- Haskell-Cafe - Haskell.org

Issue 61 :: Haskell Weekly
by Taylor Fausak 29 Jun '17

29 Jun '17

\ Haskell Weekly \/\/ Issue 61 <https://haskellweekly.news/issues/61.html> Welcome to another issue of Haskell Weekly! Haskell is a purely functional programming language that focuses on robustness, concision, and correctness. This is a weekly summary of what's going on in its community. ## Featured ## - Haxl: Making concurrency unreasonably easy <http://events.techcast.com/bigtechday10/Garmisch-1345/> > Haxl is a big hammer for doing I/O concurrently, testing I/O, and modularity (caching & memoization). - Introduction to Brick <https://samtay.github.io/articles/brick.html> > I'm going to give a short introduction to Brick, a Haskell library for building terminal user interfaces. So far I've used brick to implement Conway's Game of Life and a Tetris clone. I'll explain the basics, walk through an example snake application, and then explain some more complicated scenarios. - Simplexhc: An STG to LLVM compiler <https://pixel-druid.com/blog/announcing-simplexhc/> > STG's lowering to C-- and the decisions taken when implementing it were based in the '90s. That's not to say that the GHC team hasn't done an awesome job keeping it up to date; they have! But, I wonder what a complete rewrite of this lowering would look like. Hence, I'm trying to experiment in this space and see what happens. - Front Row is hiring a senior backend Haskell engineer <https://frontrow.workable.com/j/463B843754> > Come change how 6.5+ million K-12 US students learn Math, Language Arts, Social Studies and more. Use data, advanced type systems, great product design and deep pedagogy to change lives. - Threading responsibly: `forkIO` considered harmful <http://mazzo.li/posts/threads-resources.html> > In this post I want to argue that threads are also a resource, in the sense that they require management after we've created them. A stray thread will consume memory, CPU cycles, and really whatever resource it might need to execute. - The `Has` type class pattern <https://medium.com/@jonathangfischoff/the-has-type-class-pattern-ca12adab70…> > The value in this approach is that I don't have to think about what functions to call to collect the images: it's always `images`. In the prior example, I had to think about the how to collect the images each time, and it took brain power better spent elsewhere. - A tale of two brackets <https://www.fpcomplete.com/blog/2017/06/tale-of-two-brackets> > I hope the primary thing you take away from it is a deeper understanding of how monad transformer stacks interact with operations in the base monad, and how monad-control works in general. - `RecordWildCards` and binary parsing <https://jship.github.io/posts/2017-06-24-record-wildcards-and-binary-parsin…> > `RecordWildCards` is a GHC extension that makes working with Haskell records more convenient. The extension has been blogged about in a few places already, so this post intends to provide a different motivating example: binary parsing. - On naming things: Library design <http://www.parsonsmatt.org/2017/06/23/on_naming_things.html> > I've written a few libraries now and have tried out different naming and exporting conventions. I've developed a bit of a feel for how it is to write and use them, and I'm going to put out my personal preferences and opinions on library design here. - Haskell Bits #6: A guide to mutable references <http://www.kovach.me/posts/2017-06-22-mutable-references.html> > There are quite a few ways to store mutable data in Haskell. Let's talk about some of them! Specifically, we will focus on mutable containers that store a single value that can be modified by one or more threads at any given time. I'm not going to go into a ton of detail here --- I just want to give an overview. - Interfaces and type classes: Number APIs in C# and Haskell <https://mzabani.wordpress.com/2017/06/22/interfaces-and-typeclasses-number-…> > In C# sometimes I sorely miss something like an `INumber<T>` interface with methods add, subtract, multiply, and others. The lack of this means it is cumbersome to write generic code on numbers. ## Package of the week ## This week's package of the week is data-has, a library that provides a simple extensible product system. <https://www.stackage.org/lts-8.20/package/data-has-0.2.1.0> ## In brief ## - Announcing haskus-system 0.7 <http://hsyl20.fr/home/posts/2017-06-29-announcing-haskussystem-07.html> - Defeating evil with data structures <https://mmhaskell.com/blog/2017/6/26/fun-with-data-structures> - stack2nix first public release <https://mailman.science.uu.nl/pipermail/nix-dev/2017-June/024011.html> - Streaming combinators and extracting flat parallelism <https://futhark-lang.org/blog/2017-06-25-futhark-at-pldi.html> - Typing Nix <https://www.tweag.io/posts/2017-05-23-typing-nix.html> - When competing with C, fudge the benchmark <https://medium.com/@n0mad/when-competing-with-c-fudge-the-benchmark-16d3a91…> ## Events ## - Berlin Functional Programming Group's first meeting July 1 <https://www.meetup.com/Berlin-Functional-Programming-Group/events/241040383/> - Haskell.SG's July meetup July 5 <https://www.meetup.com/HASKELL-SG/events/240959693/>

