ChangeLog

FlexibleSUSY 2.0.0 [not released yet]

 * Feature: BSM contributions to the anomalous magnetic moment of the
   muon, a_mu, at the 1L level in any given BSM model.  Note: Diagrams
   with non-SM vector bosons are not taken into account.

   In order to let FlexibleSUSY calculate a_mu, the symbol
   FlexibleSUSYObservable`aMuon must be written into an SLHA output
   block in the ExtraSLHAOutputBlocks variable in the FlexibleSUSY
   model file.

   Example:

   ExtraSLHAOutputBlocks = {
      {FlexibleSUSYLowEnergy,
         {{21, FlexibleSUSYObservable`aMuon}}}
   };

   Thanks to Jobst Ziebell.

FlexibleSUSY-1.7.1 [not released yet]

 * Change (commit b1efa8c): Updated to GM2Calc 1.3.0.

 * Change (commit 05d8e11): The loop order of the BSM top Yukawa
   coupling at the scale M_SUSY in FlexibleEFTHiggs is now set
   automatically to match the loop order of the matching condition
   from the SM to the BSM model.

   Before this commit, the user had to set FlexibleEFTHiggs[13] = 0
   and FlexibleEFTHiggs[20] = 1 when yt(BSM) should be calculated
   using 1L QCD corrections.  Analogous, the user had to set
   FlexibleEFTHiggs[13] = 1 and FlexibleEFTHiggs[20] = 2 when yt(BSM)
   should be calculated using 2L QCD corrections.  Now,
   FlexibleEFTHiggs[13] is set automatically to FlexibleEFTHiggs[20] -
   1 when yt(BSM) is calculated in FlexibleEFTHiggs.

FlexibleSUSY-1.7.0 [September, 19 2016]

 * Feature: FlexibleSUSY is now able to generate custom spectrum
   generators using the FlexibleEFTHiggs method described in
   [arXiv:1609.00371].  The following FlexibleEFTHiggs example models
   are provided: CMSSMtower, MSSMtower, MSSMNoFVtower, NMSSMtower,
   NUHMSSMalttower, MRSSMtower, E6SSMtower.  A documentation of the
   new model file options to create a custom FlexibleEFTHiggs spectrum
   generator can be found in doc/html/FlexibleEFTHiggs.html .

 * Feature: FlexibleSUSY now provides a Mathematica interface for the
   generated spectrum generators.  For each model, an example
   Mathematica script

      models/<model>/run_<model>.m

   is generated, which illustrates the usage.  The documentation of
   the Mathematica interface and several examples can be found in
   FlexibleSUSY's HTML documentation.  Please see the section
   "Creating the soucre code documentation" in the README file for a
   description about how to generate the documentation.

 * Change: The configure options for creating dynamic libraries and
   statically linked executable have been changed.  By default, static
   FlexibleSUSY libraries and dynamically linked executables are
   created.

   To generate shared FlexibleSUSY libraries, run

      $ ./configure --enable-shared-libs ...

   To generate statically linked executables, run

      $ ./configure --enable-static ...

   Please refer to the README file for more information.

 * Bugfix (commit 39f8d36): Fix segfault when multi-threading is used
   in statically linked executables.

 * Bugfix (commit 3126ac1): Catch NaNs from inside the MSSM 2L Higgs
   mass routines of Pietro Slavich.

 * Bugfix (commit b6db614): Correcting 2-loop self energy O(alpha_t^2)
   in the Standard Model.  Before, Eq. (20) of arxiv:1205.6497 has
   been used.  However, this is incorrect, because it includes 2-loop
   contributions from the momentum iteration of the 1-loop self
   energy, which would be double counted, because FlexibleSUSY already
   does a momentum iteration of the 1-loop self energy.  To fix this,
   Eq. (20) of arxiv:1504.05200 has been used, which does not include
   these 2-loop pieces.

FlexibleSUSY-1.6.1 [August, 28 2016]

 * Bugfix (commit db67c81): Fix compilation with --disable-threads .

FlexibleSUSY-1.6.0 [August, 27 2016]

 * Feature (commit 4e9ef56): Allow user to access the beta-functions
   of the model parameters on the r.h.s. of the constraints.  BETA[p]
   represents the beta function of the parameter p using the loop
   level given in the SLHA input.  BETA[l,p] represents the l-loop
   beta function of the parameter p.

   Example in the SM:

   HighScaleInput = {
       {\[Lambda], BETA[g1] + BETA[g2] + BETA[1,Yu][3,3]}
   };

 * Feature (commit 5e0bca1): Allow user to add 3-loop QCD corrections
   of [arxiv:hep-ph/9912391] when calculating the top pole mass in
   non-SUSY models.  The 3-loop QCD corrections are added if the flag
   FlexibleSUSY[13] is set to 2 and the pole mass loop order,
   FlexibleSUSY[4], is set to a value > 2.

