• Uwe Keller (23/23) Apr 27 2009 Hi all,
• Georg Wrede (16/29) Apr 27 2009 ...

Uwe Keller <uwe.keller gmail.com> writes:

Hi all, 

I am happy to announce the first release of the evanescent project. 

evanescent is a collection of tools for reasoning in propositional logic and
provides implementations of inference engines in the D programming language.

The project is meant to be

    * a useful library for reasoning with propositional logic supporting
various computational problems (e.g. SAT) and various algorithms for these
problems.
    * a collection of efficient implementations based on state-of-the-art
techniques
    * a platform for experiments with novel algorithms and implementation
techniques
    * a stress test for the D programming language that allows to evaluate how
suitable D is for developing high performance inference engines for hard
computational problems 

Bottom line: no toy implementations; instead, we aim at building efficient
tools that are able to deal with realistic problems found in industrial
applications. 

Further, we include some demo application showchasing the wide applicability of
propositional logic in computer science 

The released version 0.1 includes:

 - deescover v0.1:  a fast conflict-driven SAT solver
 - deescoverSUDOKU v0.1: a fast SUDOKU solver based on deescover v0.1
 - Some examples that allow to test both solvers

Some first performance measurements for deescover have been done and are
available at the deescover section of the project website.

For more information and download please look at: 

http://code.google.com/p/evanescent/
or
http://www.dsource.org/projects/evanescent/

Try it out! 

We are happy to receive any sort of feedback. 

Cheers,
Uwe.

Georg Wrede <georg.wrede iki.fi> writes:

Uwe Keller wrote:
 Hi all, 
 
 I am happy to announce the first release of the evanescent project. 
 
 evanescent is a collection of tools for reasoning in propositional
 logic and provides implementations of inference engines in the D
 programming language.

...

 a stress test for the D programming language that allows to evaluate how
 suitable D is for developing high performance inference engines for hard
 computational problems 

...

 Bottom line: no toy implementations; instead, we aim at building
 efficient tools that are able to deal with realistic problems found
 in industrial applications.

Congratulations! Reading the web pages was impressive.
And actually I feel the D language could be congratulated, too, for 
gaining such an impressive applicaton as a reference!

But totally ignoring your main point (because of a lack of background in 
propositional logic :-) ), I'd like to ask:

  - What lead to choosing D?
  - What were the alternative languages?
  - Could you share some of your experiences so far with D?
     - The positive encourage us
     - The negative are valuable
  - Is there anything else you'd like to share with us?

--------

The sources seem to indicate D1 and Tango.