1 0

dsl for atari 2600 vcs?
by Raoul Duke 29 Jun '17

29 Jun '17

Along the lines of atom & ivory & tower, has anyone (I haven't turned up any hits online so I guess not) worked on a dsl in haskell for generating racing-the-beam code? Feels like all the book keeping might be something haskell could do, giving a dsl that is more high level than manual mental cycle counting.

1 0

[ANN] haskus-system 0.7
by Sylvain Henry 29 Jun '17

29 Jun '17

Hi all, I have released haskus-system 0.7. haskus-system <https://github.com/haskus/haskus-system> is a Haskell library for system programming on top of the Linux kernel. You can use it to build systems that only use: * the Linux kernel * the haskus-system library * your code This release features a new build tool that automatically downloads, configures and builds Linux, SysLinux, etc. It can also execute your system within QEMU, build ISO images and even create bootable devices. For instance, to test the Demo example from https://github.com/haskus/haskus-system-examples/ with QEMU, you just have to type: > haskus-system-build test --init Demo The full announce with some screenshots and a video: http://hsyl20.fr/home/posts/2017-06-29-announcing-haskussystem-07.html Cheers, Sylvain

1 0

[ann] hp2pretty-0.8
by Claude Heiland-Allen 29 Jun '17

29 Jun '17

Hi all, hp2pretty is a program to graph heap profiles output by Haskell programs compiled by GHC. New features in this 0.8 release: * --sort size|stddev|name and --reverse to change output order * --trace [percent] and --bands [count] to adjust output bands * --pattern to fill with patterns instead of solid colour The last option should help a lot if you want to print without using too much ink. More information and pictures exhbiting the new features: https://mathr.co.uk/blog/2017-06-29_hp2pretty-0.8_released.html There were also minor changes in the 0.7 release (earlier this week): * parsing bugfix (don't use `read` to drop quotes) thanks to Pepe Iborra * don't stroke text (fill is enough) * use HTTPS in the packaging hp2pretty still misses features from hp2ps including: * monochrome output * multipage (separate page for key) * page size options * header size options * (encapsulated) Postscript support (hp2pretty is still SVG-only) Another feature that might be nice is: * choose font(s) from command line, to be consistent across a document Patches welcome, or even just votes for which features would be most useful to you so I can prioritise my efforts next time I feel like hacking on it. Thanks, Claude -- https://mathr.co.uk

1 0

[ANN] Compose :: Melbourne, 2017, Call for Presentations Extension - Monday, 10th of July
by Lyndon Maydwell 29 Jun '17

