ChangeLog

2019-05-21  Andy Buckley  <andy.buckley@cern.ch>

	* Apply a few remaining patches from Linux distribution packagings.

2019-05-13  Andy Buckley  <andy.buckley@cern.ch>

	* Fix a bug in line-end parsing, which led reading of multiweight
	events in GenEvent format to abort on every event.

2017-10-09  Andy Buckley  <andy.buckley@cern.ch>

	* Provide a const version of GenVertex::particles().

2017-01-31  Andy Buckley  <andy.buckley@cern.ch>

	* Add GenParticle::ancestors() and descendants() methods.

	* Provide a GenEvent::beams() method, returning a vector, for HepMC3 compatibility.

	* Add a GenParticle::id() forward-compatibility alias for barcode().

	* Change new GenVertex parents and children methods to no-arg particles_in, _out methods to match HepMC3.

2017-01-23  Andy Buckley  <andy.buckley@cern.ch>

	* Change forward Gen{Event,Particle,Vertex}Ptr typedefs to
	#defines to deal with constness semantics problems.

2016-09-27 Andrii Vebytskyi <andrii.verbytskyi@mpp.mpg.de>
	* Tests are working

--------------------------  HepMC2.07-beta00  --------------------------

2016-09-06 Andrii Vebytskyi <andrii.verbytskyi@mpp.mpg.de>
	* Attepms to provide compatibility with 2.06
	* Add publication reference in CPC.
	* Small fixes in examples.

2016-09-01 Andrii Vebytskyi <andrii.verbytskyi@mpp.mpg.de>

	* Make examples working with recent Pythia8 and Herwig6.
	Fix compiller warnings. Fix some tests. Fix configure scripts.
	Small fix to IO_AsciiParticles. Implements fixes from EPEL 2.06 version.

2015-05-08  Andy Buckley  <andy.buckley@cern.ch>

	* Make sure that all macros in HepMCDefs.h have a numerical value
	(thanks to Marek Schoenherr for the heads-up).

2015-04-03  Andy Buckley  <andy.buckley@cern.ch>

	* Adding weight and weight names accessors to GenEvent, for
	forward compatibility.

2015-04-02  Andy Buckley  <andy.buckley@cern.ch>

	* IO_GenEvent.cc: Remove unnecessary copying of the *whole*
	GenEvent during write, since a reference is just fine!

2015-03-27  Andy Buckley  <andy.buckley@cern.ch>

	* Adding forward compatibility typedefs GenEventPtr, GenParticlePtr and GenVertexPtr.

	* Providing HEPMC_VERSION_CODE integer macro for preprocessed version requirements.

	* Adding HEPMC_HAS_CONSISTENT_CONST and HEPMC_HAS_SIMPLE_RANGE macros.

2015-02-15  Andy Buckley  <andy.buckley@cern.ch>

	* Make GenVertex::status() the canonical form, with id() as the alias.

	* HEPMC_HAS_NAMED_WEIGHTS #define changed from implicit value 1 to
	explict value 2, to allow detection of new capabilities.

2015-01-26  Andy Buckley  <andy.buckley@cern.ch>

	* Adding forward-compatibility GenWeights, GenPdfInfo, and GenHeavyIon typedef aliases.

	* Restructuring WeightContainer to work as intended -- weight names are now queryable.

2014-11-25  Andy Buckley  <andy.buckley@cern.ch>

	* Improving IO_AsciiParticles usability -- can now write to non-file streams.

	* Providing children() / parents() functions on GenVertex and GenParticle.

	* Providing preferred upper-case versions of IteratorRange enum values.

	* Add status() and set_status() aliases for Vertex::[set_]id().

	* Add rapidity and abs eta/rap methods to FourVector.

	* More constness improvements (hacked implementation, but improves API).

	* Add abs_pdg_id(), pid() and abspid() methods to GenParticle.

	* Allow CMake in-tree builds.

	* Add user-convenience HepMC/HepMC.h header.

2013-11-17  Andy Buckley  <andy.buckley@cern.ch>

	* Added more iterator access methods on GenVertex.

	* Adding vertices() and particles() aliases on GenEvent.

	* GenRanges clean-up and auto-including at the foot of GenEvent.h

	* Lots of "proper" inlining and noting bugs and TODOs.

2013-11-16  Andy Buckley  <andy.buckley@cern.ch>

	* Merging in code tiding changes from hg branch.

2013-10-16 Witek Pokorski

	* Conversion to Git version control (no code change)

2013-03-25  Lynn Garren

	* HepMC/HeavyIon.h: add centrality (defaults to zero)
	* HepMC/HepMCDefs.h: define HEPMC_HEAVY_ION_HAS_CENTRALITY
	* working release number is 2.07.00.a01

  --------------------------  HepMC-2.06.09  --------------------------

2012-06-05  Lynn Garren

     * configure.ac: explicitly disallow using configure with VC++
     * INSTALL*: improve build instructions

  --------------------------  HepMC-2.06.08  --------------------------

2012-02-16  Lynn Garren

     * src/GenEventStreamIO.cc: make sure we check the stream units
       if there is no unit line in the event header
     * test/testHepMC.cc.in: verify unit behavior

2012-02-10  Lynn Garren

     * cmake/Modules/HepMCVariables.cmake: deal with Windows

     * HepMC/GenEvent.h: add define_units()

  --------------------------  HepMC-2.06.07  --------------------------

2012-01-09  Lynn Garren

     * remove all VCMakefile.in files
     * Use cmake to build for Windows
     * cmake is now the preferred build method

2012-01-05  Lynn Garren

     * src/GenEvent.cc: fix a typo in error report for GenEvent::use_length_unit

     * examples/pythia8: add main31 and main32 from pythia8

2012-01-05  Lynn Garren

     * GenParticle.h, GenVertex.h, PdfInfo.h, SimpleVector.h, SimpleVector.icc:
       Rename local variables to avoid shadowing class variables.
       This problem surfaces when compiling with -Wshadow.

  --------------------------  HepMC-2.06.06  --------------------------

2011-11-09  Lynn Garren

     * add the ability to build with cmake
     * defs.h is neither created nor installed when building with cmake

2011-11-01  Lynn Garren

     * examples: move Pythia and Herwig examples to examples/fio

2011-10-31  Lynn Garren

     * remove fio/PythiaWrapper.cc
     * HepMC/PythiaWrapper.h: HepMC::GenCrossSection getPythiaCrossSection()
       is now an inline method

2011-10-25  Lynn Garren

     * fio/PythiaWrapper.cc: remove struct definitions

  --------------------------  HepMC-2.06.05  --------------------------
2011-04-21  Lynn Garren

      * HepMC/GenVertex.h: add #include <cstddef> for bug #80400

2011-04-15  Lynn Garren

      * test/testFlow.cc, test/testPrintBug.cc: define units
      * test/*.sh.in: make sure comparisons work for allowed
        momentum and length combinations

  --------------------------  HepMC-2.06.04  --------------------------
2011-01-04  Lynn Garren

      * src/HeavyIon.cc, src/PdfInfo.cc: throw if there is an invalid line type
      * test/testHepMC.cc.in, test/testStreamIO.cc.in, test/testHepMCVarious.input:
        provoke a problem with HeavyIon  (see bug #76527)

  --------------------------  HepMC-2.06.03  --------------------------
2010-10-22  Lynn Garren

      * test/testPolarization.cc: verify that Polarization is copied and written

2010-10-20  Lynn Garren

      * src/GenParticle.cc: copy flow when using the copy constructor
      * test/testFlow.cc: add check to be sure flow is copied

  --------------------------  HepMC-2.06.02  --------------------------
2010-08-18  Lynn Garren

      * src/GenEventStreamIO.cc, StreamHelpers.cc:
        cleanup unassociated event remnants if corrupt data is found

2010-07-30  Lynn Garren & Andy Buckley

      * src/GenEventStreamIO.cc: create GenCrossSection, PdfInfo, and HeavyIon
        on the stack rather than the heap to fix a memory leak

  --------------------------  HepMC-2.06.01  --------------------------
2010-06-01  Lynn Garren

      * HepMC/GenRanges.h: remove ConstGenVertexParticleRange since
        there is neither a GenVertex::particle_const_iterator nor a
        GenVertex::vertex_const_iterator
      * HepMC/GenVertex.h: remove unimplemented
        ConstGenVertexParticleRange particles()


  --------------------------  HepMC-2.06.00  --------------------------
  --------------------------  HepMC-2.06.00.b02  --------------------------
2010-05-13  Lynn Garren

      * HepMC/Version.h: version() now takes an optional output stream

      * HepMC/GenEvent.h: add helper method for reading weight names
      * src/GenEventStreamIO.cc: use new helper method to read weight names
      * testHepMCVarious.input: change order of some event info lines
      * test/testIOGenEvent.input: add named weights for testing

2010-05-12  Lynn Garren

      * "make check" now works if momentum units are MEV

2010-05-11  Lynn Garren

      * rename HepMC/GenEventIterators.h to HepMC/GenRanges.h

      * src/GenEventStreamIO.cc: fix a bug reading unnamed (old style) weights

      * test/testHepMC: add more weight testing

  --------------------------  HepMC-2.06.00.b01  --------------------------

2010-04-05  Lynn Garren

      * HepMC/StreamInfo.h: add reading_event_header flag for use by streaming input
      * src/GenEventStreamIO.cc: Input no longer expects a fixed order of
        header lines and will ignore lines that are not of a known type.

