are needlessly restrictive can get in the way too. For instance, I could just slap "text >= 1.1 && < 1.2" on my package, but this makes things needlessly hard for someone trying to make my package work in conjunction with an older one, especially if all I use are basic functions like "pack" that have been around at least since, say. text-0.11. Does anyone have a best practice for testing and verifying lower bounds? Testing the lower bounds and the upper ones is a challenge. Does anybody bother? I have my sunlight package at http://hackage.haskell.org/package/sunlight but it is really a nasty hack. I am considering cleaning it up so it is less of a nasty hack but before I do that I wondered if anyone else has ever thought about this problem. Thanks. Omari

7 8

ML Family workshop: First Call for Participation
by oleg＠okmij.org 07 Jul '14

07 Jul '14

Higher-order, Typed, Inferred, Strict: ACM SIGPLAN ML Family Workshop Thursday September 4, 2014, Gothenburg, Sweden Call For Participation http://okmij.org/ftp/ML/ML14.html Early registration deadline is August 3. Please register at https://regmaster4.com/2014conf/ICFP14/register.php This workshop specifically aims to recognize the entire extended ML family and to provide the forum to present and discuss common issues, both practical (compilation techniques, implementations of concurrency and parallelism, programming for the Web) and theoretical (fancy types, module systems, metaprogramming). We also encourage presentations from related languages (such as Scala, Rust, Nemerle, ATS, etc.), to exchange experience of further developing ML ideas. The workshop is conducted in close cooperation with the OCaml Users and Developers Workshop http://ocaml.org/meetings/ocaml/2014/ taking place on September 5. Program * Andreas Rossberg 1ML -- core and modules as one (Or: F-ing first-class modules) * Jacques Garrigue and Leo White Type-level module aliases: independent and equal * Felix Klock and Nicholas Matsakis Demo: The Rust Language and Type System * Tomas Petricek and Don Syme Doing web-based data analytics with F# * Thomas Braibant, Jonathan Protzenko and Gabriel Scherer Well-typed generic smart-fuzzing for APIs * Ramana Kumar, Magnus O. Myreen, Michael Norrish and Scott Owens Improving the CakeML Verified ML Compiler * Leo White and Frederic Bour Modular implicits * Nada Amin and Tiark Rompf Implicits in Practice * Anil Madhavapeddy, Thomas Gazagnaire, David Scott and Richard Mortier Metaprogramming with ML modules in the MirageOS * Katsuhiro Ueno and Atsushi Ohori Compiling SML# with LLVM: a Challenge of Implementing ML on a Common Compiler Infrastructure * Akinori Abe and Eijiro Sumii A Simple and Practical Linear Algebra Library Interface with Static Size Checking * John Reppy SML3d: 3D Graphics for Standard ML In addition, the joint poster session with the OCaml workshop will take place in the afternoon on September 5. The session will include posters: * Nicolas Oury Core.Sequence: a unified interface for sequences * Thomas Gazagnaire, Amir Chaudhry, Anil Madhavapeddy, Richard Mortier, David Scott, David Sheets, Gregory Tsipenyuk, Jon Crowcroft Irminsule: a branch-consistent distributed library database * Michel Mauny and Benoit Vaugon Nullable Type Inference * Edwin Toeroek LibreS3: design, challenges, and steps toward reusable libraries * Fabrice Le Fessant A Case for Multi-Switch Constraints in OPAM Program Committee Kenichi Asai Ochanomizu University, Japan Matthew Fluet Rochester Institute of Technology, USA Jacques Garrigue Nagoya University, Japan Dave Herman Mozilla, USA Stefan Holdermans Vector Fabrics, Netherlands Oleg Kiselyov (Chair) University of Tsukuba, Japan Keiko Nakata Tallinn University of Technology, Estonia Didier Remy INRIA Paris-Rocquencourt, France Zhong Shao Yale University, USA Hongwei Xi Boston University, USA