   FlexibleSUSY[13] = 0 and FlexibleSUSY[4] > 0: 1L QCD correction
   FlexibleSUSY[13] = 1 and FlexibleSUSY[4] > 1: 2L QCD correction
   FlexibleSUSY[13] = 2 and FlexibleSUSY[4] > 2: 3L QCD correction

 * Feature (commits 98bc536, e8fd56a): Speed up of the RG running in
   models with very complicated beta functions.

 * Change (commit 728b5ea): `make clean' no longer removes generated
   source files to avoid the need to re-generate them.  To remove the
   generated files use either

   $ make clean-<model>-src # deletes generated files for <model>

   or

   $ make clean-generated   # deletes all generated files

 * Bugfix (commit a5342eb): Avoid non-portable use of sed in
   createmodel.  This fixes `make install-src' on Mac.

 * Bugfix (commit 44b31fa): Fix potential race condition when
   different model classes that make use of the (N)MSSM 2-loop Higgs
   mass routines of P. Slavich call calculate_spectrum() at the same
   time.

 * Bugfix (commit 0d08b99): Do not try to generate non-squared unit
   matrices for beta function expressions that must be splitted.
   Non-squared unit matrices did appear for non-squared matrix-valued
   parameters, as for example T[hE] in the SE6SSM.

   Thanks to Dylan Harries.

FlexibleSUSY-1.5.1 [July, 12 2016]

 * Bugfix (commit 63f5361): Fix numerical instability of SOFTSUSY's B1
   function in parameter regions with p << m1,m2 and m1 close to m2.

 * Bugfix (commit fc6d509): Fix makefile bug in the tarball by
   shipping all .m files that appear in the list of dependencies for
   the generated C++ code.

FlexibleSUSY-1.5.0 [June, 29 2016]

 * Feature: Write phases to SLHA output if a SLHA output block is
   defined for them in the SARAH model file.
   Thanks to Dylan Harries.

 * Feature: Allow the user to calculate the pole masses at a fixed
   renormalisation scale at run-time, which is different from the one
   set by the SUSYScale model file variable.  The fixed
   renormalisation scale can be given via the FlexibleSUSY[17] entry
   in the SLHA input.  FlexibleSUSY[17] is equivalent to
   SPhenoInput[33] in SPheno.

 * Feature: Updated to GM2Calc 1.2.0.

 * Bugfix (commit 9a2d576): Fix compilation error due to ambiguous
   overload of operator<< .
   Thanks to Dylan Harries.

 * Bugfix (commits fc748be, 9654a52): Fix compilation in case Greek
   Symbols appear in If[] or Which[] functions in the model file.
   Thanks to Dylan Harries.

 * Bugfix: Fix compilation with g++ 4.4.7.
   Thanks to Dylan Harries.

 * Bugfix (commit 6f5e38e): Correcting convergence criterion in the
   iteration which determines the 1st and 2nd generation running
   fermion masses in the SM(5) at the low-energy scale.  After this
   correction, the running 1st and 2nd generation SM(5) fermion masses
   differ from SOFTSUSY by less than 0.5% at the electroweak scale.

FlexibleSUSY-1.4.2 [May, 09 2016]

 * Bugfix: Correcting handling of spaces in configure script if
   /bin/sh is /bin/dash.

FlexibleSUSY-1.4.1 [May, 09 2016]

 * Feature: Tab-completion for FlexibleSUSY's spectrum generators and
   scripts in the bash.

   Usage:
   . utils/install-bash_completions.bash

 * Feature: For each model an example SLHA input file is generated,
   which is located at models/<model>/LesHouches.in.<model>_generated

 * Feature (commit 2b95522): Allow user to provide specific location
   to libpthread using the --with-pthread-libdir= option.

 * Change: The algorithm to determine the running fermion masses and
   gauge couplings has been replaced by a more secure one.  The new
   algorithm performs an iteration between 2 GeV and MZ to fix all
   input parameters at their scale.  The new algorithm leads to
   running 1st and 2nd generation quark masses, which differ from
   SOFTSUSY by around 3%.

 * Bugfix (commit 59b867d): Avoid singularity in the limit MSU^2 /
   M3^2 -> MSQ^2 / M3^2 in HSSUSY.

 * Bugfix (commit f3864b8): Catch exception from SOFTSUSY's QedQcd
   class which are triggered when the input value of Mt_pole is chosen
   to be smaller than MZ_pole.

 * Bugfix (commit 077c5b9): Fixing check for SARAH installation with
   Mathematica 10.

 * Bugfix (commit e9954d6): Fixing numerical instability of SOFTSUSY's
   B0 and B22 functions for very heavy spectra and external small
   momenta.

 * Bugfix (commits bcb99bc - 8b5d87e): Fixing compilation error for
   models which don't have input parameters.

 * Bugfix (commits 637d099, 8b3a94f, 2e3a972): Fixing `make
   install-src' in case the path to the FlexibleSUSY contains spaces.