      * HepMC/HerwigWrapper.h: add hwbeam
      * rename PythiaWrapper6_2.h to PythiaWrapper6_4.h
        Add structs for more common blocks.

2010-04-02  Lynn Garren

      * HepMC/IteratorRange.h: put the enum in its own header
      * HepMC/GenEvent.h: add vertex_range() and particle_range()
      * HepMC/GenParticle.h: add particles_in(range) and particles_out(range)
      * HepMC/GenVertex.h: add particles(range),
        particles_in(GenParticle,range), and particles_out(GenParticle,range)
      * test/testHepMCIteration.cc: use the new methods

2010-03-20  Lynn Garren

      * HepMC/GenEventIterators.h: add ConstGenVertexParticleRange,
        ConstGenParticleProductionRange, and ConstGenParticleEndRange
      * HepMC/WeightContainer.h: method to check for weight name is has_key()
      * src/GenEventStreamIO.cc: use quotes to delineate weight names

2010-02-08  Lynn Garren

      * remove unused PythiaWrapper5_720.h, PythiaWrapper6_152.h, and PythiaWrapper6_152_WIN32.h
      * combine HerwigWrapper6_4.h and HerwigWrapper.h in a single file
      * HepMC/PythiaWrapper.h: add getPythiaCrossSection function
      * HepMC/HerwigWrapper.h: add getHerwigCrossSection function
      * fio/PythiaWrapper.cc: implement getPythiaCrossSection function
      * fio/HerwigWrapper.cc: implement getHerwigCrossSection function
      * remove redundant examples/HerwigHelper.h.

2010-02-04  Lynn Garren

      * doc/Makefile.am, examples/Makefile.am: use pkgdatadir to define the instation directory
        The default installation directory is now ${prefix}/share/HepMC/doc
        Override with --datadir=/some/directory/path

      * HepMC/Polarization.h: add is_defined() and set_undefined()
        is_defined returns false if the polarization is undefined
	set_undefined unsets the defined flag and zeroes the Polarization

2010-01-28  Lynn Garren

      * HepMC/GenEventIterators.h: This new header defines a set of classes
        which provide the appropriate iterators ranges.  They can be used to
	simplify code and also work directly with such utilities as std::foreach.
	These new classes are
	GenEventVertexRange, ConstGenEventVertexRange,
	GenEventParticleRange, ConstGenEventParticleRange,
	GenVertexParticleRange,
	GenParticleProductionRange, and GenParticleEndRange
      * test/testHepMCIteration.cc: use the new iterator clases and
        also illustrate the use of std::for_each

2010-01-26  Lynn Garren

      * test/testHepMCIteration: illustrate use of std::for_each

2010-01-25  Lynn Garren

      * HepMC/WeightContainer.h: add map-like functionality, allowing users to
        name weights, while retaining the original vector interface for
	complete backwards compatibility
	If a name is not specified for a particular weight, the default
	name is the weight's vector index.
	Also add == and != operators.
      * HepMC/GenEvent.h: add a private method used by the stream I/O
      * src/GenEventStreamIO.cc: add a new line for weight names
        The weight name line is optional when reading input and is
	written only if the event weights object is not empty.
      * test/testWeights.cc: add tests for WeightContainer
      * test/testHepMC: test I/O of named weights
      * examples/testPythiaCopies.cc: set some arbitrary weights

2010-01-09  Lynn Garren

      * removed ParticleDataTable.h and all supporting references.

 --------------------------  HepMC-2.05.01  --------------------------

2010-01-04  Lynn Garren

      * HepMC/GenEvent.h: add write_cross_section(std::ostream&)
      * src/GenEvent.cc: output the cross section (if it exists) when printing an event
      * src/GenVertex.cc, GenParticle.cc: save and restore output stream state
        when calling print methods

      * examples: Calculate Pythia and Herwig cross sections using
        getPythiaCrossSection from examples/PythiaHelper.h and
	getHerwigCrossSection from examples/HerwigHelper.h.
        Set the units explicitly.
      * test/testHepMC.cc.in: call write_cross_section()

2009-12-28  Lynn Garren

       * examples/VectorConversion.h: inline the functions

2009-07-08  Lynn Garren

	* HepMC/Flow.h: improve the syntax of the erase() method
	* test/testFlow.cc: adding a test of the Flow class
	see https://savannah.cern.ch/bugs/index.php?52240

  --------------------------  HepMC-2.05.00  --------------------------

2009-05-19  Lynn Garren

	* HepMC/GenCrossSection: new class to contain the generated
	cross section and cross section error.  This information is
	considered temporary and could change on an event by event basis.
	* write GenCrossSection information with GenEvent whenever the
	cross section has been set.
	* GenEvent.h: add pointer to a CrossSection (null by default).

	* examples: add example_PythiaStreamIO.cc,
	which uses streaming I/O and GenCrossSection.

2009-05-07  Lynn Garren

	* GenEvent: add streaming input and output methods:
	std::ostream& write(std::ostream&);
	std::istream& read(std::istream&);
	Add some private methods used by streaming I/O.

2009-04-17  Lynn Garren

	* GenEvent.h, GenVertex.h: explicitly disallow -- iterator operators.

	* IO_GenEvent: add precision() to set output precision (default is 16).
	IO_GenEvent uses the GenEvent streaming input and output operators.
	Many of the protected methods are no longer useful and have been removed.


	* IO_Exception class added for use when throwing errors.
	If invalid data is encountered when reading an event, input will
	read to the end of the event instead of failing and return an
	empty GenEvent or null pointer.

	* StreamHelpers.h: Define functions used by streaming I/O.
	All are in the detail namespace.

	* StreamInfo: Special class containing extra information needed
	to process HepMC ascii I/O.  This class is initialized by a call
	to a custom iomanip when any I/O stream is initialized and deleted
	by the stream destructor.  Each stream has a separate instance.

	* CommonIO is no longer necessary and has been removed.

	* GenEvent:  Add streaming input and output operators.
	Add free functions to be used with streaming I/O: set_input_units,
	write_HepMC_IO_block_begin, and write_HepMC_IO_block_end.

2009-03-02  Lynn Garren

	* HeavyIon:  Add streaming input and output operators.
	Add is_valid() method.

	* PdfInfo:  Add streaming input and output operators.
	Add is_valid() method.

2009-01-29  Lynn Garren

	* HepMC/SimpleVector: remove FourVector::mag() and ThreeVector::mag()
	Change test/testSimpleVector.cc and test/testMass.cc to reflect this.

	* HepMC/HepMCDefs.h: New header containing HEPMC_HAS_UNITS and
	similar macros so users can check for various code features.
	HepMC/GenEvent.h includes HepMCDefs.h.
	* HepMC/Version.h: Use new HEPMC_VERSION macro.

	* HepMC/GenParticle: add is_undecayed(), has_decayed(),
	and is_beam_particle() convenience methods.
	Test these methods in test/testHepMC.

2009-01-28  Lynn Garren

	* remove HepMC/IO_Ascii.h and test/testDeprecated

	* remove HepMC/HepMC_CLHEP20.h

  --------------------------  HepMC-2.04.01  --------------------------

2008-12-15  Lynn Garren

	* src/CommonIO.cc: add space requested by -Wextra
	The -Wextra flag is not a default compiler flag, request it with
	configure ... CXXFLAGS="-Wextra"

2008-12-11  Lynn Garren

	* test/testHepMCIteration.cc.in: add loop over children to the test

	* src/IO_Ascii.cc, IO_GenEvent.cc: fix error messages

2008-06-24  Lynn Garren

        * fio/IO_HEPEVT.cc: bug fix for inconsistency in HWHGUP
        see https://savannah.cern.ch/bugs/?38173 for details

  --------------------------  HepMC-2.04.00  --------------------------

2008-06-04  Lynn Garren

	* doc/HepMC2_user_manual.tex: expand description of Ascii IO format

2008-05-29  Lynn Garren

	* configure.ac: Use libtool versioning for libraries

2008-05-27  Lynn Garren

	* configure.ac: Throw an error immediately if units are not specified.

  --------------------------  HepMC-2.04.00.beta2  --------------------------

2008-05-21  Lynn Garren

	* HepMC/GenEvent.h: Duplicate the existing constructors with new
	constructors that have momentum and length as the first 2 arguments.

	* examples: all examples except example_BuildEventFromScratch,
        testHerwigCopies, and testPythiaCopies use default units

	* HepMC/IO_GenEvent.h: add use_input_units() method
	This method is needed ONLY if a) the input stream has no
	unit information and b) it is known that the units of the input
	stream do not match the default units.

2008-05-20  Lynn Garren

	* revamp Units design completely
	Replace MomentumUnits.h and PositionUnits.h with Units.h
	Allowed unit designations are GEV, MEV, MM, and CM

	* configure.ac: User MUST specify both --with-momentum_units=XX
	and --with-length_units=YY.  There are no defaults.

	* HepMC/GenEvent.h: unit data members are now enums

2008-05-12  Lynn Garren

	* HepMC/Pdfinfo.h: Input parton flavour codes id1 and id2 are expected
	to obey the PDG code conventions, especially g = 21.

