V1.54 (05/01/2011)

* A lot of improvements in the Lutin manual
* More Lutin examples
* More Lutin functionalities
    - add loop exactly n times  construct: now one can write 'loop 42' instead of 'loop [42;42]'
    - add support to call lutin program from C, lustre, or luciole 
* Automatically choose a node in the Lutin program if -main is not specified
* Fix a bug in the Lutin variable initialisation


V1.53 (18/11/2010)

* All lucky tools (lucky, xlurette, luc2c, etc.) now accept lutin (.lut) files.
  
V1.52 (18/08/2010)

* Add support for testing ocaml programs via lurette.

V1.51 (8/07/2010)

* New feature: when some SUT or environement inputs are missing, use
    luciole to generate them.

* Enhance the gnuplot-rif visualiser

V1.50 (18/05/2010)

* New feature: Compute a coverage rate. It looks at oracle outputs
which names begins with "covered_", and compute the rate of true
values among those.

V1.49 (8/04/2010)

* Add support to compile Lustre V6 programs.

V1.48 (25/03/2010)

* Lurette now accepts oracles that have more than one output. Only
  the first one should be a boolean, and is the one taken into
  account in order to decide if the test fails or not. All the
  outputs of the oracle are printed in the rif file after the
  #ORACLE_OUTPUT pragma.  That can be very convenient to track why
  the oracle failed.

  
V1.47 (19/03/2010)

* Turn around a bug in ocaml (cf
  http://caml.inria.fr/mantis/view.php?id=4932) that is triggered
  when some Floating-point environment flags (cf fenv.h) are set on.


V1.46 (22/09/2009)

* Use polkag instead of polkai to build the binaries, as the latter raises 
some seg fault when its capacity is exceded.

V1.45 (22/09/2008)

* Fix an embarassing bug in the integer solver (in dimension 1).

V1.44 (4/12/2007)

* Better error messages and a few minor improvements

V1.43 (8/12/2006)

* Add a luc2c tool, which generates C stub files to call lucky
  programs from C via the luc4c lib. luc2c has a --lustre and a
  --scade option that makes it easy to call lucky from lustre v4
  and scade (via C)

* Fix a bug (in polka) where ident could not start with a "_".

* In LuckyDraw, add the possibility to set the maximum and the minimum
  default values to output vars (the default, [-1000,1000], is not always
  suitable).

V1.42

* Fix a couple of Bugs (one in the automata product, one in the bdd traversal)

* Clean-up the distribution, split it up into several packages, etc.

V1.41

* Add External function calls from Lucky.

* The generation of default lucky files from Lurette was broken.

* In some cases (for polyhedron of dim greater that 2 for integers), false
  solutions were generated. Now, we double check that the drawn solution
  is valid. If not, a few other tries are performed. Il all fails, we
  claim that there is no solution, altough is migth not be true. But 
  this is safer than returning wrong solutions. This the price to pay
  for not having a native integer solver (which may be added in future).

* Fix warnings generated by gcc4

V1.40

* Add  a reactive  mode where  the output  of the  previous  cycle is
  returned  when  the  Lucky   environment  is  blocked  (because  no
  transition from  the current node(s)  is labelled by  a satisfiable
  constraint.

* -verbose now takes an integer (verbosity level)

* First version of the integration in the Scade editor

* Replace the Cudd package by an home made (Pascal Raymond) BDD
  package 

V1.39

* Fix a bug where the probability of a translation was sligthly wrong
  in some circumstances  (namely, when a formula happens  to be false
  because of  the numeric  part; in  such a case,  the draw  was done
  again  from the  current node  instead of  doing it  from  the last
  choice point).

* Add  a   button  in   xlurette  to  explicitely   generate  default
  environments for the current SUT.

V1.38

* When  no  environment  is  provided,  also generate  a  file  named
  "<node>_env_UD.luc"   in   addition   to   the   fake   environment
  "<node>_env.luc". That  one is  more elaborated and  contraints the
  output variables so that they move up and down (saw-edge curve).

* Fix a bug due to the fact that not all variable name start with "_"
  in the C code generated by  scade, as it was assumed (and generally
  the case).

* Fix a  bug occurring  for lucky formula  containing more  than 1200
 monomes  because   float  overflew  when  counting   the  number  of
 solutions.  The  fix  consisted  in  using  a  float/logarithm-based
 encoding, where the  mantissa is a float, and  where the exponent is
 an integer.

V1.37

* The    automatically     generated    oracle    in     now    named
  "<sut_node>_always_true.lus" instead of just "always_true.lus".

* fix a  bug where  xlurettte was crashing  down when the  seed field
  (spin button) was empty.

* Allow negative values to be set in xlurette as a seed.


V1.36

* A better  error msg  is displayed when  one try to  write something
  like "x  : int  ^ 3  ~default pre x  ^ 3",  which is  currently not
  supported.

* Fix a bug in gen_fake_lucky that was introduced in V1.33 (incorrect
 environments were generated).

V1.35

* The  file  <node>_io.c generated  by  sildex  is now  automatically
  deleted (it is included by <node>.c, and causes compile errors even
  if lurette  does not use its  content, which is only  meant for the
  sildex simulator).

* Fix a bug in  the step by step mode where the  graph was not always
  refreshed when the step was not 1.

V1.34

* Sildex is now supported

* The RELEASE-NOTES file is created...