 * Bugfix (commits ced2072, 8bc8fdd): Adding support for further
   debian-based multi-architecture linux distributions in the
   configure script.

FlexibleSUSY-1.4.0 [March, 08 2016]

 * Feature: Allow the user to chose the loop order of the RGEs to be
   generated by SARAH.  This is useful in pure low-energy models,
   where no RGE running is needed, or in very complex models, where
   the generation of the RGEs takes a very long time.

   The RGE loop order can be set in the model file using the
   `FSRGELoopOrder' variable.

   Example:
   FSRGELoopOrder = 0; (* no RGEs generated *)
   FSRGELoopOrder = 1; (* only 1-loop RGEs generated *)
   FSRGELoopOrder = 2; (* 1- and 2-loop RGEs generated (default) *)

   Note: The RGE loop order can also be specified at run-time in the
   SLHA input block FlexibleSUSY[6] .

 * Feature: FlexibleSUSY no longer requires that the weak mixing angle
   and potential Z-Z' mixing angles are provided in terms of
   Lagrangian density parameters (gauge couplings etc.).  Instead,
   FlexibleSUSY makes use of the DependenceSPheno specification given
   in the SARAH model file to calculate these mixing angles
   numerically.  In this way the effect of gauge boson mixings in
   models with extended gauge groups can be taken into account
   automatically.

   Note: If the weak mixing angle is to be fixed at the low-energy
   scale by the running W and Z masses (see FSWeakMixingAngleOptions
   option) in order to determine the electroweak gauge couplings, then
   an expression for it has to be given in either DependenceNum or
   FSWeakMixingAngleExpr .

   Example for the MRSSM:

   (* determine weak mixing angle from W and Z masses *)
   FSWeakMixingAngleOptions = FSSetOption[
       FSWeakMixingAngleOptions,
       FSWeakMixingAngleInput -> FSMassW
   ];
   (* need to provide expression for weak mixing angle *)
   FSWeakMixingAngleOptions = FSSetOption[
       FSWeakMixingAngleOptions,
       FSWeakMixingAngleExpr  -> ArcSin[Sqrt[1 - (Mass[VWm]^2 - g2^2*vT^2)/Mass[VZ]^2]]
   ];

   Important note: In the SARAH model file a mass ordering of the
   vector bosons is assumed.  For example, the statement

   DEFINITION[EWSB][GaugeSector] = {
       {{VB,VWB[3],VBp}, {VP,VZ,VZp}, ZZ},
       ...
   };

   assumes MVP < MZ < MZp.  Thus, the user has to make sure that the
   studied parameter region leads to Photon, Z and Z' masses which are
   in agreement with the relation MVP < MZ < MZp.  Otherwise, the
   calculated Z and Z' masses will be incorrect.  If a parameter
   region shall be studied where MVP < MZp < MZ, then the ordering of
   vector bosons in the SARAH model file has to be changed to

   DEFINITION[EWSB][GaugeSector] = {
       {{VB,VWB[3],VBp}, {VP,VZp,VZ}, ZZ},
       ...
   };

 * Feature: By setting the entry FlexibleSUSY[16] = 1 in the SLHA
   input file, the user can force majorana fermion masses to be
   positive.  In this case, the corresponding mixing matrix is not
   purely real and its imaginary part will be written to the output in
   addition.  Note, that setting FlexibleSUSY[16] = 1 is therefore a
   violation of the SLHA standard.

 * Feature: FlexibleSUSY calculates the effective 1-loop couplings of
   the CP-even and CP-odd Higgs -> photon + photon and Higgs -> gluon
   + gluon.
   Author: Dylan Harries

   For each model the <model>_effective_couplings class is generated
   and can be used at the C++ level to calculate the effective
   couplings.  In order to write the effective couplings to the SLHA
   output, extra SLHA output blocks have to defined in the
   FlexibleSUSY model file, which contain the symbols

   FlexibleSUSYObservable`CpHiggsPhotonPhoton
   FlexibleSUSYObservable`CpHiggsGluonGluon
   FlexibleSUSYObservable`CpPseudoScalarPhotonPhoton
   FlexibleSUSYObservable`CpPseudoScalarGluonGluon

   Example:

   Definition of an extra SLHA output block named `EFFHIGGSCOUPLINGS',
   containing the effective 1-loop CP-even and CP-odd Higgs -> photon
   + photon and Higgs -> gluon + gluon couplings:

   ExtraSLHAOutputBlocks = {
      {EFFHIGGSCOUPLINGS,
              {{1, FlexibleSUSYObservable`CpHiggsPhotonPhoton},
               {2, FlexibleSUSYObservable`CpHiggsGluonGluon},
               {3, FlexibleSUSYObservable`CpPseudoScalarPhotonPhoton},
               {4, FlexibleSUSYObservable`CpPseudoScalarGluonGluon} } }
   };

   The calculation of the effective couplings can be disabled (or
   enabled) by setting the flag FlexibleSUSY[15] to 0 (or 1) in the
   SLHA input file.