  --------------------------  HepMC-2.04.00.beta  --------------------------

2008-05-05  Lynn Garren

	* configure.ac: use separate makefiles (VCMakefile.in) for Visual C++
	The custom makefiles are invoked by default when using Visual C++.
	* doc: Stop building postscript copies of the manuals, just
	build and distribute PDF copies of the manuals.

2008-04-29  Lynn Garren

	* remove deprecated IO_ExtendedAscii

2008-04-25  Lynn Garren

	* cleanup examples: combine example_MyPythia.cc,
	example_MyPythiaRead.cc, example_MyPythiaWithEventSelection.cc,
	example_PythiaParticle.cc, and example_ReadMyPythia.cc into one example with multiple subsets.
	* add tests to list of examples in the reference manual

2008-04-23  Lynn Garren

	* src/GenVertex.cc: fix bug in edge_iterator operator++
        * src/IO_Ascii.cc: fix for VC9
        * src/IO_ExtendedAscii.cc: fix for VC9
        * src/IO_AsciiParticle.cc: fix for VC9

2008-04-18  Lynn Garren

	* HepMC/GenEvent.h: minor bug fix for gcc 4.3.0
	* HepMC/IO_ExtendedAscii.h: minor bug fix for gcc 4.3.0
	* HepMC/IO_GenEvent.h: minor bug fix for gcc 4.3.0
	* src/IO_ExtendedAscii.cc: minor bug fix for gcc 4.3.0
	* src/IO_GenEvent.cc: minor bug fix for gcc 4.3.0

2008-04-18  Lynn Garren

	* ReadMe.cygwin: changing cygwin directions for new versions of VC

	* HepMC/MomentumUnits.h: enum HepMCmomentumUnits{ UNKNOWN = 0, MEV, GEV };
	Recommended usage: MomentumUnits::GEV
	* HepMC/PositionUnits.h: enum HepMCpositionUnits{ unknown = 0, MM, CM }
	Recommended usage: PositionUnits::MM

2008-04-17  Lynn Garren

	* Rename configure.in to configure.ac, which is now the common usage.
	* configure.ac: Major restructuring.  Using libtool.
	Only Linux, MacOSX, and Windows are explicitly supported.
	All Makefile.am's have been restructured to use libtool.

2008-04-01  Lynn Garren


	* HepMC/IO_GenEvent.h: replace const char* filename with
	  const std::string& in the constructor accepting a filename.

2008-03-31  Lynn Garren

	* HepMC/GenEvent.h: random_states() now returns const std::vector<long>&
	  HepMC/GenVertex.h: position() now returns a const FourVector&
	  HepMC/GenParticle.h: flow() now returns const Flow&
	  and polarization() now returns const Polarization&

	* HepMC/IO_GenEvent.h: add a constructor that accepts the filename
	  as a string.

2008-03-28  Lynn Garren

	* HepMC/GenParticle.h: The momentum() method now returns a const
	  reference to a FourVector instead of a copy of a FourVector.
	  This change should be backwards compatible.

2008-03-27  Lynn Garren

	* examples: set Herwig and Pythia units to GeV and mm

2008-03-26  Lynn Garren

	* HepMC/MomentumUnits.h: new class to encapsulate momentum units,
	  which are represented by a single enum
	* HepMC/PositionUnits.h: new class to encapsulate position units,
	  which are represented by a single enum
	* GenEvent: add MomentumUnits and PositionUnits data members and
	  associated methods
	* GenVertex: add protected convert_position method
	* GenParticle: add protected convert_momentum method

	* HepMC/PdfInfo.h: add pdf_id1(), pdf_id2(), set_pdf_id1(), and
	  set_pdf_id2() for the pdf set id numbers

	* IO_GenEvent: get and put unit information in a new line
	  The first character of the unit line is "U".
	* IO_GenEvent: process the new PdfInfo data members

	* test/testUnits.cc: new test for MomentumUnits and PositionUnits

  --------------------------  HepMC-2.03.08  --------------------------

2008-04-23  Lynn Garren

	* src/GenVertex.cc: fix bug in edge_iterator operator++
        * src/IO_Ascii.cc: minor fix for VC9
        * src/IO_ExtendedAscii.cc: minor fix for VC9
        * src/IO_AsciiParticle.cc: minor fix for VC9

  --------------------------  HepMC-2.03.07  --------------------------

2008-04-18  Lynn Garren

	* HepMC/GenEvent.h: minor bug fix for gcc 4.3.0
	* HepMC/IO_ExtendedAscii.h: minor bug fix for gcc 4.3.0
	* HepMC/IO_GenEvent.h: minor bug fix for gcc 4.3.0
	* HepMC/ParticleData.h: minor bug fix for gcc 4.3.0
	* src/IO_ExtendedAscii.cc: minor bug fix for gcc 4.3.0
	* src/IO_GenEvent.cc: minor bug fix for gcc 4.3.0

  --------------------------  HepMC-2.03.06  --------------------------

2008-03-12  Lynn Garren

	* test: reduce size of input files

	* IO_HEPEVT: add trust_beam_particles and set_trust_beam_particles
	  methods so the user can specify that no beam particles are available.
	  This fixes a problem with pythia6 input when the user chooses "none".

2008-03-11  Lynn Garren

	* src/GenEvent.cc: The GenVertex pointer to the parent event was
	  not properly moved when the operator= method was used to copy
	  a GenEvent.  That bug is now fixed.
	* GenVertex: added protected change_parent_event_ method for swap
	  as part of bug fix.
	* test/testMultipleCopies.cc.in: Add explicit tests for operator=
	  and swap.

	* examples: minor tweaks for readability

  --------------------------  HepMC-2.03.05  --------------------------

2008-02-22  Lynn Garren

	* fio/IO_HEPEVT.cc:  According to standard HepMC design, all vertices
	in an event are expected to contain particles.  If, for instance, the
	first two particles have no parents, then they should simply not have
	a production vertex.  However, in at least one instance empty production
	vertices were created for the first two particles. IO_HEPEVT now
	protects against this.


2008-02-18  Lynn Garren

	* HepMC/HEPEVT_Wrapper.h: fix calculation of the total size
	of the common block

  --------------------------  HepMC-2.03.04  --------------------------

2008-02-05  Lynn Garren

	* test: use /bin/bash instead of /bin/sh in shell scripts

	* GenEvent, GenVertex, GenParticle: remove static counters
	These counters were only for the debugging convenience of the
	HepMC code developer.  They were not accessible by users.

2008-02-04  Lynn Garren

	* fio/IO_HERWIG.cc: fix for Z0 mother-daughter problem (from Atlas)

	* HepMC/SimpleVector.h: Remove simple math methods

2007-01-29  Lynn Garren

	* examples/testHerwigCopies.cc: compare copies of Herwig events

	* examples/testPythiaCopies.cc: compare copies of Pythia events

2007-01-15  Lynn Garren

	* TempParticleMap.h: use barcode instead of counter for ordering
	to solve a problem with changed order of particles when making copies

2007-01-14  Lynn Garren

	* HepMC/CompareGenEvent.h: new utility to compare two
	copies of a GenEvent object

	* test/testMultipleCopies.cc.in: verify that two copies of
	the same GenEvent object are, indeed, the same

	* test/testHepMCIteration: use a different output file name so the
	test will work if you build on top of the source code


  --------------------------  HepMC-2.03.03  --------------------------

2007-12-17  Lynn Garren

	* HepMC/SimpleVector.h: Add simple math methods:
	FourVector   operator +  (const FourVector &) const;
	FourVector & operator += (const FourVector &);
	FourVector   operator -  (const FourVector &) const;
	FourVector & operator -= (const FourVector &);
	FourVector   operator -  () const;
	FourVector & operator *= (double);
	The same methods were also added to ThreeVector.

  --------------------------  HepMC-2.03.02  --------------------------

2007-12-14  Lynn Garren

	* doc: The HepMC 1 and HepMC 2 user manuals have been merged
	into a single manual.

	* HepMC/IO_GenEvent.h: check operator methods for consistency

2007-12-07  Lynn Garren

	* HepMC/IO_GenEvent.h: fix problem with rdstate() and clear()

2007-12-04  Lynn Garren

	* HepMC/IO_Ascii.h, IO_ExtendedAscii.h, IO_GenEvent.h and
	  src/IO_Ascii.cc, IO_ExtendedAscii.cc, IO_GenEvent.cc:
	  common input methodologies now handled by CommonIO

2007-12-03  Lynn Garren

	* HepMC/CommonIO.h, src/CommonIO.cc:
	  new header for shared IO methods and keys

	* src/IO_GenEvent.cc: use CommonIO
	  search for any of several allowed input data types

	* HepMC/IO_GenEvent.h: disable write_particle_data_table and
	  fill_particle_data_table - that is best done outside HepMC

  --------------------------  HepMC-2.03.01  --------------------------

2007-11-26  Lynn Garren

	* configure.in: add support for g++-*

2007-11-26  Lynn Garren

	* src/IO_GenEvent.cc: bug fix - fill event scale, alphaQED, and alphaQCD

  --------------------------  HepMC-2.03.00  --------------------------

2007-10-30  Lynn Garren

	* examples/example_ReadMyPythia.cc: New example to read the file
	  written by example_MyPythia.

