Poly/ML

polyml@lists.polyml.org

2 participants
711 discussions

Fwd: Re: [polyml] Documentation of Poly/ML opcode
by xyly781＠gmail.com 01 May '10

01 May '10

forward to the list. -------- Original Message -------- Subject: Re: [polyml] Documentation of Poly/ML opcode Date: Sat, 01 May 2010 13:58:45 -0500 From: Yue Li <xyly781(a)gmail.com> To: David Matthews <David.Matthews(a)prolingua.co.uk> Hi, David, On 05/01/2010 06:38 AM, David Matthews wrote: > Hi, > The opcodes in int_opcodes.h apply only to the interpreted version of > Poly/ML and the interpreter is only used on platforms where there is > no native code-generator. > I see. Ok! > There's not much, if any, documentation for the internals of the > compiler or run-time system. Probably the most useful data structures > to look at are in mlsource/MLCompiler/CODETREESIG.ML . I did not discover that before. The IR is really helpful! Thanks! > The top-level of the compiler builds this intermediate level data > structure and the optimiser transforms it. It is then code-generated > by the bottom level using either a native code-generator; for the X86 > (32 or 64-bit), PPC or Sparc or byte code for the interpreter on any > other platform. Many operations, though, are performed by RTS > (run-time system) functions and these include such basic operations as > integer addition. These are listed in libpoly/sys.h and > basis/RunTimeCalls.ML. In the general case these actually require a > call into the C++ run-time system but simple operations are > implemented in the assembly code section or are recognised by the > code-generator and turned into the appropriate machine instructions. > No problem! Here could you point me to the code where the native and byte code generator lives? Thank you very much! Yue > Regards, > David > > Yue Li wrote: >> Hi, >> >> I'm currently studying the run time system of Poly/ML, looking at its >> source code. I found there are 130 instructions listed in >> int_opcodes.h, and the semantics is defined in interpreter function. >> I'm wondering whether there's any documentation available for >> describing the semantics of each instruction, this will definitely >> help me understand each instruction better. >> >> Thanks for your help, >> >> Yue >> _______________________________________________ >> polyml mailing list >> polyml(a)inf.ed.ac.uk >> http://lists.inf.ed.ac.uk/mailman/listinfo/polyml

1 0

Documentation of Poly/ML opcode
by xyly781＠gmail.com 01 May '10

01 May '10

Hi, I'm currently studying the run time system of Poly/ML, looking at its source code. I found there are 130 instructions listed in int_opcodes.h, and the semantics is defined in interpreter function. I'm wondering whether there's any documentation available for describing the semantics of each instruction, this will definitely help me understand each instruction better. Thanks for your help, Yue

2 1

CUFP 2010 in Baltimore in September -- Call for Participation
by sweeks＠sweeks.com 24 Apr '10

