Poly/ML

polyml@lists.polyml.org

695 discussions

Poly/ML 5.5.1
by David.Matthews＠prolingua.co.uk 16 Oct '13

16 Oct '13

I have uploaded the source for Poly/ML 5.5.1 to the SourceForge site so it is now officially released. I still need to create installers for Windows. The release notes at http://www.polyml.org/docs/ReleaseNotes.html give a general overview of the changes over the last year. There have been no big changes which is the reason it's 5.5.1 rather than 5.6. David

7 20

response to PolyML on Android and iOS thread
by monkeyrater＠hotmail.com 16 Oct '13

16 Oct '13

Sorry for making this response in a new thread, but as I am not a member of the mailing list this is the only way as this being a mailing list, everything is automatically sent instead of being able to edit material in a central location. Id also like to say its an honor to have David Mathews respond to my thread. >I'm not exactly clear about the motivation behind the question. I mentioned in my original post what my motivation is but I will state it again in more detail. Windows is no longer a viable scientific computing platform as Microsoft have gone out of their way to take away user control of how software can be installed and accessed and trying to move toward 'cloud based services' type of user experience. Linux definately allows control but is so fragmented by many different distros that its impossible to port a unified system. And OSX requires the purchase of its own expensive hardware to run on. Furthermore mobile computing is no longer a weak subset of the capabilities of PC computing but equal processor power and memory capacity exists to run equally complex software. However one could argue that mobile platforms are not ready yet due to the fact that they only run on phone or tablet from factor. Although it is possible to use a wireless keyboard with a tablet, what is really required is a netbook form factor such as the Chromebooks. Unfortunately Google has dropped the ball by creating the ChromeOS operating system that only runs 'web apps', trying to force cloud computing on its users like Microsoft is with win 8. >OCAML implementations seem to provide a small console window that allows >someone to enter and run OCAML programs. I don't know how useful that >would be in general. It might be more useful to be able to create >little ML programs. Without a linker it wouldn't be possible to create >stand-alone binaries but it might be possible to create scripts. It might be possible to include the Tiny C compiler http://bellard.org/tcc/ to create binaries. I know this sounds frivolous to go to the trouble of doing this, but it shouldnt be too difficult for any programmer who is experienced with Android programming to do this kind of port. Im sorry to take up too much of your time with this idea, Im sure you probably have more important things to concern yourselves with. But I believe that the ML family of languages are going to become much more popular in place of other functional languages like Haskell or Scala which both bring limitations that prevent adoption for general use. And I would hate to see SML lose out simply because it is not as accessible as Ocaml. Mobile computing is the future and so far Android is the most viable platform for it. Id also like to mention this class: https://www.coursera.org/course/proglang that has attracted tens of thousands of online students and will do a lot to raise awareness of ML's advantages over other functional languages. -Brad Cantrell

2 1

Compiler in SVN
by David.Matthews＠prolingua.co.uk 03 Sep '13

03 Sep '13

There's been quite a lot of work on the intermediate code optimiser since the last release and it's largely been rewritten. The main aim has been to reduce the number of cases where heap storage was allocated for argument tuples or closures. There were a number of bugs that appeared during testing but I'm not aware of anything outstanding. Isabelle2013 and HOL4 both seem to build successfully and there does seem to be some speed-up. I'd like to release this version in the next month or so but I'd like to fix any other bugs that have been introduced. Please try out the SVN version and let me know how it goes. Don't forget that after the configure and make steps you need to run "make compiler" at least once and preferably twice. Otherwise you will still be using the 5.5 compiler. David