	* doc: remove the latex subdirectory after building the reference library

	* GenEvent: add print_version( std::ostream& ) method.
	  If you don't specify an ostream, the default is std::cout.

	* GenVertex: Store particles in vectors instead of sets.
	  Root IO does not properly restore the state of the particle
	  serial number counter, so the previous solution to retaining
	  consistent particle ordering within a vertex had to be revisited.

	* IO_GenEvent: Replace the temparary particle map with TempParticleMap.

	* HepMC/SearchVector.h: Utilities to facilitate working with the
	  vector of particles.

	* HepMC/TempParticleMap.h: Temporary particle container that
	  allows us to maintain particle ordering when reading an event
	  with IO_Ascii and IO_ExtendedAscii.

	* Use fully qualified names (e.g., HepMC::GenParticle) inside class
	  templates (e.g., std::map<>) wherever they are used within a
	  header so that rootcint can make the dictionary.

  --------------------------  HepMC-2.02.01  --------------------------

2007-10-16  Lynn Garren

	* configure.in: build dynamic libraries for MacOSX - thanks to James Monk

  --------------------------  HepMC-2.02.00  --------------------------

2007-07-19  Lynn Garren

	* IO_GenEvent: uses input and output streams
	  IO_GenEvent can be constructed with a file, an input stream, or
	  an output stream.
	* IO_Ascii and IO_ExtendedAscii are deprecated.
	* examples use IO_GenEvent

  --------------------------  HepMC-2.01.10  --------------------------

2008-02-05  Lynn Garren

	* test: use /bin/bash instead of /bin/sh in shell scripts

	* doc: The HepMC 1 and HepMC 2 user manuals have been merged
	into a single manual.

2008-02-04  Lynn Garren

	* fio/IO_HERWIG.cc: fix for Z0 mother-daughter problem (from Atlas)

2008-01-14  Lynn Garren

	* test/testHepMCIteration: use a different output file name so the
	test will work if you build on top of the source code

2007-12-17  Lynn Garren

	* HepMC/Flow.h: comment changes for doxygen

	* HepMC/GenParticle.h, src/GenParticle.cc: remove remnant static counter

        * Use fully qualified names (e.g., HepMC::GenParticle) inside class
          templates (e.g., std::map<>) wherever they are used within a
          header so that rootcint can make the dictionary.

	* src/IO_ExtendedAscii.cc: make sure all info is filled when reading

	* configure.in: add support for g++-*
	build dynamic libraries for MacOSX - thanks to James Monk

  --------------------------  HepMC-2.01.08  --------------------------

2007-10-23  Lynn Garren

	* GenVertex: Store particles in vectors instead of sets.
	  Root IO does not properly restore the state of the particle
	  serial number counter, so the previous solution to retaining
	  consistent particle ordering within a vertex had to be revisited.

	* IO_Ascii, IO_ExtendedAscii: Replace the temparary particle map
	  with the more sophisticated TempParticleMap.

	* HepMC/SearchVector.h: Utilities to facilitate working with the
	  vector of particles.

	* HepMC/TempParticleMap.h: Temporary particle container that
	  allows us to maintain particle ordering when reading an event
	  with IO_Ascii and IO_ExtendedAscii.

  --------------------------  HepMC-2.01.07  --------------------------

2007-10-19  Lynn Garren

	* Use fully qualified names (e.g., HepMC::GenParticle) inside class
	  templates (e.g., std::map<>) wherever they are used within a
	  header so that rootcint can make the dictionary.

  --------------------------  HepMC-2.01.06  --------------------------

2007-08-07  Lynn Garren

	* HepMC/GenEvent.h, GenVertex.h,GenParticle.h : add swap method
	* src/GenEvent.cc, GenVertex.cc, GenParticle.cc:
	  use best practices for assignment and copy constructor

	* HepMC/SimpleVector.h: add swap method
	* HepMC/Flow.h: add swap method
	* HepMC/WeightContainer.h: add swap method

	* HepMC/Polarization.h: add swap method
	  add private valid_theta and valid_phi methods for use by constructors
	  src/Polarization.cc: use best practices for assignment and copy constructor

  --------------------------  HepMC-2.01.05  --------------------------

2007-07-19  Lynn Garren

	* HepMC/HeavyIon.h: make it nice for rootcint

  --------------------------  HepMC-2.01.04  --------------------------

2007-07-18  Lynn Garren

	* various bug fixes and protections

  --------------------------  HepMC-2.01.03  --------------------------

2007-07-11  Lynn Garren

	* src/IO*.cc: write HepMC version number

  --------------------------  HepMC-2.01.02  --------------------------

2007-07-10  Lynn Garren

	* HepMC/Version.h: add the ability to print the HepMC version number
	* src/IO_ExtendedAscii.cc: write HepMC version number

2007-07-09  Lynn Garren

	* src/IO_ExtendedAscii.cc: read/write beam particle information
	  fix a problem with reading optional info
	* test/testMass.*: compare generated and calculated masses

2007-06-29  Lynn Garren

	* HepMC/GenEvent.h, src/GenEvent.cc: define beam process methods
	   bool valid_beam_particles() const;
	   std::pair<GenParticle*,GenParticle*> beam_particles() const;
	   bool set_beam_particles(GenParticle*, GenParticle*);
	   bool set_beam_particles(std::pair<GenParticle*,GenParticle*> const &);

	* fio/IO_HEPEVT.cc, fio/IO_HERWIG.cc: set incoming beam particles

	* HepMC/PythiaWrapper*: remove old process struct

2007-06-21  Lynn Garren

	* HepMC/GenEvent.h: add methods to set and access the number of
	  multi particle interactions in the event.
	  The default value of this number is -1.
	* src/IO_ExtendedAscii.cc: read and write MPI information

2007-06-19  Lynn Garren

	* HepMC/GenParticle.h: use uint64_t for the serial number counter

  --------------------------  HepMC-2.01.00  --------------------------

2007-06-08  Lynn Garren

	* HepMC/GenEvent.h, src/GenEvent.cc: add clear() method

	* HepMC/GenParticleComparison.h:
	  define a GenParticle comparator
	* HepMC/GenParticle.h, src/GenParticle.cc:
	  set a serial number to be used by GenParticleComparison
	* HepMC/Flow.h, HepMC/GenVertex.h, src/Flow.cc, src/GenVertex.cc:
	  use std::set<GenParticle*,GenParticleComparison>
	* test: automatic comparison of output files is now possible

  --------------------------  HepMC-2.00.04  --------------------------

2007-06-01  Lynn Garren

	* HepMC/GenEvent.h, src/GenEvent.cc: change interface such that
	  GenEvent makes its own copy of HeavyIon and PdfInfo


2007-05-29  Lynn Garren

	* src/GenVertex.cc: initialize range in default constructor
	* src/IO_Ascii*.cc: make sure every object is properly deleted
	* test/testHepMCIteration.cc: add iterator test

  --------------------------  HepMC-2.00.03  --------------------------

2007-04-23  Lynn Garren

	* bootstrap, configure.in, doc/Makefile.am:
	  To avoid problems with different versions of latex, just build
	  the documents during the bootstrap step.
	* doc/buildDoc.sh: builds the documents if doxygen and latex are present

2007-04-20  Lynn Garren

        * HepMC/include/PythiaWrapper*.h:
          inline the method definitions to avoid conflicts
	* examples/initPythia.cc: put pythia intialization in a separate
	  file so we can test PythiaWrapper

  --------------------------  HepMC-2.00.02  --------------------------

2007-02-12 Lynn Garren

	* test/testHepMC.sh.in: works on Windows (Cygwin), MacOSX, and Linux

2007-02-07 Lynn Garren

	* doc/doxygen-template: provide an html template

2007-01-30 Lynn Garren

	* doc/doxygen.conf: use doxygen 1.5.1

2007-01-30 Lynn Garren

	* examples/example_BuildEventFromScratch.cc: add example of
	  conversion from SimpleVector to HepLorentzVector
	* examples/VectorConversion.h: VectorConversion.h is meant as a
	  template for user code converting to the vector of your choice.

2007-01-29 Lynn Garren

	* fix everything so doxygen will create a useful reference manual
	  nearly every file has been touched, but only the comments changed


2007-01-25 Lynn Garren

	* examples/example_MyPythiaRead.cc: read in the events you just wrote
	* examples/*.cc: put all uses of IO_Ascii within an explicit scope

	* configure.in, doc/Makefile.am: if latex is found, get the path

	* test/testHepMC.sh.in: make the diff work for MacOSX

	--------------------  HepMC-02-00-01  ---------------------------------

2007-01-17 Lynn Garren

	* test/testSimpleVector.cc: exercise the vector methods

2007-01-16 Lynn Garren

	* HepMC/SimpleVector.icc: implement ThreeVector::set(x,y,z)

2006-08-23 Lynn Garren

	* install examples in $(prefix)/examples/HepMC

	* check for latex and build documents if latex is present

	--------------------  HepMC-02-00-00  ---------------------------------

2006-08-22 Lynn Garren

	* src/IO_ExtendedAscii.cc: read and write generated mass as part of
	  particle line.

2006-08-18 Lynn Garren

	* doc/HepMC2_user_manual.tex documentation for HepMC 2

2006-08-01 Lynn Garren

	* HepMC/is_arithmetic.h, HepMC/enable_if.h supplied by Walter Brown
	  for a clean template constructor implementation.