29 Jun '17

Hi All! We're extending the Compose :: Melbourne CFP until Monday, 10th of July. Details follow: (Web version: http://www.composeconference.org/2017-melbourne/cfp/) # Compose :: Melbourne Call for Presentations Compose :: Melbourne - http://www.composeconference.org/ - was closing its call for presentations at the end of the month... BUT, now we are extending the CFP until Monday, 10th of July, 2017! Submit your presentation proposal here: https://easychair.org/conferences/?conf=cmc20170 ## About Compose :: Melbourne Compose :: Melbourne is a functional-programming conference returning for its second year which is focused on developing the community and bringing functional-programming to a wider audience. It is a 2-day event being held in Melbourne, Australia on the 28th and 29th of August 2017 at RMIT University. The first day will feature a single track of presentations followed by a second day of workshops and unconference. Compose :: Melbourne is the sister-conference of the NY based Compose :: Conference. # Important Dates - CFP Extended Until - Monday 10th July, 2017 - Notify Presenters - July 20th, 2017 - Conference Day 1: Presentations - Monday, 28th August, 2017 - Conference Day 2: Unconference - Tuesday, 29th August, 2017 # Talk and Workshop Submission ## Day 1 - Presentations Submit a Presentation Proposal via Easy Chair: https://easychair.org/conferences/?conf=cmc20170 ## Day 2 - Unconference and Workshops Contact us (composemel-admin(a)googlegroups.com) with your workshop proposals, or talk-ideas. Although we don't require ahead-of-time talk information, if you tell us about your unconference talk, we will promote it! # Submission Guidelines Talk slots will be 30 minutes: 25 minute talk and 5 minutes for questions. Provide sufficient detail in your submission to enable the reviewers to understand your proposal and clearly identify what an attendee will gain from attending your session. You should include the following information in the submission system: - Authors - The details of each of the presenters - Title - Title of the presentation (please keep it brief and specific) - Abstract - A 300-500 word description of your presentation, ideally including... - Description of the content - Keywords - List any keywords that will help the program committee and attendees categorise your presentation. For example: proofs, Haskell, music. Please keep the content of your talk on-topic and do not use this speaking opportunity as a sales pitch. # Feel We want Compose Melbourne to be all about the community. We're aiming to help foster the growth of Functional Programming in Melbourne and unite all interested parties to spark a unified presence and a feeling of camaraderie amongst FP and Theory proponents in this wonderful city! Talks should help inspire and promote this goal. # Audience Functional-Programming and Programming-Language-Theory professionals and enthusiasts. Newcomers, experts, anyone from other disciplines of fields who is interested in what FP is or how it could help them with their work, or simply make life more enjoyable! # Diversity Just like Compose-Conference, we would also like to put an emphasis on soliciting a diverse set of speakers - anything you can do to distribute information about this CFP and encourage submissions from under-represented groups would be greatly appreciated. We welcome *all* (new and established) contributions and encourage you to apply! # Day One ## Keynote Day One will kick-off with a mind-blowing excursion into the realm of function live-coded music artistry by the living legend, Andrew Sorensen! Look him up on YouTube! ## Talks All topics for proposals for presentations are invited, but not limited to explore the following topics: ## New Languages The development of new and emerging Functional Languages, and the toolsets around them. ## Libraries and Tools Exploring the use and development of new and underrepresented tools and libraries. ## Production Systems 2017 continues the trend of many previously undeployed languages and systems breaking through into production. We're looking for war-stories, success-stories and sob-stories! ## Theory Theory is the cutting edge of new functional programming. Show the world what is rising over the horizon. ## Art and Music Exciting and innovative usage of functional programming language in the arts! # Day Two The second day will be dedicated to workshops and unconference. ## Workshops If you wish to run a workshop on the second day of the conference, then please let us know. Alternatively, if there is a workshop that would interest you as an attendee, then please also let us know. Last year, there were 5 workshops, including: * Intro to Haskell * FreeMonad FreeApplicative FreeWorkshop * Writing games for Android in Haskell * Intro to Purescript * Intro to the J language ## Unconference Got something random you want to talk about? The unconference will run the second day. There will be a whiteboard with spaces where anyone can write their name down to speak. There's no need to bring anything but yourself and your ideas and speaking voice. # Sponsors We are seeking sponsors to support us in putting on this conference. If you or your company would like to sponsor us, please get in contact via composemel-admin(a)googlegroups.com. You can find info on our sponsorship tiers here: https://github.com/composeconference/Compose-Melbourne-2017/wiki/Sponsorshi… -- Compose Conference Melbourne Organising Committee http://www.composeconference.org/

