release 1.1.3 2008-02-21 ================================================

 "Maintenance" release:
 Corrected overflow of static buffer in boulder_print_pairs()
 and boulder_print_oligos() in primer3_main.c

release 1.1.2 2007-06-10 ================================================

 Updates to dpal.c so that no error forces an exit.
 Adpatation in primer3_main.c and ntdpal_main.c to 
   updated error handling in dpal().
 Arranged for primer3_main.c align() to check for
   and circumvent local alignments to "targets" < 3 nucleotides
   long, and added tests for this condition.
 Comment and code readability updates in dpal.h and dpal.c

release 1.1.1 3/37/2007 ==================================================

 Minor corrections

release 1.1.0 1/17/2007 ==================================================

 Added tags PRIMER_TM_SANTALUCIA, PRIMER_SALT_CORRECTIONS,
   PRIMER_LOWERCASE_MASKING, PRIMER_DIVALENT_CONC,
   and PRIMER_DNTP_CONC.  See the README for
   details of what these tags do.  Briefly, the first two
   tags allow the use of modern melting temperature 
   calculations.  New tests for this functionality
   were added to the tests/ subdirectory

 NOTE: The oligotm executable now takes additional
   arguments.

 Cleaned up Makefile, and created Makefile for the test
   directory with a target to clean up after tests

 Most (but not all) temporary files created by the
   tests now end in .tmp or .tmp2

 Tried to create more informative test for 
   MAX_PRIMER_LENGTH > DPAL_MAX_ALIGN in primer3_main.c

 NOTE: The oligtotm() function is now provided as
   a library, liboligotm.

 NOTE: non-backward compatible change: removed support of
   flag -2x_compat.

release 1.0.1 11/06 ==================================================

 Identical to 1.0b except for release number

release 1.0b 11/06 ==================================================

  Corrected bug involving check for template mispriming
  Minor code cleanup

release 1.0a   9/06 ==================================================

  No new functionality (bug fixes and minor code cleanup only)

  Bug fixes:

    * Corrected calculation of ..._TEMPLATE_MISPRIMING when
      INCLUDED_REGION was set (thanks to Tomoaki Nishiyama, Advanced
      Sciences Research Center, Kanazawa University, Japan).

    * Corrected error that caused crash when
      PRIMER_WT_TEMPLATE_MISPRIMING was set (thanks to Peng Yu,
      Functional Genomics Group II, Chinese National Human Genome
      Center, Beijing).

    * Added olgiotm_main.c (for executable oligotm)

release 1.0   6/04 ==================================================

  Added ability to search source sequence (template) for
  mispriming or mishybing sites.

  Added _experimental_ facility to calculate Tm on long
  oligos using a GC% based formula.

  Miscellaneous fixes and cleanup, including:

    * Avoiding closing NULL file handles when a mispriming or mishyb
      library is missing

    * Correction to long_seq_tm() in oligotm.c plus addition of
      related tests.

    * Gave user the ability to _not_ treat IUB/IUPAC ambiguity
      codes as a consensus while matching against mispriming
      or mishyb libraries.  Inability to turn off this "feature"
      caused severe problems when libraries contained strings
      of N's:  NNNNNNNNN, since if N is treated as a consensus,
      any base matches.

release 0.9  9/98 ==================================================

  Corrected errors in calculation of amplicon Tm.

release 0.8  6/98 ==================================================

  Added flexible objective function.  Various tags governing
  the penalty weights are not fully documented in this
  release.
  
  Added much changed web interface.

  Added ability to select only a single
  left or right primer or hyb probe.

  Made it possible to continue even if 
  PRIMER_{LEFT,RIGHT,INTERNAL_OLIGO}_INPUT is
  not legal.

  Added PRIMER_{LEFT,RIGHT,INTERNAL_OLIGO}_GC_PERCENT
  output tags.

release 0.7  5/98   ==================================================

  Fixed the 'case problem' for
  PRIMER_{LEFT,RIGHT,INTERNAL_OLIGO}_INPUT (and at the
  same time a related bug).

  Added experimental PRIMER_START_CODON_POSITION tag
  and related outputs.  This feature should still be
  considered experimental and the user should scrutinize
  the output.

  Added PRIMER_PRODUCT_{MAX,MIN}_TM and related outputs.

release 0.6   ==================================================

  Bug fixes (most introduced as part of 0.5 enhancements).

release 0.5   ==================================================

  ** Non-Backward compatible change:
  Fixed inconsistency in output format;
  PRIMER_INTERNAL_n_OLIGO_SEQUENCE
  -> PRIMER_INTERNAL_OLIGO_n_SEQUENCE.
  The old format can be preserved by
  compiling with the precompiler
  symbol USE_OLD_FORMAT_MISTAKE defined.

  Added end-anchored local alignment option to dpal.

  Removed some un-needed dpal implementations.

  Added end-anchored local alignments for checking
  mispriming libraries.