2006-07-23 Lynn Garren

	* HepMC/SimpleVector.h: add a templated constructor for both FourVector
          and ThreeVector that will take any lorentz vector which has the
          x(), y(), z(), and t() methods.  This should enable existing code
          that uses CLHEP Vector classes to keep working.

2006-07-19 Lynn Garren

	* replace CLHEP/Vector/LorentzVector.h and CLHEP/Vector/ThreeVector.h
	  with HepMC/SimpleVector.h
	  implement some of the basic vector properties so user code won't break


	--------------------  HepMC-01-28-00  ---------------------------------

2006-08-08 Lynn Garren

	* HepMC/IO_ExtendedAscii.h: Extended format writes PdfInfo and HeavyIon
	  if they are present in the event.  This is otherwise identical
	  to IO_Ascii.h.

2006-07-30 Lynn Garren

	* HepMC/HEPEVT_Wrapper.h: fix for 64bit machines

2006-07-26 Lynn Garren

	* HepMC/PdfInfo.h implements information requested by CMS
	* GenEvent.h, GenEvent.cc make sure m_pdf_info is initialized to 0

	--------------------  HepMC-01-27-02  ---------------------------------

2006-06-19 Lynn Garren

	* fix test for Windows
	* add ReadMe.cygwin-VC71 and setup.cygwin-VC71


2006-06-14 Lynn Garren

	* GenEvent.cc, GenParticle.cc,  GenVertex.cc use standard C++ output
	* testPrintBug new test for output problems with gcc 4.x
	* HepMC/HepMC_CLHEP20.h defines several typedefs needed when compiling
	   with CLHEP 2.0.x
	* HEPEVT_Wrapper.h, HerwigWrapper6_4.h, PythiaWrapper6_2.h
	   need extern "C" statements for gcc 4.x
	* HepMC_CLHEP20.h, Polarization.h, GenVertex.h, GenParticle.h
	   HepMC will work with both CLHEP 1.9.x and 2.0.x


	--------------------  HepMC-01-27-01  ---------------------------------

2006-03-31 Lynn Garren

	* GenEvent.h, GenEvent.cc make sure m_heavy_ion is initialized to 0


2006-03-29 Lynn Garren

	* HepMC/ParticleData.h remove dependency on CLHEP/Units by HepMC_hbarc
	* HepMC/Polarization.h remove dependency on CLHEP/Units by HepMC_pi

	--------------------  HepMC-01-27-00  ---------------------------------

2006-03-07 Lynn Garren

	* add simple check in the test subdirectory

2006-02-17 Lynn Garren

	* HepMC/HeavyIon.h names have been changed to match both HepMC and
	wishes of heavy ion users

2006-01-12 Lynn Garren

	HepPDT 1.26 is available at https://savannah.cern.ch/projects/hepmc/

	* HepMC/HeavyIon.h implements information requested by CMS
	* HepMC/GenParticle.h has pointer (null by default) to HeavyIon
	* code providing interfaces to Fortran common blocks has been moved
	out of libHepMC and into libHepMCfio

	--------------------  HepMC-01-02-26  ---------------------------------

	--------------------  HepMC-01-02-21  ---------------------------------

2005-04-27 Matt Dobbs and Giorgos
	* src/IO_HERWIG.cxx implemented a bug fix suggested by borut,
	which keeps the remapping of daughters/motherrs from going "over
	the end" on herwig events (was noticed in the tauola events for
	rome DC).

2004-08-04  David Quarrie  <David.Quarrie@cern.ch>

	* cmt/requirements: Change dependencies on CERNLIB and CLHEP
        to AtlasCERNLIB and AtlasCLHEP to avoid name clashes


	--------------------  HepMC-01-02-19  ----(gcc 3.2 compatible)---------


2004-04-22  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* src/IO_HERWIG.cxx (HepMC): fixed bug reported by Hinchliffe in
	IO_HERWIG wherein (for the special case of min bias events only)
	number of entries in HEPEVT was being zero-ed by the IO_HERWIG
	class. Added extra protection in HEPEVT_Wrapper for this as well.


	--------------------  HepMC-01-02-18  ----(gcc 3.2 compatible)---------

2004-03-27  Matt Dobbs  <Matt.Dobbs@Cern.CH>

        Removed the #define statements that allow backwards compatibility
	with gcc 2.95, and specialized the 3.2 compliant iterators to be
	forward iterators such that they work properly with std
	algorithms.

	Thanks to Ulrik Egede for pointing this out.

	* HepMC/GenVertex.h, .cxx HepMC/GenEvent.h, .cxx: changed
	std::forward_iterator<GenParticle*,ptrdiff_t>
	to
	std::iterator<std::forward_iterator_tag,GenParticle*,ptrdiff_t>
	for standard
	compliance, and removed #define statements.

	* HepMC/IO_Ascii.h,.cxx: changed open_mode to openmode for standard
	compliance, and removed #define statements.


2003-12-17  David Rousseau  <droussea@lxplus072.cern.ch>

	* cmt/requirements: remove HepMC_libraries

2003-10-22  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* src/GenEvent.cxx and lots of other files + latex documentation.
	REmoved any mention of units--- GeV, MeV etc, since HepMC merely
	accepts whatever is put in. I do this because, though HEPEVT has
	used GeV/mm, ATLAS plans to use CLHEP units of MeV/mm.

2003-10-15  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* src/GenParticle.cxx (HepMC): added a default constructor for
	GenParticle as requested by Giorgos S.

	--------------------  HepMC-01-02-14  ----(gcc 3.2 compatible)---------

2003-04-21  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* src/IO_HERWIG.cxx:
	minor changes to handling of herwig event record

	--------------------  HepMC-01-02-11  ----(gcc 3.2 compatible)---------

2003-03-10  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* HepMC/IO_BaseClass.h:
	Fixed bug reported by Peter Loch, wherein the virtual destructor
	for IO_BaseClass was missing.

	--------------------  HepMC-01-02-10  ----(gcc 3.2 compatible)---------

2003-03-03  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* HepMC/GenVertex.h: added extra ifdef statement to allow
	compilation on gcc 3.2. (forward_iterator and open_mode problems).
	* HepMC/IO_Ascii.h: as above.
	* src/IO_Ascii.cxx (HepMC): as above.

2003-01-17  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* src/IO_PDG_ParticleDataTable.cxx (HepMC): if the PDG table does
	not exist, we abort, rather than just returning false.

	--------------------  HepMC-01-02-04  ----(5.0.0)------------
2002-11-04  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* src/GenEvent.cxx (HepMC): set_barcode(particle, barcode),
	set_barcode(vertex, barcode) Fixed a bug reported by Giorgos S.
	For this bug, a user suggests a barcode for a vertex that is
	already part of an event... the vertex is inserted in the vertex
	map with the new barcode, but HepMC forgets to erase the old
	entry... such that the vertex now appears twice in the map.  This
	is fixed, and the corresponding error for particles is also fixed.

2002-10-31  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* HepMC/IO_HERWIG.h: modified IO_HERWIG to include a switch
	(default m_no_gaps_in_barcodes=on) which removes null entries from
	Herwig HEPEVT and reshuffles the common block

2002-10-22  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* HepMC/HerwigWrapper6_4.h, HepMC/HerwigWrapper.h: new interface
	to HERWIG Version 6.4 is added. There is an example for using it
	at examples/example_MyHerwig.cxx. To use this example you will
	have to download HERWIG version 6.4 and modify the Makefile to
	link it.

	* HepMC/IO_HERWIG.h: New class for reading the Herwig version of
	the HEPEVT common block is added. READ THE COMMENTS IN THE .h file
	carefully before using it!


2002-07-29  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-02-03  ----(4.0.2)------------

	* src/GenVertex.cxx (HepMC): remove_particle()
	* HepMC/GenVertex.h:
	removed confusing comment about the use of GenVertex::remove_particle()
	in .h file, and clarified the comment in .cxx file.

	* src/GenEvent.cxx (HepMC): operator=
	bug discovered by <Malte.Muller@cern.ch>.
	Formerly, the GenEvent::operator= method relied on the particles
	attached to each vertex to always be in the same order. However,
	since the particles are stored in set<>, this is not always true
	-- making the results of the operator= method unreliable.
	This is fixed by mapping the vertices explicitly.

	* src/IO_PDG_ParticleDataTable.cxx (HepMC):
	added a line in read_entry that ignores empty lines (previously
	empty lines in the pdg data table generated a cerr message).
	This change was requested by Davida Costanzo.


2002-03-02  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-02-02  ----(3.0.1)------------

	* src/GenVertex.cxx (HepMC):
	GenVertex::edge_iterator::edge_iterator( .. )
	Extra error protection added in the constructor.
	Sometimes the edge_iterator fails when we have a vertex with no
	in_particles. This never happens in standalone mode, but when ran
	with the ATLAS framework, it does happen. New code identifies the
	scenario explicitly, and catches the error, which has to do with
	the past-the-end value of the edge_iterator.

	* test/test_iterators.cxx: new test program added for iterators,
	to test for the above problems.