24 Apr '10

Commercial Users of Functional Programming Workshop (CUFP) 2010 Call for Participation Sponsored by SIGPLAN Co-located with ICFP 2010 Baltimore, Maryland Sep 27-29, 2010 Submission Deadline: 15 June 2010 Functional programming languages have been a hot topic of academic research for over 35 years, and have seen an ever larger practical impact in settings ranging from tech startups to financial firms to biomedical research labs. At the same time, a vigorous community of practically-minding functional programmers has come into existence. CUFP is designed to serve this community. The annual CUFP workshop is a place where people can see how others are using functional programming to solve real world problems; where practitioners meet and collaborate; where language designers and users can share ideas about the future of their favorite language; and where one can learn practical techniques and approaches for putting functional programming to work. # Giving a CUFP Talk # If you have experience using functional languages in a practical setting, we invite you to submit a proposal to give a talk at the workshop. We're looking for two kinds of talks: *Experience reports* are typically 25 minutes long, and aim to inform participants about how functional programming plays out in real-world applications, focusing especially on lessons learned and insights gained. Experience reports don't need to be highly technical; reflections on the commercial, management, or software engineering aspects are, if anything, more important. You do not need to submit a paper! *Technical talks* are expected to be 30-45 minutes long, and should focus on teaching the audience something about a technical technique or methodology, from the point of view of someone who has seen it play out in practice. These talks could cover anything from techniques for building functional concurrent applications, to managing dynamic reconfigurations, to design recipes for using types effectively in large-scale applications. While these talks will often be based on a particular language, they should be accessible to a broad range of functional programmers. If you are interested in offering a talk, or nominating someone to do so, send an e-mail to francesco(at)erlang- consulting(dot)com or yminsky(at)janestreet(dot)com by **15 June 2010** with a short description of what you'd like to talk about or what you think your nominee should give a talk about. Such descriptions should be about one page long. There will be no published proceedings, as the meeting is intended to be more a discussion forum than a technical interchange. # Program Committee # * Francesco Cesarini, Erlang Training and Consulting (Co-Chair) * Tim Dysinger, Sonian Networks * Alain Frisch, LexiFi * Nick Gerakines, Chegg * Adam Granicz, IntelliFactory * Amanda Laucher * Romain Lenglet, Google Japan * Yaron Misky, Jane Street (Co-Chair) * Mary Sheeran, Chalmers * Don Stewart, Galois * Dean Wampler, DRW Trading # More information # For more information on CUFP, including videos of presentations from previous years, take a look at the CUFP website at <http://cufp.org>. Note that presenters, like other attendees, will need to register for the event. Presentations will be video taped and presenters will be expected to sign an ACM copyright release form. Acceptance and rejection letters will be sent out by July 15th.

1 0

Code examples in Parallel Poly/ML and Isabelle/ML paper
by yli＠cse.tamu.edu 24 Apr '10

