Ann: SWI-Prolog 5.6.55

[Jan Wielemaker]

Hi,

I'm happy to announce SWI-Prolog 5.6.55.  Full ChangeLog is below.  Some
highlights:

        * RDF: Fix possible deadlock on literal removing (bug introduced in 5.6.54),
        Fix to predicate hashing, which can give significant speed improvements one
        some operations (notably rdf_save/2).

        * Renamed hash_term/2 --> term_hash/2; added term_hash/4.

        * Added several missing ordered set operations.
       
        * Enhancements to clp(fd) and when/2 (Thanks Markus)
       
        * Enhancements to PlUnit
       
        * Added SICStus 4 compatible library(process).  This notably allows
        creating processes and manage stdin/stdout/stderr of the new process
        in any combination.

        * The usual small stuff ...

        Enjoy

                --- Jan

================================================================
[May 12 2008]

 * PORT: Ported library(process) to Windows

[May 11 2008]

 * ADDED: library(process), providing SICStus 4 compatible process interaction
   to the clib packages.  Currently runs on Unix systems only (tested Linux
and
   MacOS X 10.4).
   
   This patch also extends the foreign interface with the PL_get_signum_ex()
   and PL_cleanup_fork().
   
[May 9 2008]

 * ADDED: Allow specifying sto mode for an entire test set

[May 7 2008]

 * FIXED: toplevel for ground queries using
set_prolog_flag(prompt_alternatives_on, groundness).  Bill Wilson.

[May 5 2008]

 * FIXED: Make plwin.exe menus asking for a file use prolog_file_type/2

 * FIXED: Make PceEmacs mode selection sensitive to prolog_file_type/2.

 * MODIFIED: global_url/3 now raises a syntax error instead of failing if the
   URL is invalid.
   
 * FIXED: rdf_quote_uri/3 not to map & to %26.  Jacco van Ossenbruggen

[May 2 2008]

 * FIXED: on_signal/3 instantiation error

[May 1 2008]

 * DOC: plunit and non termination

[Apr 29 2008]

 * MODIFIED: CLP(FD): At most one labeling option from each category can be
                      specified, and an option must not occur repeatedly.
                      After discussion on the mailing list; thanks to
                      Fernando Sáenz Pérez for opinions and suggestions.
   
[Apr 23 2008]

 * MODIFIED: CLP(FD): respect several optimisation options, which are
                      interpreted from left to right. For example:
                      ?- [X,Y] ins 10..20, labeling([max(X),max(Y)],[X,Y]).
                      %@ X = 20, Y = 20  ;
                      %@ etc.
   
[Apr 28 2008]

 * ADDED: flag tests using `fixme'.  Suggested under the name `desired' by
   Ulrich Neumerkel.  Let us see how this works out in practice!
   
 * ADDED: Bug#354L Support to compile .QLF file contents from a stream.

 * ENHANCED: Perform a GC at the end of a communication with a client.  Note
that
   at that moment we have little active memory (=cheap GC), while otherwise we
may
   remain idle for a long time with large stacks.
   
 * PORT: Make some scripts for building the manuals not depend on the perl
path.
   Jacco van Ossenbruggen.
   
[Apr 27 2008]

 * FIXED: RDF-DB: predicate hash error that caused most predicates to be
   hashed on the same key. Fixes slow rdf_save/2 reported by Jacco van
   Ossenbruggen. Should also positively effect determinism and reliability
   of rdf_estimate_complexity/4.
   
[Apr 26 2008]

 * DOC: Fixed some paths in semweb documentation.  Jacco van Ossenbruggen

[Apr 25 2008]

 * FIXED: Destruction of xterm interactor window and backspace handling

 * ENHANCED: PceEmacs: show location of import

[Apr 24 2008]

 * ENHANCED: When interrupting a thread, indicate which one.

 * ENHANCED: Make interrupt handler use the message system for synchronous
interrupts

 * ADDED: PlUnit: running_tests/0 to find what test is running.

 * FIXED: wrong dynamic declaration in plunit.pl

[Apr 23 2008]

 * ADDED: Allow using --home=DIR to specify the home directory

 * ENHANCED: GUI tracer: expose variable details window

[Apr 22 2008]

 * ADDED: ord_union/2; fixed is_ordset/1 (Ulrich Neumerkel)

 * ADDED: is_ordset/1 to library(ordsets)

[Apr 21 2008]

 * ADDED: ord_intersection/4 to library ordsets.

 * ENHANCED: PceEmacs: show location of autoload (as well as documentation)

[Apr 20 2008]

 * FIXED: Bug#352: atom_concat(X,X,...).  Carl Friedrich Bolz.

[Apr 19 2008]

 * ENHANCED: when/2: more readable residual goals

 * TEST: working

 * DOC: minor typos

 * ENHANCED: when/2: error on invalid condition

 * FIXED: no domain error on ?- when(ground(g), fail).

[Apr 18 2008]

 * ADDED: term_hash/4: depth-bounded term hashes

 * MODIFIED: Renamed hash_term/2 to term_hash/2.  Added hash_term/2 to
   library(backcomp), so most code should not notice this.
   
[Apr 17 2008]

 * FIXED: Avoid deadlock in RDF literal index when removing a literal



------------
For further info, please visit http://www.swi-prolog.org/

To unsubscribe, send a plaintext mail with "unsubscribe prolog <e-mail>"
in its body to majordomo@...

[Jorge Cabrera-2]

Hi Jan,

After upgrading to the new version, the help text links seem to be corrupt.

Thanks.

Jorge Cabrera

------------
For further info, please visit http://www.swi-prolog.org/

To unsubscribe, send a plaintext mail with "unsubscribe prolog <e-mail>"
in its body to majordomo@...

[Jan Wielemaker]

On Tuesday 13 May 2008 14:54, Jorge Cabrera wrote:
> Hi Jan,
>
> After upgrading to the new version, the help text links seem to be corrupt.

Hmmm.  Moved to latest GIT on Windows which does cr/nl mapping, but now the
file-offsets are wrong :-(  I'll see what can be done to fix that.

        --- Jan



------------
For further info, please visit http://www.swi-prolog.org/

To unsubscribe, send a plaintext mail with "unsubscribe prolog <e-mail>"
in its body to majordomo@...

[Jan Wielemaker]

Hi Jorge,

On Tuesday 13 May 2008 14:54, Jorge Cabrera wrote:
> After upgrading to the new version, the help text links seem to be corrupt.

I changed the administration to be line oriented to avoid this trouble.
The changes are in the latest GIT source. I replaced both Windows
installers (32 and 64-bits). Waiting for the next CR/LF nightmare :-(

        Thanks
       
                --- Jan


------------
For further info, please visit http://www.swi-prolog.org/

To unsubscribe, send a plaintext mail with "unsubscribe prolog <e-mail>"
in its body to majordomo@...