Here is the latest Caml Weekly News, for the week of March 27 to April 03, 2012.
Archive: https://sympa-roc.inria.fr/wws/arc/caml-list/2012-03/msg00330.htmlAlain Frisch said:
I'd like to ask for the community help in evaluating the benefits of a new strategy to control boxing/unboxing of floats in ocamlopt. I've implemented this new strategy in the more_unboxing branch of the SVN. You can find some description (and micro-benchmarks) there: http://caml.inria.fr/mantis/view.php?id=5204 If numerical code is the bottleneck of one of your applications, it would be helpful if you could you try out the more_unboxing branch (forked from the trunk in january 2011, IIRC) and report the results on Mantis.Dario Teixeira then replied:
Thanks for this! I've given the 'more_unboxing' branch a brief try, using the N-body problem from the infamous shootout as benchmark. Below are the results for executables compiled with 3.12.1, today's SVN trunk, and today's 'more_unboxing' branch. The compiler options are the same as used in the shootout; the tests consisted of 10^9 iterations on an x86_64 machine: Version 3.12.1: time ./nbody.native 1_000_000_000 1111.23user 0.13system 18:32.61elapsed 99%CPU (0avgtext+0avgdata 3472maxresident)k 0inputs+0outputs (0major+273minor)pagefaults 0swaps SVN trunk (2012/03/29): time ./nbody.native 1_000_000_000 969.29user 0.62system 16:14.66elapsed 99%CPU (0avgtext+0avgdata 3488maxresident)k 0inputs+0outputs (0major+273minor)pagefaults 0swaps SVN branches/more_unboxing (2012/03/29): time ./nbody.native 1_000_000_000 960.72user 0.30system 16:03.51elapsed 99%CPU (0avgtext+0avgdata 3472maxresident)k 0inputs+0outputs (0major+272minor)pagefaults 0swaps Though the improvement from trunk to more_unboxing is small (about 1%) and may in fact disappear altogether if one were to run a serious statistical analysis, the improvement from 3.12.1 to 4.00 is quite striking: about 13%! What has changed in the compiler that would explain this? Also, a note to those using GODI: it's fairly straightforward to have multiple co-existing OCaml installations, making the test of trunk or experimental branches less disrupting. You'll find at the end the steps I used to get the 'more_unboxing' branch running. (Note that I had to run each godi_make step individually because of step 6, though I'm sure the GODI ninjas will find ways to streamline this process even further.) Best regards, Dario Teixeira ============================================================================ 1) Tell GODI the SVN location to checkout. Edit $GODI/etc/godi.conf: OCAML_SVN_CHECKOUT = yes OCAML_SVN_PATH = /ocaml/branches/more_unboxing 2) cd $GODI/build/godi/godi-ocaml-src 3) godi_make clean 4) godi_make fetch 5) godi_make extract 6) GODI expects the source code directory to be called 'work/ocaml'. So we must do some renaming before proceeding: mv -T work/more_unboxing work/ocaml 7) godi_make patch 8) godi_make configure 9) godi_make build 10) godi_delete godi-ocaml-src-3.12.1godi1 11) godi_make install 12) godi_make package 13) Start godi_console and mark package godi-ocaml for rebuilding. This will automatically force an overall rebuild. 14) Enjoy!
Archive: https://sympa-roc.inria.fr/wws/arc/caml-list/2012-03/msg00340.htmlVincent Balat announced:
Archive: https://sympa-roc.inria.fr/wws/arc/caml-list/2012-03/msg00322.htmlJonathan Protzenko announced:
I've spent the past few days improving the OCaml installer for windows. This should solve all issues that have been mentioned previously. This in preparation for the upcoming 4.0 release ; therefore, the "beta installer" below will install a fairly recent trunk version of OCaml. http://yquem.inria.fr/~protzenk/caml-installer/ocaml-4.01.0+dev0-i686-mingw64.exe Changelog: - The installer packages a working and well-configured findlib, out of the box. This means you can open up cmd.exe, type ocaml, then type #use "topfind";; in the top-level. - The installer no longer blasts the PATH variable if it is too long. Moreover, it is now able to deal with PATHs that are up to 8192 bytes long. - The installer is now able to deal with multi-user privileges: in particular, you can use this installer on a limited account: OCaml will properly install into your local "Application Data" folder if you don't have administrator rights on your machine. This should be particularly useful for school computers that run Windows. - Various fixes: better cleanups on exit, minor fixes for OCamlWinPlus, properly warn the user if it was unable to download ActiveTCL from the internet. - The installer packages a newer version of flexlink, so you shouldn't have any more issues with gcc not accepting the -mno-cygwin option. However, the new, official toolchain uses the mingw64 32-bit compilers. This means flexlink will look for a i686-w64-mingw32-gcc in your path whenever you try to do native compilation. The official, recommended way to work is to install cygwin and its mingw64 packages. While in theory it should be possible to work within a MSys environment, some issues  currently prevent you from doing so.  http://caml.inria.fr/mantis/view.php?id=5465 Please let me know if anything doesn't work as expected. This will be the official installer for the 4.0 release.Jonathan Protzenko later added:
Following all the good suggestions in this thread, I've updated the installer. It now downloads and runs cygwin's setup.exe so as to provide a fully working environment for OCaml on windows after the installer completes. The cygwin installer runs in silent mode, that is, the progress window shows up, but there is no user prompt. I've also looked into odb compatibility, and it looks like after writing a few patches, odb now runs fine on Windows. However, Oasis-generated setup.ml files do not work at all on windows. Any help in that area would be highly appreciated.
Thanks to Alp Mestan, we now include in the Caml Weekly News the links to the recent posts from the ocamlcore planet blog at http://planet.ocamlcore.org/. Tracking Asynchronous IO Using Type Systems: http://blog.incubaid.com/2012/04/02/tracking-asynchronous-io-using-type-systems/ The topology of the set of all types: http://math.andrej.com/2012/03/30/the-topology-of-the-set-of-all-types-2/ Getting started with Snap-Auth: http://blog.dbpatterson.com/post/20132704797 Ocsigen Eliom 2.1 released: http://ocsigen.org/ Programming tools UX: How we made MongoDB even easier to use with Opa: http://blog.opalang.org/2012/03/programming-tools-ux-experience-how-we.html
If you happen to miss a CWN, you can send me a message and I'll mail it to you, or go take a look at the archive or the RSS feed of the archives.
If you also wish to receive it every week by mail, you may subscribe online.