24 Apr '10

Hi, Makarius, Thank you very much for these information, and I apologize that I reply this so late. I will start working on topics related to parallel computer algebra during this summer, and is very interested in knowing how these examples work so well from your Poly/ML paper so that I can take advantage of the mechanism in the paper to my work, therefore I was asking for these useful examples presented in the paper. I will ask more if I have further questions. Thank you very much, Yue On Fri, Feb 12, 2010 at 6:11 PM, Yue K. Li <yue.kevin.li(a)gmail.com> wrote: > ---------- Forwarded message ---------- > From: Makarius <makarius(a)sketis.net> > Date: Fri, Feb 12, 2010 at 2:21 PM > Subject: Re: [polyml] Code examples in Parallel Poly/ML and Isabelle/ML paper > To: Yue Li <yue.kevin.li(a)gmail.com> > Cc: polyml(a)inf.ed.ac.uk > > > On Thu, 11 Feb 2010, Yue Li wrote: > >> Recently I read the paper "Efficient Parallel Programming in Poly/ML and Isabelle/ML" published in DAMP'10. It's really a nice paper for me to read. Especially, I'm deeply impressed by the good scalability and cpu utilization of multithreading implementation as shown in the data section. I'm very interested in the parallel implementation of Poly/ML, and would like to try to use it. Here may I ask is there any way for me to obtain the testing examples used in the paper's experiment so that I could study and play with the parallel features? > > In the paper we point to a particular repository version of Isabelle > (with Mercurial changeset id), but it is easier to use the official > Isabelle2009-1 release that came out a few weeks later -- the parallel > infrastructure did not change much in between. > > See http://isabelle.in.tum.de and > http://isabelle.in.tum.de/installation.html in particular. ?A suitable > Poly/ML compilation for many platforms is also included here -- this > is actually required to use the Isabelle logic images from that > website. > > To work with Isabelle/ML, you can mostly forget about everything > involving Proof General and Emacs, which causes the main problems for > first-time installation. ?If you have the GNU "rlwrap" ("readline > wrapper") tool on your system, you already get a convenient tty > interface. ?E.g. run it like this: > > ?$ Isabelle2009-1/bin/isabelle tty -l HOL > ?Welcome to Isabelle/HOL (Isabelle2009-1: December 2009) > ?> exit > ?ML> > > Strictly speaking, the "Pure" image is sufficient, instead of the > slightly bulky "HOL" one, but the latter is available as precompiled > on the Isabelle download site. > > If you want to compile only the core Isabelle/Pure system you can do > it like this: > > ?$ Isabelle2009-1/build Pure > ?... > ?Press RETURN to compilation of > > ?Pure > > ?Started at Fri Feb 12 21:01:22 CET 2010 (polyml-5.3.0_x86-linux on ...) > ?Building Pure ... > ?Finished Pure (0:00:11 elapsed time, 0:00:10 cpu time, factor 0.90) > ?Finished at Fri Feb 12 21:01:33 CET 2010 > ?0:00:11 elapsed time, 0:00:10 cpu time, factor 0.90 > > And invoke it as above, replacing "HOL" by "Pure" in the command line. > > > The "applications" from the paper are all relatively big Isabelle > logic sessions, with formal definitions and proofs being processed by > semi-automatic tools. ?The ideas behind parallelizing formal document > processing is further explained in > http://www4.in.tum.de/~wenzelm/papers/parallel-isabelle.pdf > > You need to use "isabelle usedir ..." and/or "isabelle make" with > certain parameters in certain directories to reproduce the examples. > I can give you further details on demand -- even some Scala script to > produce the graphs from the paper. ?Note that certain server-class > hardware sometimes needs extra tweaking to get good results, while > high-end workstations usually work without further worries (notably > Mac Pro). > > > Probably your main interest will be the Isabelle/ML library for > working with synchronized variables, future values etc. ?All of this > is in Isabelle2009-1/src/Pure/Concurrent/. ?The ML names in the paper > have been slightly shortened. > > Here is an example: > > ?val x = Future.fork (fn () => 1 + 1); > > The ML toplevel pretty printer will display the value of x, just type > "x" again at the toplevel later. ?You can also use Future.join > explicitly. > > Here is the example from section 4.2: > > ?fun loop () : int = loop (); > > ?val g = Task_Queue.new_group NONE; > ?val x = Future.fork_group g loop; > ?val y = Future.fork_group g (fn () => 1 div 0); > > > The library implementation is fine-tuned towards the particular needs > of Isabelle proof checking, but it should be straight forward to turn > this into a fully generic ML library. ?(I would be willing to assist > anybody in doing this, but I can myself only maintain the Isabelle/ML > version.) > > > ? ? ? ?Makarius >

2 1

PolyML.print/makestring vs. toplevel pretty printing
by makarius＠sketis.net 12 Apr '10

12 Apr '10

Two questions concerning the builtin pretty printing facilities of Poly/ML 5.3 (or pre-5.4 SVN version 1125). (1) The Poly/ML toplevel pretty prints values nicely, according to the print functions installed by the user. It seems that PolyML.print (and PolyML.makestring) produce an unformatted string only via PolyML.uglyPrint instead of PolyML.prettyPrint. Is this a fundamental limitation, or could one get regular pretty printed output here as well? (2) When mapping Pretty.fbrk (forced break) of Isabelle to the Poly/ML model we simulate it via a very large regular break, greater than the usual line length. This produces funny output with PolyML.print/makestring, because it does not do the formatting as observed above. Is it possible to force breaks more directly, apart from embedding a literal "\n" which is not counted properly? Makarius

2 2

Profiling on Mac OS X
by David.Matthews＠prolingua.co.uk 04 Apr '10

04 Apr '10

There's a problem using PolyML.profiling 1 on Mac OS X and think this has been discussed before. Time profiling works by setting a timer and then examining the program counter when the timer goes off to determine which function is being executed. Linux supports per-thread timers and when a timer goes off the thread that is executing receives the signal. Mac OS X has only per-process timers and all signals are sent to the "main" thread, the thread that handles garbage-collection, which means that the profiler cannot work out what function was actually being executed. I've now added an option to "configure", --without-threads, that builds a version of poly that is single-threaded and so allows profiling on Mac OS X. Note that disabling threads also disables signal handling. David