 * Feature: Allow the user to temporarily re-define model parameters
   in the boundary conditions, which are restored to their original
   values after the calculations in the boundary condition has been
   finished.

   Example: Temporarily scale the gauge coupling g1 by a factor 1/2
   and set the up-quark Yukawa coupling to zero:

   LowScaleInput = {
      {Temporary[g1], g1 / 2},
      {Temporary[Yu[1,1]], 0},
      ...
   };

 * Feature: The three THDM-like models, which have been used in
   arxiv:1512.07761, are provided.  The models implement the 1- and
   2-loop threshold corrections of arxiv:1508.00576 and
   arxiv:hep-ph/9307201.  The models are named:
   * THDMIIMSSMBC (THDM with boundary condition to the MSSM)
   * HTHDMIIMSSMBC (THDM + Higgsinos with boundary condition to the
        MSSM)
   * HGTHDMIIMSSMBC (THDM + Higgsinos + gauginos with boundary
        condition to the MSSM)

 * Feature: In non-SUSY models the 3-loop (Standard Model) QCD
   corrections to the MS-bar Yukawa coupling of the order O(alpha_s^3)
   [arxiv:hep-ph/9911434, arxiv:hep-ph/9912391] are added
   automatically.  They are taken into account at run-time if the
   threshold correction loop (FlexibleSUSY[7]) order is set to a value
   > 2 in the SLHA input file.

   The generation of 3-loop QCD corrections can be disabled by setting
   in the model file

   UseYukawa3LoopQCD = False;

 * Change (commit f2f913e, 002c904): When threshold corrections are
   disabled, the charged lepton and top quark pole masses are used to
   determine the corresponding Yukawa couplings.  Before commit
   f2f913e, the running Standard Model masses were used.  This change
   makes it easier to compare the mass spectrum with SPheno when
   threshold corrections are disabled.

 * Change (commit 1c7e4a7): The 2-loop QCD contribution to the top
   Yukawa coupling [hep-ph/0210258 Eq. (60)-(61), hep-ph/9803493
   Eq. (17)] is taken into account only if the threshold correction
   loop order (flag FlexibleSUSY[7]) is set to a value > 1.  Before
   commit 1c7e4a7 the 2-loop QCD contribution was always taken into
   account and could not be disabled.  This change allows the user to
   consistently disable 2-loop contributions.

 * Bugfix (commit f7ff872): Support models which have couplings
   proportional to the epsilon tensor in color space.

 * Bugfix (commit 8c1ca39): Enabling support to use
   LowEnergyConstant[MZ] as scale for the susy-scale contraint.
   LowEnergyConstant[MZ] will be replaced in the C++ code by the
   user-defined SLHA input value of the Z pole mass.

   Example:
   SUSYScale = LowEnergyConstant[MZ];

 * Bugfix (commit 0a7934e): Fix compilation error in models in which a
   multiplet exists, which consists only of Goldstone bosons.

 * Bugfix (commit a87042f): Rename enum entries for matrices to
   prevent compilation errors in models which have mixing matrices
   larger than 10x10.

 * Bugfix (commit 61fb1ca): Fix compilation errors in models which
   don't contain SM-like neutrinos.

 * Bugfix (commit 919347d): Correcting the phase of Dirac fermion
   singlets if their mass is less than zero: Before commit 919347d,
   the phase of Dirac fermion singlets was set to e^(i Pi/2) if their
   mass is less than zero, which is wrong, because in SARAH only one
   Weyl component of the Dirac spinor receives a phase.  After this
   commit, the phase of Dirac fermion singlets is set to e^(i Pi) if
   their mass is less than zero.

 * Bugfix (commits 060b492, a6f7741, 306385b): Implement massless
   limits in C0, D0 and D27 functions.

 * Bugfix (commits d62886d): Ensure that only Standard Model goldstone
   bosons are removed to obtain "heavy" W and Z self-energies.

 * Bugfix (commits 60d68af): Fix compilation error in models where the
   left-handed electron and neutrino mass matrices are of equal size,
   but larger than 3x3.

FlexibleSUSY-1.3.2 [January, 10 2016]

 * Bugfix (commit d76ca79): Fix compilation error with Eigen
   3.3-beta1.

FlexibleSUSY-1.3.1 [January, 08 2016]

 * Bugfix (commit aa8dc76): Re-enable the output of gauge eigenstate
   masses of 1st and 2nd generation sfermions in the CMSSMNoFV for
   SLHA-1 compatibility.

FlexibleSUSY-1.3.0 [January, 08 2016]

 * Feature: The output of the spectrum generator can be written into
   an SQLite database using the --database-output-file= option.  At
   the C++ level, a to_database() and from_database() function is
   provided for each model, which write/read a model object (including
   the DR-bar parameters and the pole mass spectrum) to/from a
   database file.