6 16

Warning building Poly/ML on Mac OS X
by rda＠lemma-one.com 01 Sep '13

01 Sep '13

I just built Poly/ML from the latest sources in SVN on Mac OS X 10.8.4 and got the following warning: /bin/sh ./libtool --tag=CC --mode=link gcc -DMACOSX -Wall -fno-strict-aliasing -O3 -I../libffi/include -o poly polyexport.o libpolymain/libpolymain.la libpolyml/libpolyml.la -lpthread -lm -ldl -lstdc++ -lgcc libtool: link: gcc -DMACOSX -Wall -fno-strict-aliasing -O3 -I../libffi/include -o poly polyexport.o libpolymain/.libs/libpolymain.a libpolyml/.libs/libpolyml.a -lpthread -lm -ldl -lstdc++ -lgcc ld: warning: could not create compact unwind for _ffi_call_unix64: does not use RBP or RSP based frame ld: warning: PIE disabled. Absolute addressing (perhaps -mdynamic-no-pic) not allowed in code signed PIE, but used in area1 from polyexport.o. To fix this warning, don't compile with -mdynamic-no-pic or link with -Wl,-no_pie I don't have any problems running Poly/ML after this, but I wondered if it was something that somebody somewhere might care about. Regards, Rob.

1 0

pretty printing infrastructure
by Michael.Norrish＠nicta.com.au 28 Aug '13

28 Aug '13

First a feature request: could we please have a version of the basic function for adding a string that takes an optional argument indicating how wide the Oppen algorithm should treat the string? At the moment, the infrastructure just seems to count the bytes emitted, and this causes premature line breaks if I'm emitting UTF-8 characters. Of course, if the infrastructure could magically become aware of UTF-8 or wide characters that would solve this particular use-case, but I'd still ask for this feature because I also want to emit byte-sequences that are terminal control codes. (On appropriate terminals, I emit vt100 codes to colour certain substrings, for example.) Secondly, I'm a bit concerned that the new, recommended approach to pretty-printing requires me to translate into the Pretty structure before anything happens. I'd feel more comfortable if the structure could be produced and consumed lazily so that I don't end up with a large (strictly unnecessary) Pretty structure in memory. Michael ________________________________ The information in this e-mail may be confidential and subject to legal professional privilege and/or copyright. National ICT Australia Limited accepts no liability for any damage caused by this email or its attachments.

3 10

call Poly/ML from not native Poly/ML threads
by pob_001＠mail.ru 15 Aug '13

15 Aug '13

Hi, Please forgive me for troubling you but I have a small question. Whether it is possible to make callback from C to PolyML from threads that not created by Poly/ML ? Thanks, Dmitry

4 11

Printing of function types
by rda＠lemma-one.com 12 Aug '13

12 Aug '13

Poly/ML seems to be a little inconsistent about when it uses type names supplied by the user to abbreviate types as shown in the following transcript: > type COMPLEX = {re : real, im : real}; type COMPLEX = {im: real, re: real} > val unity : COMPLEX = {re = 1.0, im = 0.0}; val unity = {im = 0.0, re = 1.0}: COMPLEX > fun negate ({re, im} : COMPLEX) : COMPLEX = {re = ~re, im = ~im}; val negate = fn: COMPLEX -> {im: real, re: real} Regards, Rob.

3 2

Linking exported object files
by David.Matthews＠prolingua.co.uk 02 Aug '13

02 Aug '13

I've updated the "polyc" script so that as well as accepting an ML source file to compile and link it can also be used to link object files created with PolyML.export. Switching to static linking by default in SVN has turned up a few problems. In particular it seems that it's important to include the libraries required by libpolyml. The problem is that this can depend on how it was built. In particular, the configure script will use GMP if it is present but falls back to the internal arbitrary precision package if it is not. That has the effect that if GMP was detected during the build process then the linking line must include -lgmp. However, if GMP is not present then including -lgmp will result in an error message. This makes it rather difficult to write a general purpose linking script. The linking line found at http://www.polyml.org/FAQ.html#standalone has gone through several versions and still has to be tweaked. It should now just require polyc -o hello hello.o Although "polyc" is a shell script it is produced during the build process and includes the appropriate options for linking depending on what the configure script found. The name "polyc" was meant to be analogous to "cc" or "gcc". I don't know whether it might be better to use the name "poly" for both the "batch" and interactive versions depending on the options. Any ideas? David