  Added mispriming estimates for primer _pairs_
  (see PRIMER_PAIR_MISPRIMING_LIBRARY),
  with related new output tags.  Trivial change
  in maximum allowable library sequence weight.

  Added an experimental primer position component to
  the objective function (PRIMER_{IN,OUT}SIDE_PENALTY),
  with related new output tags.

  Added sequence quality scores as constraints
  in primer picking (PRIMER_MIN_{,END_}_QUALITY,
  PRIMER_QUALITY_RANGE_{MIN,MAX}) with related
  new output tags.

  Added more error reporting to primer3.cgi when primer3
  exits with a non-0 termination status.

  Fixed bug when non-0, non-1 value was supplied
  for pick internal oligos.

  Added PRIMER_{LEFT,RIGHT,INTERNAL_OLIGO}_INPUT
  for checking or for designing around existing primers or
  internal oligos.

  Added error checking for duplicates of certain
  tags, such as SEQUENCE.

  Added PRIMER_MAX_END_STABILITY and related changes
  to output.

release 0.4b   ==================================================

  Corrected bug in parsing of INCLUDED_REGION tag.

release 0.4   ==================================================

  Added optional check of oligos against "mispriming" or
  "mishyb" libraries  with related changes to formated output
  and to the contents of primer list files.

  Modified search algorithm to compute complementarity/
  similarity measures as late as possible.

  Got rid of quotes (") around error and warning strings.

  Fixed numerical overflow on input sequences longer than
  MAX_SHRT.  (Thanks to Pete Young for finding this one.)

release 0.3   ==================================================

  Changes to dpal to make it substantially faster on the maxgap=1
  case for local and end-anchored complementarity measures.

  Removed restriction on maximum sequence length (by removing
  ftar, fexcl, fexcl_int, fnn, fn, fgc arrays, which necessitated
  internal changes to methods for calculating overlap with
  excluded regions and targets and to calculate spanning of
  targets.

  Modified 'primer list' format (clean up and hopefully got rid
  of spurious regression test discrepancy on some systems due to
  differences in rounding).

  Exit codes for some errors changed.

  Text of error messages for PRIMER_PRODUCT_SIZE -> PRIMER_PRODUCT_SIZE_RANGE.

  Some errors that should have been global (fatal) were not.
  These were fixed and error handling rationalized somewhat.
  
  An error in the display of the start position of targets, and
  excluded regions in formated output was corrected.

  An error in which overlapping product size ranges caused
  primer3 to return duplicate primer pairs was corrected.

  Made regression tests more complete.
  
  Made the maximum number of primer pairs to return a 
  parameter (PRIMER_NUM_RETURN).

  Added optional n-based indexing in input and output
  (PRIMER_FIRST_BASE_INDEX=n).  Has not been tested
  for n < 0.

release 0.2b (10/11/96)  ==================================================
  Added PRIMER_WARNING tag.

  Made it legal for excluded regions and targets to lie outside
  of included region.  Documentation fixes (especially to
  PRIMER_DNA_CONC).

release 0.2a ==================================================
  Tweaks to formated output to make it more complete
  (included complementarity measures);  adjusted tests.

release 0.2 ==================================================
  Made 0-length excluded region legal, and adjusted tests.
  Tweaked error reporting of negative-length excluded regions,
  internal oligo excluded regions, and target.


==============================================================
==============================================================
Compared to non-primer3 progenitors:

Compared to primer 0.5
---------------
Completely different input format.  

It has been reported the 0.5 deleted Ns when they occurred in
primers.  

More stringent self-complementarity defaults.

Primer3 selects internal oligos on request (and produces .int
files if requested).

Compared to both primer 0.5 and primer v2
---------------------------
The format of the contents of .for, .rev (and .int) files is
different.

Primer3 returns a user-specifiable number of primer pairs (or
triples) sorted by "goodness".

Primer3 will find a primer pair if any acceptable pair exists.

Optional n-based indexing into source sequence.

Use of sequence quality and 3' stability as constraints in primer
picking.  Optional positional component to objective function.

Compared to primer v2
-------------
Tag name changes.  However, primer3 should understand most or
all Primer v2 input tags, and should produce v2-compatible output
tag names when the -v2_compat command-line switch is used.

The one exception is that the PRIMER_RECOMMEND tag is no longer
produced. Instead primer3 produces the PRIMER_x_EXPLAIN output
tags.  The format of the data in this tags is different from the
data in v2's PRIMER_RECOMMEND output tag.

Numerous fixes.

Uses the PRIMER_SELF_ANY and PRIMER_SELF_END parameters to govern
maximum allowable complementarity between left and right primers,
as well as complementarity between copies of a single oligo or
within a single oligo.  This behaviour is very close to that of
primer 0.5; self complementarity calculations in v2 were
unreliable.

Primer3 produces much more output information, including the TMs
and self complementarity measures of selected primers.