   Example:
   $ models/CMSSM/run_CMSSM.x \
     --slha-input-file=model_files/CMSSM/LesHouches.in.CMSSM \
     --slha-output-file= --database-output-file=point.db

   Example using the scan script:
   $ utils/scan-database.sh \
     --spectrum-generator=models/CMSSM/run_CMSSM.x \
     --slha-input-file=model_files/CMSSM/LesHouches.in.CMSSM \
     --scan-range=MINPAR[3]=1~30:21 \
     --database-output-file=scan.db

 * Feature: Models can now be matched to the Standard Model at Q =
   MZ_pole, Q = MT_pole or any other dynamically calculated scale, as
   MT_DRbar for example.
   To match at MZ_pole set in the model file: LowScale = LowEnergyConstant[MZ]
   To match at MT_pole set in the model file: LowScale = LowEnergyConstant[MT]
   To match at MT_DRbar set in the model file: LowScale = M[Fu[3]],
   depending on your chosen name for the top quark.

 * Feature: 3-loop beta-functions can now be used in the real MSSM.
   To enable the 3-loop MSSM beta-functions, set UseMSSM3LoopRGEs =
   True; in the model file (enabled by default in all real MSSM models
   that are shipped with FlexibleSUSY).  The expressions have been
   obtained from http://www.liv.ac.uk/~dij/betas/allgennb.log and
   include family mixing.

   Note: The 3-loop beta-functions for the vacuum expectation values
   vu and vd are not available so far.  Furthermore, the 3-loop MSSM
   beta-functions miss the "tadpole" contributions corresponding to
   the renormalisation of the Fayet-Iliopoulos D-term, see the note in
   Section 2, page 4 of arxiv:hep-ph/0308231 .

 * Feature: The anomalous magnetic moment of the muon, (g-2)/2, can be
   calculated in all MSSM models without sfermion flavour violation
   (e.g. the MSSMNoFV and CMSSMNoFV).  The calculation is performed
   with GM2Calc 1.1.0 [arXiv:1510.08071] up to the 2-loop level
   including tan(beta) resummation.

   In order to enable the calculation of (g-2)/2, the symbols

      FlexibleSUSYObservable`aMuonGM2Calc
      FlexibleSUSYObservable`aMuonGM2CalcUncertainty

   have to be added to ExtraSLHAOutputBlocks variable in the
   FlexibleSUSY model file (they are already added in the MSSMNoFV and
   CMSSMNoFV example models).  In addition, the SLHA input file entry
   FlexibleSUSY[15] has to be set to 1 to perform the calculation.  If
   FlexibleSUSY[15] is set to 0, (g-2)/2 is not calculated.

 * Change (commit d553af8): No SLHA output is written if the option
   --slha-output-file= is set to the empty string.  To write the SLHA
   output to stdout, set --slha-output-file=- (this is the default).

 * Change (commit ac70fec): In the SM the Higgs pole mass is no longer
   calculated at the scale Qin (= the scale where lambda is input),
   but at the scale Q = M_top.

 * Bugfix (commit 1b4fc20): Correcting W contribution in beta-function
   of alpha_em in the SM with 5 active quark flavours.
   Imported from SOFTSUSY (commit 0139daa).

 * Bugfix (commit d7dbeb6): Adding neutrino charge, Qv, to list of
   input parameters in the UMSSM.  This fixes a compilation error with
   SARAH 4.6.0.

 * Bugfix (commit f1752a7): Correcting the trilinear couplings and the
   effective mu parameter in the NMSSMRUN SLHA output block in the
   models: NMSSM, NMSSMCPV, NUTNMSSM, SMSSM and NUTSMSSM.

 * Bugfix (commit 9ccdb4d): Workaround a SARAH issue where the list
   SARAH`Masses[EWSB] contains replacement rules of the form `0 ->
   MassGiven[X]', instead of `Mass[X] -> MassGiven[X]'.  Due to this
   issue some massless particles have been missing in FlexibleSUSY
   before commit 9ccdb4d.

FlexibleSUSY-1.2.4 [October, 27 2015]

 * Change (commit 33af37c): The spectrum generator, run_<model>.x,
   will no longer overwrite the user-given input parameters of the
   SMINPUTS block.

 * Bugfix (commit 9067f3a): There was an internal programming error in
   the meta code concerning the assignment of tadpole diagrams to the
   Higgs fields, which resulted in a compilation error in the SSM.
   Thanks to John McDowall.

 * Bugfix (commit 77d2a86): Ensure that in the calculation of the pole
   mass of a fermion singlet the prefactor of the self-energies is the
   positive tree-level mass.  Before commit ce1ef83, the prefactor of
   the gluino self-energies in MSSM for example was the soft-breaking
   parameter M3.  If M3 < 0 the gluino pole mass was not calculated
   correctly.
   Thanks to Dylan Harries and Roman Nevzorov.