1 0

PASCO 2010: Extended submission deadline
by Daniel.Cordeiro＠imag.fr 29 Mar '10

29 Mar '10

(Please accept our apologies if you receive this message multiple times) ------------------------------------------------------------------------ Parallel Symbolic Computation 2010 (PASCO 2010) July 21-23, 2010 -- Grenoble, France http://pasco2010.imag.fr/ Third Announcement and C A L L f o r P A P E R S *** Submission date extended to April 18 *** Overview: The International Workshop on Parallel and Symbolic Computation (PASCO) is a series of workshops dedicated to the promotion and advancement of parallel algorithms and software in all areas of symbolic mathematical computation. The pervasive ubiquity of parallel architectures and memory hierarchy has led to the emergence of a new quest for parallel mathematical algorithms and software capable of exploiting the various levels of parallelism: from hardware acceleration technologies (multi- core and multi-processor system on chip, GPGPU, FPGA) to cluster and global computing platforms. To push up the limits of symbolic and algebraic computations, beyond the optimization of the application itself, the effective use of a large number of resources --memory and general or specialized computing units-- is expected to enhance the performance multi-criteria objectives: time, energy consumption, resource usage, reliability. In this context, the design and the implementation of mathematical algorithms with provable and adaptive performances is a major challenge. Earlier meetings in the PASCO series include PASCO'94 (Linz, Austria), PASCO'97 (Maui, U.S.A.), PASCO'07 (London, Canada). PASCO 2010 is affiliated with the 2010 International Symposium on Symbolic and Algebraic Computation (ISSAC) in Munich, Germany. Immediately prior to the ISSAC 2010 meeting, PASCO will be held in Grenoble, France. The workshop PASCO 2010 will be a three-day event including invited presentations and tutorials, contributed research papers and posters, and a programming contest. Specific topics include, but are not limited to: * Design and analysis of parallel algorithms for computer algebra * Practical parallel implementation of symbolic or symbolic-numeric algorithms * High-performance software tools and libraries for computer algebra * Applications of high-performance computer algebra * Distributed data-structures for computer algebra * Hardware acceleration technologies (multi-cores, GPUs, FPGAs) applied to computer algebra * Cache complexity and cache-oblivious algorithms for computer algebra * Compile-time and run-time techniques for automating optimization and platform adaptation of computer algebra algorithms Call for Papers: The conference invites submission of papers presenting original research, either in the form of extended abstracts (2 pages) or full papers (up to 10 pages) in ACM format. As in previous years, PASCO 2010 will publish formal proceedings of the accepted papers. The PASCO 2010 Conference proceedings will be published by the Association for Computing Machinery (ACM) http://www.acm.org/ and will be made available in the ACM Digital Library. The proceedings will also be