2 2

Linking exported object files
by David.Matthews＠prolingua.co.uk 01 Aug '13

01 Aug '13

On 01/08/2013 16:23, Rob Arthan wrote: > > On 1 Aug 2013, at 16:07, David Matthews wrote: >> ... In particular, the configure script will use GMP if it is >> present but falls back to the internal arbitrary precision package >> if it is not. That has the effect that if GMP was detected during >> the build process then the linking line must include -lgmp. >> However, if GMP is not present then including -lgmp will result in >> an error message. > > What are the advantages in building with GMP? If you are working with large arbitrary precision values particularly multiplying and dividing GMP is likely to be significantly faster. The default code is fairly naive. It's unlikely to make any difference for values that are only slightly longer than short precision (31-bits on 32-bit; 63-bits on 64-bit) and doesn't affect arithmetic on short precision values. Regards, David

1 0

Fwd: Performance experiment with Poly/ML
by jesus-maria.aransay＠unirioja.es 30 Jul '13

30 Jul '13

Hi David and Matthew, thanks for your encouraging answers. The SML code directly generated from Isabelle/HOL is available from the input [ADb] in the paper bibliography, in http://www.unirioja.es/cu/jodivaso/Isabelle/Gauss-Jordan/ under the link "View the generated code". The matrices used for the benchmarks are in a zip file (9MB compressed, 160MB uncompressed) in the link "Download the benchmarks" in the previous website. Uncompressing the folder three different folders appear (RAT, REAL, Z2). In each of these folders there is a "Readme" file where we describe how we carried out the computations (we would be very glad to hear if there is a more proper way, both in MLton and Poly/ML). The segmentation fault was raised with Poly/ML 5.5 trying to compute the matrix "Benchmarks/REAL/800_real.sml". There is also the issue of lack of performance from Poly/ML 5.5 to Poly/ML 5.2. As an example, the matrix "Benchmarks/REAL/200_real.sml" sums up 3.7 (processing) + 0.7 (executing) seconds with Poly/ML 5.5 and 31.0 (processing) + 0.4 (executing) seconds with Poly/ML 5.2 (the version shipped as an Ubuntu package); the same matrix in MLton took 9145.0 (processing) + 0.5 (executing) seconds. Thanks in advance for your advice, do not hesitate to contact us for further questions, Jesus On Thu, Jul 25, 2013 at 4:59 PM, Matthew Fluet <matthew.fluet at gmail.com> wrote: > On Thu, Jul 25, 2013 at 8:03 AM, David Matthews > <David.Matthews at prolingua.co.uk> wrote: >> On 25/07/2013 11:29, Jesus Aransay wrote: >>> >>> Jose Divas?n and I wrote a short paper for a Spanish workshop on >>> functional programming ("Taller de Programaci?n Funcional 2013", >>> http://babel.ls.fi.upm.es/tpf2013/) in which we test a SML program >>> obtained with the Isabelle code generator tool to perform the Gauss >>> Jordan elimination algorithm over matrices. >> >> That is interesting work. Is it possible to get a copy of the code you ran >> for the various tests so I can check this out? > > Seconded! I would love to investigate why MLton is performing so > poorly on the input processing. > > -Matthew > > -- Jes?s Mar?a Aransay Azofra Universidad de La Rioja Dpto. de Matem?ticas y Computaci?n tlf.: (+34) 941299438 fax: (+34) 941299460 mail: jesus-maria.aransay at unirioja.es ; web: http://www.unirioja.es/cu/jearansa Edificio Luis Vives, c/ Luis de Ulloa s/n, 26004 Logro?o, Espa?a -- Jes?s Mar?a Aransay Azofra Universidad de La Rioja Dpto. de Matem?ticas y Computaci?n tlf.: (+34) 941299438 fax: (+34) 941299460 mail: jesus-maria.aransay at unirioja.es ; web: http://www.unirioja.es/cu/jearansa Edificio Luis Vives, c/ Luis de Ulloa s/n, 26004 Logro?o, Espa?a

4 3

← Newer
1
...
27
28
29
30
31
32
33
...
70
Older →

Jump to page:

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