1 0

Any ideas on what this should generate and not generate...?
by Jurriaan Hage 29 Jun '17

29 Jun '17

We had a strange phenomenon while debugging a large compiler. This is the smallest example, but we have seen this behavior in our own monad. module TracEx where import Debug.Trace runit :: IO () runit = traceShow "Bla" $ do traceShow "Bloe” runit and this generates Bla Bloe and then it hangs. Nothing else is generated. We are guessing this has to do with optimisations of some kind. And what about runit :: IO () runit = traceShow "Bla" $ do n <- traceShow "Num" print 2 traceShow "Bloe” runit Now we get Bla Num 2 Bloe 2 Bloe 2 … No more Bla and no more Num. Any ideas whether this is a bug or a feature? Jur and Tibor

4 3

Haskell Infrastructure Jobs @ Swift Navigation (SF)
by Bhaskar Mookerji 28 Jun '17

28 Jun '17

Hello! I work at Swift Navigation, a high-accuracy GPS receiver manufacturer in San Francisco. We are currently looking to grow out our platform infrastructure team in the next few months and have **several positions** open for experienced Haskell developers. We definitely are open to remote candidates. You can **apply through the posting** here: https://jobs.lever.co/swift-nav/eb80d943-8157-4c9b-afef-8fe98d287c28. The text of this is reproduced at the end of the email. If you have any questions, feel free to comment or email me directly at mookerji(a)swiftnav.com. * About the Work We are a very small team supporting a zoo of libraries and internal services at Swift, including a mature ETL pipeline for evaluating our hardware product, as well an upcoming real-time GPS data product. We open source our work where possible and appropriate (https://github.com/swift-nav?utf8=%E2%9C%93&q=&type=&language=haskell) The ideal candidate will have experience designing and implementing Haskell backends in a commercial web services environment. A Haskell open-source background or production experience using another functional programming language (Scala, Clojure, Erlang, etc.). would also work. We value good communication skills, technical fluency, and responsibility in our day-to-day, and would enjoy seeing these personal qualities in future contributors. * About the Company We design, test, manufacture, and sell a Linux-based GPS/GLONASS receiver for machine automation applications, specifically targeting the autonomous vehicle market. Our work environment is very multidisciplinary, with ~30 engineers covering many areas of engineering (electrical/firmware, embedded Linux, statistics, satellite navigation, web, etc.). You definitely don't need knowledge of these subjects to work here, but you may find it a refreshing change of pace from typical web app development. You can read more about the product here: https://www.swiftnav.com/sites/default/files/piksi_multi_product_summary.pdf Thanks, Buro ================= Swift Navigation is looking for an outstanding software engineer in web infrastructure. You will work with a small team to push the state of the art in satellite navigation technology, making high accuracy positioning ubiquitous and easy to use across a wide variety of industries and applications in unmanned aerial vehicles, robotics, and autonomous transportation. As a member of the infrastructure team, you'll contribute to software services for upcoming networking services as well as internal infrastructure shared broadly by Swift's R&D teams: * Services for Networked Sensors. Service-oriented architecture, scientific modeling services, and distributed message brokers for connected satellite navigation receivers. * Integration Infrastructure. Much of our current work involves rigorously demonstrating that our hardware product works as intended when leaving the lab for a variety of customer scenarios in real-world environments. You'll build and use automation frameworks and assemble instrumentation for hardware-in-the-loop testing, fault injection, and regression testing across these scenarios, as well as data analysis pipelines and metrics monitoring for test units under firmware and manufacturing validation. Candidates should have skills and experience with: - Software engineering in a production environment (commercial web services) - Algorithms, data structures, distributed systems, and relational databases - Haskell, or other typed functional programming languages - Amazon Web Services (AWS) and its related ecosystem (Docker, etc.) for application automation and monitoring - Exceptional problem solving, communication, and collaboration skills. Flexibility and enthusiasm for learning new programming languages, tools, and ideas. Strong candidates will also possess skills in one or more of - Infrastructure engineering for "real-time" sensor networks - Python / NumPy / SciPy / Pandas, R and other analysis/scripting platforms - Version control tools (git), software automation - Open-source development