distributed on CDs to the participants of the workshop. Programming Contest: A programming contest will be organized during the PASCO 2010 meeting. Each participating team will be competing on one or more programming challenges proposed by the organizing committee and by the participants. Details can be found at the PASCO 2010 web site. Important Dates: (extended w.r.t. previous calls for papers) Paper submission deadline: April 18 (Sun), 2010 Programming Challenge proposal April 18 (Sun), 2010 Notification of acceptance: May 31 (Mon), 2010 (for papers and challenge proposals) Camera-ready version of paper due: June 13 (Sun), 2010 Intent of participation to the contest: June 13 (Sun), 2010 Tutorials and Workshop in Grenoble: July 21-23 (Wed-Fri), 2010 Conference Chairs: Marc Moreno Maza (University of Western Ontario, Canada) <moreno(a)csd.uwo.ca> Jean-Louis Roch (Grenoble University, France) <jlroch(a)imag.fr> Local Arrangements Chairs: Jean-Guillaume Dumas (Grenoble University, France) <jean-guillaume.dumas(a)imag.fr> Thierry Gautier (INRIA Grenoble, France) <thierry.gautier(a)imag.fr> Clement Pernet (Grenoble University, France) <clement.pernet(a)imag.fr> Publicity Chair: Daniel Cordeiro (Grenoble University, France) <daniel.cordeiro(a)imag.fr> Administration: Daniele Herzog (INRIA Grenoble, France) <daniele.herzog(a)inrialpes.fr> Christian Seguy (CNRS, France) <christian.seguy(a)imag.fr> Ahlem Zammit-Boubaker (INRIA Grenoble, France) <ahlem.zammit-boubaker(a)inrialpes.fr> Program Committee: Daniel Augot (INRIA Saclay, France) Jean-Claude Bajard (Montpellier II University, France) Olivier Beaumont (INRIA Bordeaux, France) Bruce Char (Drexel University, USA) Gene Cooperman (Northeastern University, USA) Gabriel Dos-Reis (Texas A&M University, USA) Jean-Christophe Dubacq (LIPN, University Paris 13, France) Jean-Guillaume Dumas (Grenoble University, France) Jean-Charles Faugere (INRIA - UPMC, France) Matteo Frigo (Axis Semiconductor, USA) Thierry Gautier (INRIA Grenoble, France) Pascal Giorgi (LIRMM, France) Stef Graillat (University Paris 6, France) Jeremy Johnson (Drexel University, USA) Erich Kaltofen (NCSU, USA) Herbert Kuchen (University of Muenster, Germany) Philippe Langlois (University of Perpignan, France) Anton Leykin (Georgia Tech, USA) Gennadi Malaschonok (Tambov State University, Russia) Michael Monagan (Simon Fraiser University, Canada) Winfried Neun (Zuse Institute Berlin, Germany) Clement Pernet (Grenoble University, France) Nicolas Pinto (MIT, USA) Manuel Prieto-Matias (Complutense university of Madrid, Spain) Markus Pueschel (Carnegie Mellon University, USA) Nathalie Revol (INRIA - LIP, France) David Saunders (University of Delaware, USA) Eric Schost (University of Western Ontario, Canada) Wolfgang Schreiner (RISC, Austria) Arne Storjohann (University of Waterloo, Canada) Sivan Toledo (Tel-Aviv University, Israel) Gilles Villard (CNRS, France) Yuzhen Xie (University of Western Ontario, Canada) Kazuhiro Yokoyama (Rikkyo University, Japan) Links: PASCO 2010 http://pasco2010.imag.fr/ ISSAC 2010 http://www.issac-conference.org/2010/ Submission http://www.easychair.org/conferences/?conf=pasco2010

