./lang/ocaml, The latest implementation of the Caml dialect of ML

[ CVSweb ] [ Homepage ] [ RSS ] [ Required by ] [ Add to tracker ]


Branch: pkgsrc-2010Q3, Version: 3.12.0nb1, Package name: ocaml-3.12.0nb1, Maintainer: adam

Objective Caml is the latest implementation of the Caml dialect of ML. The main
novelties compared with its ancestor, Caml Light, are:

* Full support for objects and classes -- here combined for the first time
with ML-style type reconstruction.
* A powerful module calculus in the style of Standard ML (but retaining
separate compilation).
* A high-performance native code compiler (in addition to a Caml
Light-style bytecode compiler).


Required to build:
[devel/gmake]

Master sites:

SHA1: 33ebbfb1115806f117808f37e40d206c8994943d
RMD160: 3d10d5f7431d8200688fa74122e7ed528126a0fa
Filesize: 2577.196 KB

Version history: (Expand)


CVS history: (Expand)


   2010-11-28 15:31:55 by Matthias Scheler | Files touched by this commit (5)
Log message:
Pullup ticket #3287 - requested by is
lang/ocaml: portability fix

Revisions pulled up:
- lang/ocaml/Makefile.common			1.27
- lang/ocaml/PLIST				1.23
- lang/ocaml/PLIST.opt				1.17
- lang/ocaml/distinfo				1.52
- lang/ocaml/patches/patch-aw			1.8
---
Module Name:	pkgsrc
Committed By:	is
Date:		Sun Nov 28 08:41:33 UTC 2010

Modified Files:
	pkgsrc/lang/ocaml: Makefile.common PLIST PLIST.opt distinfo
	pkgsrc/lang/ocaml/patches: patch-aw

Log message:
Make this package work again on architectures where we don't compile
natively (that is, only to byte-code).

This consists of two parts:

a) a patch to ocamldoc/Makefile to make it create the man pages using
the interpreted ocamldoc - this exists for both types of architectures,
so is safe. (This will be sent up-stream).

b) move a common shared library file to the common PLIST, and a lot of
files (natively compiled versions of the ML modules and natively compiled
versions of a few binaries) to PLIST.opt.

This has been build-tested on i386 (cross-compiled from amd64) and on
arm. unison builds and works.

"make test" shows the same amount of passed and failed tests (mostly
non-found libraries) as before; but this needs more investigation.

An additional issue to solve (with upstream) is that there's no easy way
to run the part of the test suite that would work on byte-code-only
architectures.