1 0

Control.Exception.bracket is broken(?)
by Yuras Shumovich 07 Jul '14

07 Jul '14

Hello, >From docs: "If an exception is raised, then bracket will re-raise the exception (after performing the release)." But what if the release action will throw an exception? There are 3 options: 1) suppress the exception from release action and rethrow the original one 2) shutdown the program 3) suppress the original exception and throw the exception from release action. Current behavior is #3 -- suppress the original exception. I think it the worse choice. What other languages do: - java has try-with-resources statement which is the nearest equivalent of haskell's "bracket". It uses the option #1 (rethrow the original exception) - in c++ it is common to use RAII for resource management, so someone's destructor performs resource cleanup. But if destructor throws exception during stack unrolling, the program is terminated (option #2) I prefer #2 because suppressing exception is a really bad idea. What is your choice? And the related thing. Release action should not use any interruptible action. Otherwise async exception can occur while processing other exception, so one of them will be suppressed. Is it correct? Control.Exception module contains a list of uninterruptible actions (see "Interruptible operations" section), and e.g. "hClose" is not there. Is "hClose" interruptible or not? If async exception interrupts "hClose", the file descriptor will leak. So, should we use uninterruptibleMask in release action for "bracket"? Thanks, Yuras

3 4

how to turn on RTS options?
by Johan Larson 06 Jul '14

06 Jul '14

I need to change some runtime options. I updated the options line in my cabal file: ghc-options: -Wall -fno-warn-orphans -O2 -rtsopts But when I rebuild and run, it's as though runtime options are still disabled. C:\johan\Haskell\projects\bio-acai>cabal build .\bio-acai.cabal has been changed. Re-configuring with most recently used options. If this fails, please run configure manually. Resolving dependencies... Configuring bio-acai-0.1.0.0... Building bio-acai-0.1.0.0... Preprocessing executable 'acai' for bio-acai-0.1.0.0... Preprocessing test suite 'tests' for bio-acai-0.1.0.0... C:\johan\Haskell\projects\bio-acai>dist\build\acai\acai.exe +RTS -H128m -K64m *acai.exe: Most RTS options are disabled. Link with -rtsopts to enable them.* My best guess is that the "ghc-options" line is the wrong place to try to turn on RTS option handling. But what's the right one? -- Johan Larson -- Toronto, Canada

2 2

[ANN] codex 0.1.0 - generate tags file from dependencies
by Alois Cochard 06 Jul '14

06 Jul '14

Hi, Just to let you know that I released a new version of `codex` (a tool which generate a tags[1] file for a given cabal project using the sources of all the dependencies of that project). This release include some bug fixes as well as two new features: - Support modules added using `cabal sandbox add-source` as workspace dependencies - Support `emacs` format by disabling sorting on merge of tags files Thanks to everyone who reported bugs or submitted patch! [1]Those tags file basically contain references to functions/types definition in source code and allow "jump to definition" like functionality in text editors. source: http://github.com/aloiscochard/codex hackage: http://hackage.haskell.org/package/codex -- *A\ois* http://twitter.com/aloiscochard http://github.com/aloiscochard

1 0

async simplify code
by grant weyburne 06 Jul '14

06 Jul '14

Hi Cafe, Is there any way to simplify this async code to somehow merge the withAsync code into the STM code? Looking at the code for tstABorCD it is not easy to see that it is really just (A&&B)||(C&&D). Here is the code: http://lpaste.net/106945 Thanks for any pointers, Grant

4 3

Restarting doc build on hackage
by Erik Hesselink 05 Jul '14

05 Jul '14

Hi cafe, Does anyone know if there is a way to restart a documentation build on hackage? We have a package where the documentation failed to build due to a dependency lacking upper bounds. That has recently been fixed (using the new in-place cabal file editing feature), and we'd like to get documentation for our package now. Is there any way to do this (apart from uploading a new version without changes)? Erik

5 6