1 0

segmentation fault during build
by ramana.kumar＠gmail.com 10 Mar '10

10 Mar '10

I decided to update Poly/ML svn today and ran into this problem. Not a big deal since my old version still works, but maybe there's a bug to be found. ... Use: basis/PrettyPrinter.sml Use: basis/FinalPolyML.sml Use: basis/TopLevelPolyML.sml /bin/sh: line 1: 16849 Segmentation fault ./polyimport -H 10 polytemp.txt -I . < ./exportPoly.sml make[2]: *** [polyexport.o] Error 139 make[2]: Leaving directory `/home/ramana/builds/polyml-svn/src/polyml/polyml' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/ramana/builds/polyml-svn/src/polyml/polyml' make: *** [all] Error 2 svn info Path: . URL: https://polyml.svn.sourceforge.net/svnroot/polyml/trunk Repository Root: https://polyml.svn.sourceforge.net/svnroot/polyml Repository UUID: 57040599-48ee-4647-8227-d26cbd406fcb Revision: 1077 these are the commands I was trying to run (not sure where it got up to): make distclean ./configure --prefix=/usr make || return 1 make compiler make || return 1 make DESTDIR=$pkgdir install

3 4

PASCO 2010: News and 2nd Call for Papers
by Daniel.Cordeiro＠imag.fr 08 Mar '10

08 Mar '10

(Please accept our apologies if you receive this message multiple times) ------------------------------------------------------------------------ Parallel Symbolic Computation 2010 (PASCO 2010) July 21-23, 2010 -- Grenoble, France http://pasco2010.imag.fr/ Second Announcement and C A L L f o r P A P E R S News: - PASCO 2010 Programming contest - PASCO 2010 in cooperation with ACM Overview: The International Workshop on Parallel and Symbolic Computation (PASCO) is a series of workshops dedicated to the promotion and advancement of parallel algorithms and software in all areas of symbolic mathematical computation. The pervasive ubiquity of parallel architectures and memory hierarchy has led to the emergence of a new quest for parallel mathematical algorithms and software capable of exploiting the various levels of parallelism: from hardware acceleration technologies (multi- core and multi-processor system on chip, GPGPU, FPGA) to cluster and global computing platforms. To push up the limits of symbolic and algebraic computations, beyond the optimization of the application itself, the effective use of a large number of resources --memory and general or specialized computing units-- is expected to enhance the performance multi-criteria objectives: time, energy consumption, resource usage, reliability. In this context, the design and the implementation of mathematical algorithms with provable and adaptive performances is a major challenge. Earlier meetings in the PASCO series include PASCO'94 (Linz, Austria), PASCO'97 (Maui, U.S.A.), PASCO'07 (London, Canada). PASCO 2010 is affiliated with the 2010 International Symposium on Symbolic and Algebraic Computation (ISSAC) in Munich, Germany. Immediately prior to the ISSAC 2010 meeting, PASCO will be held in Grenoble, France. The workshop PASCO 2010 will be a three-day event including invited presentations and tutorials, contributed research papers and posters, and a programming contest. Specific topics include, but are not limited to: * Design and analysis of parallel algorithms for computer algebra * Practical parallel implementation of symbolic or symbolic-numeric algorithms * High-performance software tools and libraries for computer algebra * Applications of high-performance computer algebra * Distributed data-structures for computer algebra * Hardware acceleration technologies (multi-cores, GPUs, FPGAs) applied to computer algebra * Cache complexity and cache-oblivious algorithms for computer algebra * Compile-time and run-time techniques for automating optimization and platform adaptation of computer algebra algorithms Call for Papers: The conference invites submission of papers presenting original research, either in the form of extended abstracts (2 pages) or full papers (up to 10 pages) in ACM format. As in previous years, PASCO 2010 will publish formal proceedings of the accepted papers. The PASCO 2010 Conference proceedings will be published by the Association for Computing Machinery (ACM) http://www.acm.org/ and will be made available in the ACM Digital Library. Programming Contest: A programming contest