TAP (Thick Ada-Prolog bindings)
Version 0.2
Alexander Kopilovitch
aek@acm.org, aek@vib.usr.pu.ru

************************************************************************
*  This is Release Notes file for the TAP (Thick Ada-Prolog) bindings  *
************************************************************************

Release 0.2

Here is the list of changes from the previous release:

1. Error in Prolog.Synonyms package (missing "return Boolean" in Retract
   function) was corrected, and the parameter name in the Retry function
   was changed (now it matches the name of the corresponding parameter
   of Query function). An appropriate test (Test-Synonyms directory) was
   added. Also, a note about the parameter names in that package is added
   to the TAP manual.
   
2. A drawback in implementation of Remove (Retract) function is eliminated:
   it appeared that a dot (Full_Stop) at the end of argument string cause
   Remove's failure. Now Remove function tests for the ending dot and deletes
   it automatically before calling Logic Server API.

3. An attempt was made to reduce the number of conversions between Standard
   and Unbounded Strings within the Prolog.adb file (for better readability). 

4. Some dead code inside Prolog.adb file was deleted. Also, some other files
   were slightly edited (without significant changes).
   
5. Internal function Unify was renamed to Unify_Predicate_Parameter (in private
   part of Prolog package spec and in bodies of Prolog and Prolog.Unification
   packages).

6. clean.bat and run-all.bat command files were added (for convenience only).

Release 0.1 .

Here is the list of known weak points in the release:

1. Source code doesn't follow any good style.

2. GNAT 3.14p issues 2 warnings for the package Prolog.Library (GNAT 3.13p
   doesn't issue them). These warnings aren't significant, but anyway it would
   be better if they were not appear.
   
3. The choice of the LGPL for licensing status was certainly approximate.
   I have no wish to dive into those issues, but if someone tell me that another
   kind of license may be better for a particular good purpose, I'll consider
   that, and perhaps I'll agree to add another license or replace LGPL with
   another one.

4. I did not try TAP with Unix/Linux, although I believe that it will work
   somehow in that environment, at least partially (that is, without Ada shared
   objects as Logical Extension Libraries).