[Ecls-list] [ANN] ECL 9.7.1

[Greg Santucci]

Thank you for the new release.

When trying to build on Win32, the build is interrupted when ecl_min.exe
crashes on its first invocation.

On Tue, Jul 14, 2009 at 1:39 AM, Juan Jose Garcia-Ripoll <
juanjose.garciaripoll at googlemail.com> wrote:

> ECL 9.7.1:
> ==========
>
> * Printer:
>
>  - Readtables have a new flag that makes them read-only. Function
>        (si::readtable-lock readtable &optional new-flag)
>   either reads the state of this flag (if NEW-FLAG is missing) or
>   also sets it.
>
>  - The compiler now uses cl_core to efficiently access certain lisp
>   constants (standard packages, standard readtables, etc)
>
>  - EXT:OUTPUT-FLOAT-INFINITY works now fine when *print-circle* is T.
>
>  - WITH-STANDARD-IO-SYNTAX now also rebinds *print-pprint-dispatch*.
>
> * Compiler:
>
>  - When printing out data and reading it from a FASL file, the compiler now
>   uses more strict settings avoiding usual interferences such as the user
>   changing the readtable or the printing rules.
>
>  - The compiler no longer enters the debugger when it finds an error in the
>   code it compiles.
>
>  - C:COMPILER-ERROR no longer derives from ERROR or SERIOUS-CONDITION.
>
>  - Trapping of compiler conditions now works even for compiler errors.
>
>  - Handlers of compiler errors can now call (ABORT). COMPILE-FILE will
>   interrupt the compilation process, and the outputs values that signal
>   the failure to terminate.
>
>  - ECL no longer emits compiler notes when deleting unreachable code.
>
>  - Compiler messages now show the offending form and file position.
>
>  - DEFCONSTANT only had the expected compile-time side effects in the
> bytecodes
>   compiler, not in the lisp->C translator.
>
>  - New type information propagators for AREF, ASET, +, *, /, -, COS, SIN,
> ACOS,
>   ASIN, COSH, ACOSH, ASINH, SINH, ATAN, ATANH, EXP, EXPT, CIS, SQRT, ISQRT
> and
>   ABS. allow ECL to produce better code even without an additional
>   back-propagation phase.
>
>  - Code to perform a type inference pass, with which-way information has
>   been added to ECL. The code is still in a debug phase, with lots of debug
>   output printed and rather incomplete, but it is by default deactivated
> and
>   should not interfere with usual functioning.
>
>  - The code for inlining functions has been improved. Some safe inline
>   expansions that were already available are now actively used thanks to
> these
>   fixes.
>
>  - Formerly, by historical reasons, many library functions that take a
> fixed
>   number of arguments were declared as functions with a variable number of
>   arguments. This slows down ECL and prevents detecting certain errors. In
> an
>   ongoing effort to solve this problem, the C prototypes of COSH, SINH,
> CIS,
>   ACOSH, ASINH and ISQRT now take only one argument (i.e. cl_cosh(cl_object
> x)
>   instead of cl_cosh(cl_narg narg, cl_object x, ...))
>
> * Visible changes:
>
>  - The toplevel only traps SERIOUS-CONDITIONS.
>
>  - "make clean" now wipes out the entire content of the build directory.
> Please
>   use this before re-configuring again ECL so that header and configuration
>   files are properly updated.
>
> --
> Instituto de Física Fundamental, CSIC
> c/ Serrano, 113b, Madrid 28006 (Spain)
> http://juanjose.garciaripoll.googlepages.com
>
>
> ------------------------------------------------------------------------------
> Enter the BlackBerry Developer Challenge
> This is your chance to win up to $100,000 in prizes! For a limited time,
> vendors submitting new applications to BlackBerry App World(TM) will have
> the opportunity to enter the BlackBerry Developer Challenge. See full prize
> details at: http://p.sf.net/sfu/Challenge
> _______________________________________________
> Ecls-list mailing list
> Ecls-list at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ecls-list
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/ecl-devel/attachments/20090714/fee0941b/attachment.html>