2002-02-19  Matt Dobbs  <Matt.Dobbs@Cern.CH>

        CHANGES TO ALLOW HEPMC TO COMPILE WITH SOLARIS CC 5.2

	* cmt/requirements: added the line
	macro_append cppflags "" Solaris      " -D__SUNPRO_CC "
	to handle Solaris CC 5.2 features.

	* HepMC/ParticleDataTable.h:
	turns on MISSING_FORWARD_ITERATOR when __SUNPRO_CC is defined.
	This defines forward_iterator as iterator

	* HepMC/GenVertex.h:
	same as ParticleDataTable.h, but also turns on
	NEED_SOLARIS_FRIEND_FEATURE when __SUNPRO_CC is defined.

	* src/Polarization.cxx (HepMC):
	replaced abs() with theta = ( theta>0 ? theta : -theta );
	for compatibility with Solaris.

2002-02-15  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-02-01  ----(3.0.0)------------

	* HepMC/GenEvent.h: bug discovered by <srinir@bnl.gov> with help
	from Paolo and Hong.
	The GenEvent::particle_iterator, const_particle_iterator,
	              vertex_iterator, const_vertex_iterator
	had methods like:
	  bool operator !=(const particle_iterator& a) const
	  {  return !(**this == *a); }
	which doesn't work because you are not allowed to de-reference
	the end() iterator  [the above coding is a relic of the old
	              walking iterators which did not inherit from
	              forward_iterstor, and for which *end() was well
	              defined.]
	The correct usage is:
	{ return m_map_iterator == a.m_map_iterator; }

	Note: this does NOT affect the GenVertex::***_iterator's, since
	for those iterators *end() is well defined.


2002-01-23  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-02-00  --------------------
	                  ( Public Version 1.2 )

	User manual, webpage, and public version updated.

	The CLHEP names branch, called HepMC-00-00-96-CLHEP, has
	been moved to the head of the main trunk in cvs, and commited.
	Thus the main trunk is the only development branch.


2002-01-22  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-01-12  --------------------

	* src/GenEvent.cxx (HepMC): delete_all_vertices()
	modified the increment of the vertex iterator when looping over
	vertices for deletion.
	* src/GenVertex.cxx (HepMC): delete_adopted_particles()
	modified the increment of the particle iterator when looping over
	particles for deletion.

2002-01-21  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-01-07  --------------------

	* HepMC/GenEvent.cxx (delete_all_vertices):
	Possible bug reported here by ATLfast group. Added extra error
	checking, and changed the deleting of the vertices such that each
	one is explicitly erased from the map.


2002-01-18  Matt Dobbs  <Matt.Dobbs@Cern.CH>

  More changes for Windoes MS Visual C++ compatibility:

	* HepMC/PythiaWrapper6_2.h, HepMC/PythiaWrapper6_152.h:
	included new wrapper methods like call_pyinit, which hide the
	funny syntax necessary to call fortran routines from C++.
	* examples/example_MyPythiaWithEventSelection.cxx:
	* examples/example_MyPythiaOnlyToHepMC.cxx:
	* examples/example_MyPythia.cxx:
	* examples/Benchmarks_f77/example_PythiaToLCWrite.cxx:
	modified to use the call_*** pythia methods.


2002-01-15  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-01-06  --------------------

	* Makefile.standalone: updated to use cernlib 2001 and pythia6152

	* examples/example_UsingIterators.cxx:
	The input file was from the old non-barcodes version, and so
	didn't work. Replaced it with a proper input file.


These changes suggested by Witold Pokorski <Witold.Pokorski@Cern.Ch> from
LHCb to allow for Windows Visual C++ compliance. All of these changes are
fully backwards compatible(!).

	* HepMC/GenEvent.h, HepMC/GenVertex.h: change forward_iterator to
	simply iterator (which in the STL standard apparently implies
	forward_iterator)

	In Visual c++, a for (int i ... ) {} statement does not scope out
	the i variable ... so had to modify the code to ensure no
	variables were used twice in the same method:
	* src/IO_Ascii.cxx (HepMC):
	  in lines 155, 161, 175 "int i" is repeated
	  in lines 293, 314, 321 "int i" is repeated
          in lines 293, 314, 321 "p" is repeated
	* src/IO_HEPEVT.cxx (HepMC):
	  in lines 59,65,88 "i" is repeated
          in lines 118, 129 "p" is repeated
	* src/IO_PDG_ParticleDataTable.cxx (HepMC):
	  in lines 80, 103, 124 "id_i" is repeated
	* src/GenVertex.cxx (HepMC):
	  in lines: 64 and 72; 168 and 178; 195 and 201; 260 and 273;
	            322 and 331   variable "part" is repeated

	* have verified all STL calls for set, map, less, greater, cout,
	endl, list, cerr, ios, iostream, flush are properly prefixed with
	std::

	* HepMC/GenEvent.h: bug fix:
	added a std::greater<int> qualifier to the m_map_iterator of the
	vertex iterators.

	* HepMC/GenVertex.h: bug fix:
	made the GenVertex::particle_iterator a friend of GenVertex.

	* HepMC/GenEvent.h, HepMC/GenVertex.h:
	move all of the friend class declarations
	into statements separate from the class declaration, i.e.
	  class edge_iterator { ... };
	  friend class edge_iterator;
	instead of
	  friend class edge_iterator{ ... };
	Change all   std::ptrdiff_t   to  ptrdiff_t

	* HepMC/IO_Ascii.h: ios::openmode changed to ios::open_mode in the
	constructor, for compatibility with windows Visual C++.

	* Change #include <iostream.h> to #include <iostream>
	Note: this will spoil compatibility with HPUX 10.2 CC.


2001-11-29  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-01-05  --------------------
	* Fix typo in user manual "particle barcodes are positive"


2001-11  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-01-04  --------------------
	* update user manual to explain barcodes. NO changes to code.


2001-11-10  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-01-03  --------------------
	* cmt/requirements: jetset74 requirement removed, as requested by
	I.Hinchliffe.

2001-11-04  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	--------------------  HepMC-01-01-02  --------------------
	* HepMC/GenVertex.h: removed merge_vertex() method.

	* HepMC/IO_HEPEVT.h:
	removed the merge_vertices possibility, but added a
	build_production_vertex possibility to build_end_vertex.

	--------------------  HepMC-01-01-01  --------------------
	* src/HEPEVT_Wrapper.cxx (HepMC):
	new method check_hepevt_consistency() added.
	new method zero_everything() added.

2001-11-02  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	--------------------  HepMC-01-01-00  --------------------
	* src/IO_HEPEVT.cxx (HepMC):
	Modified such that the HEPEVT indices are used for the particle
	barcodes.
	* HepMC/IO_Ascii.h:
	Modifed to properly read/write the barcodes and the
	GenEvent new data members.

        Unique reference numbers called "barcodes" are added as data
	members to GenParticles and GenVertices.
	They are cross-referenced in GenEvent via a map<int,GenXXX>.
	The GenEvent::m_vertices container of vertices is then no longer
	necessary, so it is removed.
	The GenEvent::particle_iterator and GenEvent::vertex_iterator
	are updated to walk over the maps. The iterator change should be
	transparent to the user.
	Constant versions of both iterators are included.
	The old GenEvent::vertex_iterator was a typedef of the set:
	  typedef std::set<GenVertex*>::iterator vertex_iterator;

	The maps get filled via the:
		set_barcode( GenXXX* x, int suggested_barcode )
	methods, which in turn are called only by the vertex and particle:
	GenVertex::set_parent_event_()
	GenPartcle::set_production_vertex_()
	GenPartcle::set_end_vertex_(}

	* HepMC/GenEvent.h:
	New data members and access methods are added for consistency with
	the HepUP standard, see hep-ph/0109068
	double GenEvent::m_event_scale
	double GenEvent::m_alphaQCD
	double GenEvent::m_alphaQED

	* src/IO_HEPEVT.cxx (HepMC): Fixed some const warnings with the
	 GenEvent::vertex_iterator usage.

2001-10-11  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	--------------------  HepMC-01-00-01  --------------------
	* cmt/requirements: changed use CERNLIB v2001 External  to  use
	CERNLIB CERNLIB-01-* External   as per r.d.'s request

2001-10-10  Matt Dobbs  <Matt.Dobbs@Cern.CH>

        --------------------  HepMC-01-00-00  --------------------
	Identical to HepMC-00-03-08.  This tag name follows atlas
	naming conventions ... increment the major tag when code is NOT
        backwards compatible. Since Atlas is moving from tag
	HepMC-00-01-12 (old atlas names) to HepMC-00-03-08 (clhep names)
	this change is not backwards compatible.

2001-09-05  Matt Dobbs  <Matt.Dobbs@Cern.CH>
        ------- below this line is tag HepMC-00-03-08 ----------
	(clhep names version to be used in atlas. The changes have not
	been made in the HepMC-00-01-** branch)

	* src/IO_HEPEVT.h (HepMC):
	* src/IO_HEPEVT.cxx (HepMC): When inconsistent info between
	mother and daughters is found in the
	IO_HEPEVT::build_production_vertex method, IO_HEPEVT will no
	longer merge_vertex (the 2001-03-29 "fix") by default. In order to
	get this behaviour you have to explicitly ask for it with the
	IO_HEPEVT::set_merge_vertex_switch() method.

2001-09-05  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	* cmt/requirements: added to the CLHEP names branch (its just a
	direct copy of what is in the trunk ATLAS names).