FlexibleSUSY-1.2.3 [October, 18 2015]

 * Feature: Adding support for If[] and Which[] statements at the
   r.h.s. of contraints.  In addition, the IsClose[a,b,eps] and
   IsCloseRel[a,b,eps] functions have been added to allow for a
   comparison of parameters.

 * Feature: New model SplitMSSM, which implements low-energy EFT of
   the MSSM where the sfermions and one Higgs doublet have been
   integrated out.  The model implements the 1- and 2-loop matching
   conditions from arxiv:1407.4081.  The Higgs pole mass is calculated
   at complete 1-loop order plus 2-loop contributions O(alpha_t^2) and
   O(alpha_t alpha_s) from arxiv:1205.6497 plus 3-loop leading-log
   contribution from the gluino O(alpha_t alpha_s^2)
   (arxiv:1312.5220).

 * Feature: New model HSSUSY, which implements a high-scale SUSY
   scenario, where the sfermions, the gauginos, the Higgsinos and one
   Higgs doublet have been integrated out, leaving the Standard Model
   as low-energy EFT.  The model uses the 3-loop Standard Model RGEs
   (arxiv:1303.4364, arXiv:1307.3536) and implements the 1- and 2-loop
   matching conditions to lambda(MSUSY) from arxiv:1407.4081.
   Furthermore, the 1-loop matching conditions O(alpha_b) and
   O(alpha_tau) as well as the 2-loop matching condition O(alpha_t^2)
   from SUSYHD (arxiv:1504.05200) are implemented.  The Higgs pole
   mass is calculated at complete 1-loop order plus 2-loop
   contributions O(alpha_t^2) and O(alpha_t alpha_s) from
   arxiv:1205.6497.  The calculation of the Higgs pole mass in the
   HSSUSY model coincides with the one obtained with SUSYHD 1.0.2 with
   a relative deviation of < 0.06%.

 * Feature: Allow adding 3-loop gluino contribution to Higgs
   self-energy in split-SUSY models with a physical singlet Higgs.
   The 3-loop gluino contribution is enabled by default in the
   SplitMSSM.

 * Change (commit f7cd242): The `test' and `examples' modules are no
   longer loaded into the makefile by default.  To load them, run
   ./configure --with-optional-modules="test,examples"

 * Change (commit e86d23a): The FlexibleSUSY test suite is no longer
   shipped with the release tarball.  It can be obtained from the
   official git repository at
   https://github.com/FlexibleSUSY/FlexibleSUSY .

 * Change (commit 372bb96): Use FlexibleSUSY's own dependency file
   generator instead of using the corresponding compiler capabilities.

 * Bugfix (commit 20e88db): Use correct self-energy for 1st and 2nd
   generation charged leptons in *NoFV models.  Before commit 20e88db,
   the (heavy) tau self-energy was used to convert the running MS-bar
   electron and muon masses to DR-bar masses in *NoFV models.
   Corresponding test case:
   test_CMSSMNoFV_low_scale_constraint::test_delta_Yf()

FlexibleSUSY-1.2.2 [September, 08 2015]

 * Feature: The scale at which the EWSB output parameters are fixed
   can now be chosen by the user via the FSSolveEWSBFor[{...}] symbol.
   By default, the susy-scale is chosen.

 * Change (commit 5b9d653): If ./configure is run without the
   `--with-models=<models>' argument, no models will be build.  In
   former FlexibleSUSY versions if the `--with-models=<models>'
   argument was missing, all models were build.

 * Bugfix (commit 5530bf9): Defining a scale to be a running mass, for
   example SUSYScale = M[hh], resulted in a compilation error.

 * Bugfix (commits 2d6c0d2, 87cfe28): use SLHA input value of the Z
   pole mass as low-energy scale, instead of the hard-coded value MZ =
   91.1876 GeV.

 * Bugfix (commit 1ac0aa0): Use math/physics index convention (index
   starting with 1) in the comments of the extra user-defined SLHA
   output blocks.

 * Bugfix (commit 0737c4d): Properly convert greek symbols in function
   arguments.  Fixes #5.  Thanks to Dylan Harries.

 * Bugfix (commit f4eed5d): Put class Complex into softsusy namespace
   to avoid ambiguities in operator*().  Fixes #6.  Thanks to Dylan
   Harries.

FlexibleSUSY-1.2.1 [July, 07 2015]

 * Feature: The model name is printed in SPINFO[5] and the SARAH
   version is printed in SPINFO[9].

 * Bugfix (fea4d59): The MODSEL block was not read if SLHA input is
   passed to the spectrum generator via stdin.  Thanks to Peter
   Drechsel.

FlexibleSUSY-1.2.0 [June, 26 2015]

 * Feature: Allow the user to add 3-loop beta-functions in the SM.
   The beta-functions are taken from SUSYHD v1.0.1 (arXiv:1504.05200)
   and arxiv:1303.4364.