1 0

I think we should stop "fixing" other people's packages on Hackage silently
by Niklas Hambüchen 28 Jun '17

28 Jun '17

I'm writing this after having spent many hours of debugging why the packages "entropy" and "happy" have different compile-time behaviour when taken from Hackage vs when taken from their upstream git repositories. After lots of frustration, I finally figured out the problem: These packages have "Hackage revisions", where their cabal files were modified from its upstream equivalent: * https://hackage.haskell.org/package/happy-1.19.5/revisions/ * https://hackage.haskell.org/package/entropy-0.3.7/revisions/ Here, "setup-depends" were added to the cabal file, in order to make the package build correctly. These changes are on Hackage only, they cannot be found in the equivalent release of the upstream git repository. This is *extremely* confusing. The practice of "fixing up" other people's cabal files behind their back instead of making those fixes upstream prevents me to use their upstream git packages to reproduce my issues. It hides the fact that there is a problem with the upstream package. In the case of happy, there is not even any upstream fix in the master branch for the missing setup-depends: * https://hackage.haskell.org/package/happy-1.19.5/revision/2.cabal vs * https://github.com/simonmar/happy/blob/2d84ca0/happy.cabal So in this case, I suspect that the silent "fix" has prevented the upstream maintainer to even notice that there is a problem with his package, and so it will be broken again when he uploads the next version. I understand that the concept of Hackage revisions was introduced in a commendable effort to improve the package users' experience: Nobody likes when packages they want to use are broken. And it is clear that something has to be done when upstream authors don't reply to pull requests. But I believe this is not the case here. The upstream maintainers of "happy" and "entropy" are Simon Marlow and Thomas M. DuBuisson, who are very reliable getting stuff fixed (if they know about it!) and they apply PRs quickly and their repositories are well and alive. I believe we're shooting ourselves in the foot with these fixes: The time it takes to make an upstream fix and release it as a proper revision is dwarfed by the time it takes you to figure out bugs when the code you download from Hackage is not the code that the maintainer put there, and when package maintainers don't even notice that their packages are broken because of "silent fixing". So I would welcome if we could make use of the "Hackage revisions" feature only in the utmost necessary cases, or even better, never, and always make properly versioned releases, where a change to any file implies a bump of the version, so that one can clearly see if one is dealing with the unmodified upstream code or not. Greetings, Niklas

9 16

+RTS -n option
by Ruben Astudillo 28 Jun '17

28 Jun '17

Dear list I got a doubt on how the -n flag works. With 2 capabilities, is -A2m -n512k equivalent to just -A512k and 4 less times the garbage collection? It seems a win on every situation. I know I am being naive, but I don't know what could be missing. Care to share your thought on this? -- -- Ruben Astudillo -- PGP: 4EE9 28F7 932E F4AD

1 0

Domain specific error diagnosis in Haskell, a video recorded at Curry On!
by Jurriaan Hage 27 Jun '17

27 Jun '17

Dear Haskellers, The work of Alejandro Serrano Mena and myself on domain-specific error diagnosis in GHC was accepted for a presentation at Curry On! in Barcelona, co-located with ECOOP. I gave the talk last week on what you can do with our branch of GHC. A few well-known Haskellers were around. They had some good questions, and were generally quite positive. In case you missed it, there is a video of the whole affair on YouTube. Here is the link: https://www.youtube.com/watch?v=LuqSkWOcnSA&feature=youtu.be best regards, Jur

1 0