2001-07-17  Matt Dobbs  <Matt.Dobbs@Cern.CH>
	(note: HepMC-00-01-24 is an accidental tag which should be ignored)
	----- below this line is  HepMC-00-01-23, HepMC-00-03-07 -----

	* src/GenVertex.cxx (HepMC):
	GenVertex::vertex_iterator::follow_edge_()  added extra error
	checking for the very special-rare case where a particle might
	point to the same vertex for both production and end.
	Meant to handle Ian H.'s bug of 02 Jul 2001

2001-06-28  Matt Dobbs  <Matt.Dobbs@Cern.CH>
        --------------------  HepMC-00-03-06  --------------------

	* examples/GNUmakefile.in:
	updated the makefile to link to  cernlib 2001 version of pythia.
	This involves a change -lpythia6136 to -lpythia6152 -lpythiad
	Note that for xample_MyPythiaWithEventSelection.cxx 1/100 or
	36/1000 events pass cuts (with 6.136 it was 3/100 or 37/1000).

	* HepMC/GenParticle.h: now has a parent_event() access method,
	which returns the parent_event of the particle's container
	vertex's event.

2001-06-27  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	---> Make the Vertex know which event it in in. <---
	* HepMC/GenVertex.h: added private data member m_event, public
	access method parent_event() and protected access method
	set_parent_event, which is only to be used by GenEvent. Thus
	GenEvent is made to be a friend of GenVertex.
	* HepMC/GenEvent.h, src/GenEvent.cxx (HepMC): added logic to
	GenEvent::remove_vertex and GenEvent::add_vertex. The vtx's
	m_event pointer is set to point back to the event.

	* HepMC/ParticleData.h:
	* HepMC/GenEvent.h:
	* HepMC/GenParticle.h:
	* HepMC/GenVertex.h: made method counter() protected. It was never
	intended as public, merely as a method for detecting memory
	leaks. If no-one complains (they shouldn't, it was commented
	"temporary for debugging") I'll remove it entirely later.

2001-03-29  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* src/IO_HEPEVT.cxx (HepMC):
	When m_trust_mothers_before_daughters=0, the vertex position is
	now filled. In ISAJET sometimes disjoint vertex structures exist,
	in this case a merger of vertices is done, which accounts for it.

	* HepMC/GenVertex.h:
	* src/GenVertex.cxx (HepMC):
	Added a            void merge_vertex_contents( GenVertex* v_in );
	method which takes all the contents of v_in and places it in
	the vertex. It does not delete v_in, nor does it remove
	v_in from an event.
	It will mainly be used by IO_HEPEVT, but it is forseen that users
	may want to use it when reducing the overall size of the event
	(deleting intermediate vertices) such as will be done by atlas.

2001-02-28  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* Makefile.standalone:
	using
	  HepMCdir = $(shell pwd)
	instead of
	  HepMCdir = /afs/cern.ch/user/m/mdobbs/HepMC/myCheckoutV1.01
	for better portability.

	* src/GenVertex.cxx (HepMC): added extra protection to the
	GenVertex::edge_iterator::edge_iterator constructor to give
	sensible result for the special case when a vertex has incoming
	particles, but none outgoing and the user requests a children
	iterator.  The old version would have
	m_set_iter points to m_particles_in.begin() and so the first
	dereference might be wrong.

	* HepMC/IO_HEPEVT.h: Two new switches are added to give more
	flexibility in handling the possible inconsistencies arrising
	from the bi-directional pointers in HEPEVT. The switches are:
		bool m_trust_mothers_before_daughters;
		bool m_print_inconsistency_errors;
	which are described by comments in the code. Also makes minor
	changes to src/IO_HEPEVT.cxx

	* HepMC/Typedef_Version0names.h:
	This header file added. It contains typedef statements which
	should allow the new version 1.XX class names to be used with code
	written with the old version 0.XX class names.
	If the user wants to take advantage of these typedefs he needs to
	define type variable HEPMC_SHORT_NAMES in his compiler options as
	follows: -DHEPMC_SHORT_NAMES
	or in his code using #define HEPMC_SHORT_NAMES
	Note that since the header file names also changed, this ALSO
	requires changing the include statements in the user code.

	* doc/latex_user_manual/physicist_visualization.eps
	fixed a mis-labelling in this figure... the labels were all
	shifted and intermingled... its now right. Unfortunately
	the mislabelled version appears in the CPC publication.
	also affects doc/latex_user_manual/HepMC_user_manual.ps

******************************************************
Everything below this line corresponds to Version 1.01
******************************************************

2001-01-11  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* examples/example_UsingIterators.cxx:
	Changed the class IsFinalState such that a particle must have no
	end vertex AND have status==1 to be considered stable.
	Does not effect package, only this specific example.

	* HepMC/IO_BaseClass.h:
	Bug reported by Lynn Garren
	"There appears to be a problem with const-ness in HepMC/IO_BaseClass.h.
	I have made the following changes in order to compile HepMC 1.0
	with g++.
	virtual const GenEvent*& operator<<( GenEvent*& );
		becomes
	virtual       GenEvent*& operator<<( GenEvent*& );

	virtual const ParticleDataTable*& operator<<( ParticleDataTable*& )
	        becomes
	virtual       ParticleDataTable*& operator<<( ParticleDataTable*& )
	In other words, I have made the const-ness match."
	...
	These changes are implemented by Matt in the atlas repository.

	* HepMC/HEPEVT_Wrapper.h:
	Bug reported by Andreas Dell'Acqua
		HEPEVT_Wrapper::last_child
		HEPEVT_Wrapper::last_parent
	There methods were returning 0 when exactly one parent exists. This
	caused the number_children and number_parent methods to return 0
	every time there was exactly 1 parent/child. This error caused
	major problems in the IO_HEPEVT class, with the end result that
	the HEPEVT graphs were interpretted incorrectly [all particles are
	correctly interpretted, but their relationships are incorrect for
	the case where a particle has exactly ONE mother: in this case the
	particles appeared as orphans.]
	This bug was
	introduced in version HepMC-00-00-94, so it has been present for
	some time. This fix is for HEPEVT_Wrapper::last_child():
	// Returns the Index of the LAST child in the HEPEVT record
	// for particle with Index index.
	// If there is only one child, the last child is forced to
	// be the same as the first child.
	// If there are no children for this particle, both the first_child
	// and the last_child with return 0.
	// Error checking is done to ensure the child is always
	// within range ( 0 <= parent <= nhep )

	* src/IO_HEPEVT.cxx (HepMC):
	only comments in IO_HEPEVT::build_particle have changed. No change
	to code.

******************************************************
Everything below this line corresponds to Version 1.0
******************************************************

2000-11-10  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	MAJOR NAME CHANGES AFFECT ALMOST ALL FILES:
	GeneratorEvent --> GenEvent
	Particle       --> GenParticle
	Vertex         --> GenVertex

	user manual and online documentaion are updated to reflect this.

	* HepMC/PythiaWrapper6_152.h:
	A new wrapper for Pythia 6.1 is created to replace the old Pythia
	5.7 wrapper. It works with the versions of Pythia in both
	CERNlib2000 and CERNlib2001 (versions 6.136 and 6.152 and probably
	any Pythia 6).
	A pointer file:
	* HepMC/PythiaWrapper.h:
	is created so the used doesn't have to worry about pythia version
	numbers. A test file is located at:
	* test/test_PythiaWrapper.cxx

	All examples are updated to use Pythia 6.


2000-11-09  Matt Dobbs  <Matt.Dobbs@Cern.CH>

	* HepMC/WeightContainer.h:
	Created this class to house the weights in the vertex and event
	classes. It is just an interface to std::vertex<double>.

	The vertex class is modifid to use this. Required changes to:
	* HepMC/Vertex.h:
	* src/Vertex.cxx:
	* src/IO_Ascii.cxx (HepMC):

******************************************************
Everything below this line corresponds to Version 0.94
******************************************************


2000-07-05  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* HepMC/HEPEVT_Wrapper.h: Completely new wrapper for HEPEVT_Common
	is implemented. This wrapper treats the common as a series of
	bytes and interprets them "on the fly". The user can set the size
	(in bytes) of the integer and real numbers to be interpretted from
	HEPEVT, and also the total number of entries in HEPEVT. This
	allows the user to interface to two applications which use
	different HEPEVT definitions.
	Benchmarks: 1000 events pythia only 34 seconds.
	            1000 events Pythia+HepMC(using this new wrapper) 40s
        on my particular machine. (extra time is 6 seconds.) Compare this
	to the extra 5 seconds reported in the HepMC user manual for the
	old HEPEVT wrapper ... so the difference is small compared to the
	old wrapper considering the added functionality.
	(all examples have been tested and they Run!)

	* HepMC/HEPEVT_Common.h: Deleted, this common is now defined
	inside HEPEVT_Wrapper.h

	* src/HEPEVT_Wrapper.cxx (HepMC): This file is created since it is
	necessary to instantiate the static members of HEPEVT_Wrapper in a
	.cxx file.

	* Makefile.Standalone: modified to build src/HEPEVT_Wrapper.cxx into
	the HepMC library

	* HepMC/IO_HEPEVT.h: Comment modified only so as to treat
	HEPEVT_Wrapper consistently.