 * Feature: Allow the user to add 2-loop Higgs self-energy corrections
   O(alpha_t^2 + alpha_t alpha_s) in the SM.  The self-energy
   corrections were taken from arxiv:1205.6497.

 * Feature: Allow the user to provide SLHA input via stdin if the SLHA
   input file name is set to - .
   Example:
   $ cat model_files/CMSSM/LesHouches.in.CMSSM | \
        models/CMSSM/run_CMSSM.x --slha-input-file=-

 * Feature: Allow the user create standalone executables that don't
   depend on dynamically linked libraries.  See README for more
   details.

 * Bugfix (commit 3843ea7): Rewrite pole mass tachyon check to fix a
   confusion between goldstone and Higgs bosons in the CP-violating
   MSSM.

 * Bugfix (commit e2009f7): Adding missing declaration of input
   parameters in the generated DependenceNum functions.  This fixes a
   compilation error in the NE6SSM or the UMSSM if ThetaWp is set to
   an expression that involves the charges.

 * Bugfix (commits d80c30f, e6c8dda): Correcting input scale of
   tan(beta) in the lowNMSSM according to SLHA-2 convention.  The
   model file lowNMSSMTanBetaAtMZ has been added, where tan(beta) is
   input at MZ.

FlexibleSUSY-1.1.1 [June, 08 2015]

 * Bugfix (commit e1ea433): Catch NaNs from Slavich's NMSSM 2-loop
   self-energies.

FlexibleSUSY-1.1.0 [May, 31 2015]

 * Feature: Calculation of DR-bar weak mixing angle from Fermi
   constant and Z pole mass.  The implementation is based on
   expressions from SOFTSUSY and works for the SM, MSSM, NMSSM and
   their variants.  The method for the calculation of the weak mixing
   angle can be selected via the FSWeakMixingAngleInput variable in
   the FlexibleSUSY model file.
   Example:
   FSWeakMixingAngleInput = FSFermiConstant; (* or FSMassW *)
   Note: To achieve the maximum accuracy available, set the threshold
   corrections loop order to 2 (FlexibleSUSY block entry 7)

 * Feature: Support for non-SUSY models, renormalized in the MS-bar
   scheme.

 * Feature: 2-loop QCD corrections can be added when calculating the
   top pole mass from the top DR-bar mass.  These 2-loop contributions
   can be enabled/disabled using entries 13 or 4 of the FlexibleSUSY
   block in the SLHA input file.

 * Feature: In the shipped FlexibleSUSY model files, the corresponding
   default SARAH model file is specified.  This allows a user to
   create a new model with the simplified command:
   ./createmodel --name=CMSSM
   The default SARAH model file to be used with a given FlexibleSUSY
   model file can be set via FSDefaultSARAHModel = <model>;

 * Feature: Complex model parameters are now supported.

 * Feature: The CKM and PMNS matrix can now be used as low-energy
   inputs.  They are read from the VCKMIN and UPMNSIN input blocks,
   respectively.  Linked to this, the new model file CMSSMCKM was
   added to demonstrate the input of the CKM matrix at low energies.

 * Feature: Mark parameter points as invalid, for which the
   calculation of one of the pole masses failed due to
   non-convergence.

 * Feature: New (non-templated) intermediate model class
   <model>_mass_eigenstates, which is able to calculate the pole and
   running mass spectrum.  <model>_mass_eigenstates is derived from
   <model>_soft_parameters.  The templated model class
   <model><Two_scale> is now derived from <model>_mass_eigenstates .

 * Bugfix (commit 6da2cbd, 8113e32a): ensure that the MSSM-like CP-odd
   Higgs mass is used in the two-loop Higgs self-energies and
   tadpoles.  Before, there were cases where a Goldstone boson mass or
   a singlet-like pseudoscalar mass was used.

 * Bugfix (commit 29a0833): incorporate tadpole contributions in pole
   masses of singlets

 * Bugfix (commit c64a333): Softsusy's B1 function is now thread-save.
   Before commit c64a333, the tau pole mass was varying due to a race
   condition, if multi-threading is enabled and neither fflite nor
   looptoos is used.

 * Bugfix (commit d035544): Ignore trivial EWSB eqs.  Makes the MRSSM
   work in FlexibleSUSY with SARAH 4.5.x.

 * Bugfix (commit d8a1521): The SM() preprocessor macro has been
   renamed to LowEnergyConstant() in order to avoid collisions with
   the copy constructor of the SM model class.

 * Bugfix (commit 0c7a7ac): chop beta-function values smaller than the
   zero-threshold to avoid failures of the RK integrator.  The
   zero-threshold is 1e-11 by default and can be changed via
   Beta_function::set_zero_threshold() or entry 14 in the SLHA input
   file.

 * Bugfix (commit 29a1578): Ignore goldstone boson "pole mass"
   tachyons.

FlexibleSUSY-1.0.4 [January, 15 2015]

 * Add new user example program run_cmd_line_<model>.x to run a
   parameter point using command line parameters instead of an SLHA
   input file.

