Here is the latest Caml Weekly News, for the week of 16 to 23 August, 2005.
Archive: http://thread.gmane.org/gmane.comp.lang.caml.general/30141Damien Doligez announced:
We have the pleasure of announcing the release of Objective Caml version 3.08.4 This is mostly a bug-fix release; see below for the list of changes. Upgrading is not urgent unless you have problems with one of the bugs listed below. Please note that we do not guarantee binary compatibility with previous versions (including 3.08.3). Only the source is available for the moment. We will provide some binaries in the near future; in the meantime, the binaries on the Web page are still for OCaml 3.08.3. OCaml 3.08.4 is available at http://caml.inria.fr/ocaml/release.en.html -- Damien Doligez for the Caml Team Objective Caml 3.08.4: ---------------------- New features: - configure: find X11 config in some 64-bit Linux distribs - ocamldoc: (**/**) can be canceled with another (**/**) PR#3665 - graphics: added resize_window - graphics: check for invalid arguments to drawing primitives PR#3595 - lablbrowser: use windows subsystem on mingw Bug fixes: - ocamlopt: code generation problem on AMD64 PR#3640 - wrong code generated for some classes PR#3576 - fatal error when compiling some OO code PR#3745 - problem with comparison on constant constructors PR#3608 - better detection of cycles when using -rectypes - missing case of module equality PR#3738 - better error messages for unbound type variables - stack overflow while printing type error message PR#3705 - assert failure when typing some classes PR#3638 - bug in type_approx - better error messages related to type variance checking - camlp4: cryptic error message PR#3592 - camlp4: line numbers in multi-line antiquotations PR#3549 - camlp4: problem with make depend - camlp4: parse error with :> PR#3561 - camlp4: ident conversion problem with val/contents/contents__ - camlp4: several small parsing problems PR#3688 - ocamldebug: handling of spaces in executable file name PR#3736 - emacs-mode: problem when caml-types-buffer is deleted by user PR#3704 - ocamldoc: extra backslash in ocamldoc man page PR#3687 - ocamldoc: improvements to HTML display PR#3698 - ocamldoc: escaping of @ in info files - ocamldoc: escaping of . and \ in man pages PR#3686 - ocamldoc: better error reporting of misplaced comments - graphics: fixed .depend file PR#3558 - graphics: segfault with threads and graphics PR#3651 - nums: several bugs: PR#3718, PR#3719, others - nums: inline asm problems with gcc 4.0 PR#3604, PR#3637 - threads: problem with backtrace - unix: problem with getaddrinfo PR#3565 - stdlib: documentation of Int32.rem and Int64.rem PR#3573 - stdlib: documentation of List.rev_map2 PR#3685 - stdlib: wrong order in Map.fold PR#3607 - stdlib: documentation of maximum float array length PR#3714 - yacc: avoid name capture for idents of the Parsing moduleAleksey Nogin then said:
I've compiled a number of binary packages of OCaml 3.08.4 (Fedora Core 2, 3, and 4, Mandrake 10.1, Red Hat Enterprise Linux 4). They are available at http://rpm.nogin.org/ocaml.html
Archive: http://thread.gmane.org/gmane.comp.lang.caml.general/30159Matt Wade announced:
Wanted: One OCaml Expert to do a technical review on an upcoming book. Pay is decent for the amount of time involved. You'll also receive a full page acknowledgement in the book (with picture if you like) and a copy of the book to use as a business card. If you are interested, please contact me off list with your qualifications and resume. --Matt Wade Open Source Editor http://www.apress.com/
Archive: http://thread.gmane.org/gmane.comp.lang.caml.general/30166Michael Wohlwend asked and Eric Cooper answered:
> I know this question was asked some time ago, but those little > devices are getting better every day :) The tungsten T3/T5 have > 400MHz and 64MB, T5 has 256 MB ram; my sony has 200MHz and 32MB Ram > (o.k., not much). Could ocaml (not caml-light) run on such a T3 or > T5? Or just the runtime and cross-compiling? OCaml runs fine on a Sharp Zaurus (400 MHz ARM with 64 MB of memory); the ocaml interpreter runs on the device as well as cross-compiled native code. But of course it's running Linux, not PalmOS. Porting the runtime would be the main obstacle, I guess, depending on how much POSIX-compatibility you can get from PalmOS. I've also heard rumors that Palm will move to an embedded Linux platform, so maybe the answer is just to be patient.
Archive: http://thread.gmane.org/gmane.comp.lang.caml.general/30177Nicolas Cannasse announced:
I have released Neko 1.0 which is an intermediate programming language with its virtual machine. The compiler is written in OCaml and the VM in C. It might be interesting for people involved in language design, since Neko is providing a common reusable runtime for language designers. You can have a look at the documentation and download the release at http://nekovm.org Comments and Questions are welcome, Nicolas
Archive: http://thread.gmane.org/gmane.comp.lang.caml.general/30190Norman Ramsey asked and Jacques Garrigue answered:
> I wish to define an abbreviation for the following type: > > [> `Nil > | `Number of float > | `String of string > | `F