	* examples/example_MyPythiaWithEventSelection.cxx: Modified to use new
	HEPEVT_Wrapper

	* examples/example_MyPythia.cxx: Modified to use new
	HEPEVT_Wrapper

	* examples/example_MyPythiaOnlyToHepMC.cxx (main): Modified to use new
	HEPEVT_Wrapper

	* PACKAGE: added linkset HepMC -lib:HepMC
	(Modified GNUmakefile.in to use libHepMC.so as HepMC linkset)

2000-07-04  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* examples/GNUmakefile.in: $(top_srcdir) replaced by
	$(srcdir)

	* GNUmakefile.in (libHepMC.so_SRC): The location of src .cxx files
	is now explicitly specified with $(top_srcdir)/src/*.cxx
	This makefile now builds oth a shared library libHepMC.so and
	libHepMC.a (requested by Ian H.)
	* the initpydata.f routine is no longer compiled into the the
	HepMC library, but is still packaged with the other routines.
	The user needs to link to it explicitly. See the GNUmakefile.in
	in the examples directory for an example of how to do this.
	(requested by Ian H.)
	* ./src is replaced with $(top_srcdir)/src everywhere in the
	GNUmakefile.in

	Files changed this time:
	GNUmakefile.in, examples/GNUmakefile.in, Makefile.standalone,
	Changelog

2000-05-18  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* HepMC/IO_BaseClass.h: There is now a fill_next_event method.
	- bool fill_next_event( GeneratorEvent* evt ) is a new abstract method
	  which fills the passed event, and also returns true if successful.
	- GeneratorEvent* read_next_event() is now make concrete. it
	  creates a new event, fills it using the above abstract method,
	  and returns the event
	This change is backwards compatible since the method
	read_next_event() has identical behavior.
	Identical changes are made for the methods:
	- bool fill_particle_data_table( ParticleDataTable* pdt )
	- ParticleDataTable* read_particle_data_table()
	The methods are updated [ i.e. GeneratorEvent* read_next_event()
	is changed to bool fill_next_event( GeneratorEvent* evt ), etc. ]
	in:
	- IO_HEPEVT.h
	- IO_Ascii.h
	- IO_PDG_ParticleDataTable.h
	- IO_HEPEVT.cxx
	- IO_Ascii.cxx
	- IO_PDG_ParticleDataTable.cxx
	(This change requested by Ian Hinchliffe and Marjorie Shapiro.)


2000-04-24  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* HepMC/HEPEVT_Wrapper.h: set_momentum and set_position now
	properly cast input doubles as type HEPEVT_Precision,
	previously they were hardwired as cast to double, this
	meant writing to real*4 hepevt would fail.

******************************************************
Everything below this line corresponds to Version 0.91
******************************************************

2000-04-11  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* doc/latex_user_manual/HepMC_user_manual.tex: User manual is
	updated to replect changes since V0.9
	- HEPEVT and position in [mm]

2000-04-06  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* examples/example_MyPythia.cxx
	* examples/example_MyPythiaOnlyToHepMC.cxx
	* examples/example_MyPythiaWithEventSelection.cxx:
	By commenting out the initpydata call (which may be required on
	some systems to initialize the Pythia PYDATA block data as
	external) the MyPythia examples now run on HPUX (and still on
	Linux). Have not tested other platforms.
	* Makefile: HPUX with gcc requires the -lf library which contains
	the fortran intrinsic function.

2000-04-05  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	The changes below will have no effect on the user or package, they
	just serve to remove warnings when compiling.

	* HepMC/Vertex.h: 313:342:364: warning: unused parameter `enum
	::HepMC::IteratorRange dummy_range' (all dummy range parameters
	now appear as particles_end( IteratorRange /* dummy_range */ ) )

	* src/IO_Ascii.cxx (HepMC): 506: warning: ANSI C++ forbids
	variable-size array `c' (the array is now declared as char* c =
	new char[key_length +1]; and deleted before returning)

	* HepMC/IO_PDG_ParticleDataTable.h: 53: warning: control reaches
	end of non-void function read_next_event()
	(this is a dummy function anyway -- return 0)
	* HepMC/IO_HEPEVT.h:55: (as above)

	* src/Flow.cxx (HepMC): 26: warning: default argument given for
	parameter 1 of `void ::HepMC::Flow::print(class ostream & = cout)
	const' (done)

	* HepMC/IO_HEPEVT.h:57: warning: base class `class
	::HepMC::IO_BaseClass' should be explicitly initialized in the
	copy constructor (done)
	* HepMC/IO_Ascii.h: 98: (as above)
	* HepMC/IO_PDG_ParticleDataTable.h: 56: (as above)

	* HepMC/ParticleDataTable.h: 155: warning: suggest parentheses
	around assignment used as truth value (done)

	* src/IO_Ascii.cxx (HepMC): 37: warning: default argument given
	(removed default argument)

	* HepMC/Flow.h:21: warning: multiline `//' comment
	(added space after \ to remove compiler warning)

	* HepMC/Particle.h: removed =0 in the inline declaration
	inline void Particle::set_flow( int code_index, int code = 0 )
	(caused warning).


2000-04-04  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* HepMC/HEPEVT_Wrapper.h:  HepMC/HEPEVT_Common.h:
	replace the classes HEPEVT_BaseClass.h HEPEVT_Double2000.h
	HEPEVT_Double4000.h HEPEVT_Real2000.h HEPEVT_Real4000.h

	Using different classes for each of the HEPEVT common block
	varieties proved very troublesome - users had trouble and 4
	similar classes were being maintained. The new version uses only
	static methods and cannot be instantiated (logical since it has
	no data members). The HEPEVT common block is contained in the
	HepMC/HEPEVT_Common.h header. The precision and number of entries
	is specified using #define commands, default is double precision,
	2000 entries. The user will see only one change - the IO_HEPEVT
	constructor now has no arguments, so he need not define the
	any HEPEVT_XXX class. He may need to set the precision and number
	of entries. See the header for instructions, and
	examples/example_MyPythia.cxx for an example.

	example_MyPythia.cxx, example_MyPythiaOnlyToHepMC.cxx,
	example_MyPythiaWithEventSelection.cxx are modified to take this
	into account.

	* HepMC/IO_HEPEVT.h: This class now extracts information from the
	fortran HEPEVT common using HEPEVT_Wrapper instead of having to
	pass it a class which inherits from the obsolete HEPEVT_BaseClass.

2000-04-03  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* HepMC/PythiaWrapper5_720.h: Credit for writing the wrapper is
	to Silvia Resconi.
	Inclusion of HEPEVT wrapper header file is not necessary --- hence
	removed.

2000-03-31  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* HepMC/ParticleData.h: clifetime is now represented in [mm]
	instead of [cm] ... previously HepMC had internal inconsistencies
	in this --- it claimed to use [cm], but sometimes used [mm] (as
	when it read from HEPEVT, and othertimes used [cm] (as when it
	transformed a width into a lifetime. Now all is consitent and [mm]
	is used everywhere as it should be.
	* HepMC/Vertex.h: src/ParticleData.cxx:
	Changed comments which say clifetime is in [cm], to say
	clifetime is in [mm].

	* Makefile (CLHEPdir):
	Changed
	CLHEPdir      = /afs/cern.ch/sw/lhcxx/specific/Linux/CLHEP/pro
	to
	CLHEPdir      = /afs/cern.ch/sw/lhcxx/specific/@sys/CLHEP/dev
	to make it platform independent for CERN users.
	And some cosmetic changes to the Makefile.

	* HepMC/IO_BaseClass.h:
	Using the compiler options -ansi -pedantic cause this error:
	no match for `::HepMC::IO_Ascii & << ::HepMC::GeneratorEvent *&'
	Even though it recognizes:
	operator <<(::HepMC::IO_BaseClass &,
	            const ::HepMC::GeneratorEvent *&) <near match>
	as the "best <near match>".

	To fix this I make operator<<, operator>> members of the
	IO_BaseClass rather than friends, and I add versions where the
	argument of operator<< is not constant.
	This will not affect users in any way (unless they were using
	these compiler options, in which case HepMC didn't work anyway...)

2000-02-16  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>

	* GeneratorEvent.h: set_signal_process_vertex() modified to ensure
	that any vertex specified as signal process is ALSO in the
	m_vertices set.

	* This change log documents changes since Version 9 of HepMC was
	"released" February 11, 2000

pre 2000-02-16  Matt Dobbs  <Matthew.Adam.Dobbs@Cern.CH>
Changes Before Version 0.9 (i.e. changes in going from V0.1 -->0.9)
	    HepMC improvements:
	        +backup V0.1  (done)
	        +save particle data as id in Particle  (done)
		+update IO strategies  (done)
		+test  (done)
		+backup  (done)
	     -change set<Vertex*> in event to include all vertices (done)
		+change iterators to reflect this (done)
		+update IO strategies (done)
		+test (done)
	     -fix all operator= to first destruct dependents. (done)
	     -access functions to iterate over all flows  (done)
	     -check for stl namespace  (done)
	      grep -n cout HepMC/*.h src/*.cxx | grep -v std::cout |
	        less
             -add Hepevt:write event  (done)
	     -implement lujet strategy (not done!)
	     -update documentation -- enumerating recent changes.  (done)
	     -user manual written