Yass-1.01 [dec 2002]
- first public version in december 2002 : in fact, i started this program,
together with many other things that were never finished, during my
master internship that started around march/april 2002... good
old time :-) 

Yass-1.02 [september 2003]
- output entropy filter on alignments
- transition constrainted spaced seeds added
- alignments methods revised (alignement speed ~ x1.3)
- hit criteria revised (speed ~ x1.2)
- many bugs solved 

Yass-1.03 [october 2003]
- BLAST tabular like output
- multifasta support with file selection (-S)
- more scoring matrices
- less global variables ;-)
- autoconf format added (that's helpfull)

Yass-1.04 [november 2003]
- -r 2 option changed
- bug with missed alignments solved (alignandfree take the MAX now)
- -T option only affect direct repeats on one single file ... (no
palindromic ones)
- align algorithm changed (adequate greedy selection made more efficiently)

Yass-1.05 [august 2004]
- Blast tabular output / header
- Matrix cost options
- Lambda/K value can be changed by hand.
- Xdrop modifications / Expectation value to cut off
- Seed has changed (optimised on random sequences / with constrain on
jokers and transitions free elements) 
- several bugs solved (Thanks, Martin !!)
- problem with W,S letters partially solved (replacement) 
- first version of the web interface was done by Ougas ELMI HOUSSEIN [during summer internship]

Yass-1.06 [november 2004]
-  -W option for grouping DNA fragmented alignements
   [ work done by Steven CORROY (in only two months during summer 2004!! good job),
     -  some specifics bugs have been solved since.
     -  several Warnings and optimizations for the -W option.
     -  improved methods both in implementation and also definition.
   ]
-  optimized alignment between seeds and greedy one over seeds (now i have to optimize the core "assemble" part).
-  -O0 non optimization bug removed (strange one, isnt it ?)
-  "||||15120" number on alignment display removed
-  random walk polynom computed now with qpow (faster, of course ;-) )

Yass-1.07   (work done by Christophe VALMIR) [february 2005]
- Cache efficient list structure 
- Add Diagonal hash table        
- One multi fasta BUG solved

Yass-1.08   (work done by Christophe VALMIR) [july 2005]
- Extension MA BUG solved
- Regroup BUG solved
- Lambda and K computed cordance with sequences selected
- Make -C option for modifed  scoring matrices
- Yass used a multi-seed
- The substitution matrix was changed (default:-M 3)
- Bug for compiling MAC solved

Yass-1.09   (work done by Christophe VALMIR) [october 2005]
- thread for the -r 2 option (only with "configure --with-thread").
- -r 2 option is the default option
- Bug for scoring solved.
- More regrouped.

Yass-1.10   (back to LN) [october 2005 - december 2006]
- cosmetic and optimisation work (debugging of -r 2 )
- redefinition of several functions
- key-compress bug solved 
  (now all these options can be selected from the util.h)
- one-multifasta bug solved
- several macro redefined 
- extention algorithms updated
- memory check option for standalone version
- statitics for each thread
- More regrouping.

Yass-1.11 [january 2007 - april 2007]
- lowercase regions parameter
- unusual letters complements bug solved
- entropy on spaced words ...
- output sorted according to 4 criteria (+ 4 )
- new parsing script
- due to a general demand, i have changed the "." and ":", now they
represent tranversions "." and transition ":" repectively
- key only composed of pairs of amino acids are removed
- MEMOPT type bug solved (i forget some copy)

Yass-1.12 [sept 2007 - december 2007]
- single hit added as an option ... group size has been removed.
- low complexity now does not interfer between threads ... global
variables were shared before.
- Gumbel Law Lambda and K are computed with full consideration of the
background nucleotide distribution for both sequences ...
- bug solved when left alignment was computed : some unreleavants
positions were added.
- bug solved when runs of 'N' were complemented.
- bug solved on multiseed span linking
- bug solved on the rightmost alignments
- bug solved on initial -M scoring parameter

Yass-1.13
  "Houston, we've had a problem here"

Yass-1.14 (* work done by Antoine DE MONTE) [january 2008 - november 2009]
* multifasta "full support" (for the first file too, with the "-S 0" parameter)
* multithreading improved, parametrized for the multifasta first file
* 64 bits support (files larger than 2Gb) 
- PSL output format added
- bug solved on multiseed span linking
- bug solved on the rightmost alignments
- bug solved on the alignment_SG_Border part
- bug solved on the min_max_span of the default seed (inverted ?? does
not appear in previous versions)
- bug on the "mask copy" (RedMask function)
- new implementation of the greedy "align tuples" function
- most of the alignments are now provided by the heuristic "Border" functions (banded alignment)
- bug on "-c 1" tuple size (was corrected during extention ... now solved during hit)
- many improvements in the "align.c" main method x 2 ... work in progress on this part ...