will be organized during the PASCO 2010 meeting. Each participating team will be competing on one or more programming challenges proposed by the organizing committee and by the participants. Details can be found at the PASCO 2010 web site. Important Dates Paper submission deadline: April 2 (Fri), 2010 Programming Challenge proposal April 2 (Fri), 2010 Notification of acceptance: May 10 (Mon), 2010 (for papers and challenge proposals) Camera-ready version of paper due: May 28 (Fri), 2010 Intent of participation to the contest: May 28 (Fri), 2010 Tutorials and Workshop in Grenoble: July 21-23 (Wed-Fri), 2010 Conference Chairs: Marc Moreno Maza (University of Western Ontario, Canada) <moreno(a)csd.uwo.ca> Jean-Louis Roch (Grenoble University, France) <jlroch(a)imag.fr Local Arrangements Chairs: Jean-Guillaume Dumas (Grenoble University, France) <jean-guillaume.dumas(a)imag.fr Thierry Gautier (INRIA Grenoble, France) <thierry.gautier(a)imag.fr> Clement Pernet (Grenoble University, France) <clement.pernet(a)imag.fr> Publicity Chair: Daniel Cordeiro (Grenoble University, France) <daniel.cordeiro(a)imag.fr> Administration: Daniele Herzog (INRIA Grenoble, France) <daniele.herzog(a)inrialpes.fr> Christian Seguy (CNRS, France) <christian.seguy(a)imag.fr> Ahlem Zammit-Boubaker (INRIA Grenoble, France) <ahlem.zammit-boubaker(a)inrialpes.fr> Program Committee: Daniel Augot (INRIA Saclay, France) Jean-Claude Bajard (Montpellier II University, France) Olivier Beaumont (INRIA Bordeaux, France) Bruce Char (Drexel University, USA) Gene Cooperman (Northeastern University, USA) Gabriel Dos-Reis (Texas A&M University, USA) Jean-Christophe Dubacq (LIPN, University Paris 13, France) Jean-Guillaume Dumas (Grenoble University, France) Jean-Charles Faugere (INRIA - UPMC, France) Matteo Frigo (Axis Semiconductor, USA) Thierry Gautier (INRIA Grenoble, France) Pascal Giorgi (LIRMM, France) Stef Graillat (University Paris 6, France) Jeremy Johnson (Drexel University, USA) Erich Kaltofen (NCSU, USA) Herbert Kuchen (University of Muenster, Germany) Philippe Langlois (University of Perpignan, France) Anton Leykin (Georgia Tech, USA) Gennadi Malaschonok (Tambov State University, Russia) Michael Monagan (Simon Fraiser University, Canada) Winfried Neun (Zuse Institute Berlin, Germany) Clement Pernet (Grenoble University, France) Nicolas Pinto (MIT, USA) Manuel Prieto-Matias (Complutense university of Madrid, Spain) Markus Pueschel (Carnegie Mellon University, USA) Nathalie Revol (INRIA - LIP, France) David Saunders (University of Delaware, USA) Eric Schost (University of Western Ontario, Canada) Wolfgang Schreiner (RISC, Austria) Arne Storjohann (University of Waterloo, Canada) Sivan Toledo (Tel-Aviv University, Israel) Gilles Villard (CNRS, France) Yuzhen Xie (University of Western Ontario, Canada) Kazuhiro Yokoyama (Rikkyo University, Japan) Links: PASCO 2010 http://pasco2010.imag.fr/ ISSAC 2010 http://www.issac-conference.org/2010/ Submission http://www.easychair.org/conferences/?conf=pasco2010

1 0

run-time cost of single-constructor datatypes
by brianh＠cs.pdx.edu 08 Mar '10

08 Mar '10

Hello all, In Haskell, when I declare the following type: data Foo = MkFoo (Int, Bool) the constructor "MkFoo" has a run-time cost: A pointer must be dereferenced whenever I pattern-match on MkFoo, and memory is allocated when I apply the MkFoo constructor. On the other hand, if I declare this type: newtype Foo = MkFoo (Int, Bool) then the MkFoo constructor has no existence at run-time: Types Foo and (Int, Bool) have the same representation on the heap, and the MkFoo constructor function is a no-op. Here's my question: When I declare a single-constructor datatype in Poly/ML, is it more like Haskell's "data" or "newtype"? Here's the reason I'm asking: I'm thinking of using single-constructor datatypes to implement abstract types, by defining the datatype within the module, but not including the constructor function in the module signature. (This is exactly how people usually do abstract types in Haskell - just don't export the constructors.) I'd like to know if there are any pros or cons of this method compared to using abstype. - Brian

2 3

← Newer
1
...
45
46
47
48
49
50
51
...
72
Older →

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Poly/ML