 * Allow input parameters in first guesses of scale definitions, for
   example
   SUSYScaleFirstGuess = Sqrt[Sqrt[LHInput[mq2[3,3]] * LHInput[mu2[3,3]]]]

 * Adding support for FlexibleSUSY addons.  They are placed inside the
   addons/ directory and can be configued and compiled via
   ./configure --with-addons=<addon> && make

 * Adding EWSB solvers using a fixed-point iteration (FPIRelative,
   FPIAbsolute, FPITadpole).  FPIRelative is now the first default
   solver used.  Thanks to Dylan Harries.

 * Adding new NMSSM model file `NUTNMSSM' with non-universal soft
   Higgs masses (EWSB output) and non-universal trilinear couplings
   A_lambda and A_kappa at MX.

 * Read user input W boson pole mass form SMINPUTS block entry 9.

 * Read user input Z boson pole mass from SMINPUTS block entry 4.

 * Automatic check for non-perturbative dimensionless model parameters
   at the high-scale.  The check can be disabled by stetting
   FSCheckPerturbativityOfDimensionlessParameters = False;
   in the model file.  The threshold can be set via the
   FSPerturbativityThreshold variable.  The default threshold is
   N[Sqrt[4 Pi]] = 3.54491.

 * Check for tree-level tachyons at each scale (MZ and M_SUSY)

 * Allow to force SLHA output for unphysical points (for example where
   tachyons exist) in FlexibleSUSY block, entry 12.

 * Bugfix (commit 6f7d3de): allow plain model parameters for scale
   definition, for example in the form SUSYScale = vu .

 * Bugfix (commit 44baa73): allow model parameters in first guesses of
   scale definitions, for example
   SUSYScaleFirstGuess = Sqrt[mq2[3,3] * mu2[3,3]]

 * Bugfix (commit 77dce8b): correct momentum guess for the calculation
   of the self-energies with LowPrecision.

 * Bugfix (commit fb0b906): Fix compilation with g++ 4.5.3.

 * Support Intel C++ compiler versions 12.1 and 13.x (commits 78d73e7
   and bf5a08e)

 * More descriptive error message when an exception is thrown.

FlexibleSUSY-1.0.3 [November, 21 2014]

 * Allow selection of Higgs 2-loop contributions in SLHA input file

 * Allow extra user-defined SLHA output blocks

 * Allow user-defined matrix- or vector-like SLHA input parameters

 * Support low-energy quark flavour violation via CKM matrix

 * Bugfix (commit 5f78968): perform residual color contractions before
   stripping group factors.  Thanks to Philip Diessner and Wojciech
   Kotlarski.

 * Bugfix (commit 7160095): Correcting check for tachyons in pole
   masses of scalar particles, calculated with LowPrecision

 * Bugfix (commit a7a33d3): Implement reading of data from multiple
   SLHA blocks with the same name.  Subsequent block entries will
   overwrite former entries.

FlexibleSUSY-1.0.2 [July, 15 2014]

 * Bugfix (commit 689141da): Enable non-quadratic superpotential
   coupling matrices.

 * Bugfix (commit d0e9cdb): Correctly set low-energy data (read from
   the SLHA input file) in the low-energy constraint.

 * Bugfix (commit 6414e46): Convert fermion masses and mixing matrices
   to SLHA convention in the SLHA output.

 * Install specimen SLHA input files in the model directory when one
   runs the createmodel script.

 * Work around fields in Cp[] carrying an invalid index that cause
   Part::partw when passed to SARAH`Vertex[].

 * Support Cygwin on MS Windows

 * New model file for the TMSSM (triplet Higgs model)

FlexibleSUSY-1.0.1 [June, 11 2014]

 * Bugfix (commit 4dc897e): consts.hpp is not distributed but appears
   in the list of installed headers

FlexibleSUSY-1.0.0 [June, 10 2014]

 * Bugfix (commit 399a1c8): renaming SoftsusyMSSM and SoftsusyNMSSM
   model classes and files to make `make all-test' work on HFS (fixes
   #2).

 * Bugfix (commit cfc2562): correcting MS-bar to DR-bar conversion of
   fermion masses mb and mtau.

 * Bugfix (commit ceecc4a): fixing compilation error with Intel icpc
   14.0, Build 20130728 (and GNU STL 4.6.4 and 4.8.1).

 * Bugfix (commit db60205): fixing linking error of
   test/test_MSSM_NMSSM_linking.x in case LoopTools is used.

 * Bugfix (commit 32c3222): generalizing color summation routine to
   handle single-generation fields and non-fundamental
   representations.

 * Bugfix (commit 3fd2699): Correcting the determination of the number
   of EWSB eqs. in case of CP violating models.

 * Bugfix (commit c9cc34f): Reset fermion phases when clear() is called.

 * Bugfix (commit faa0fb6): adding boost include directory to CPPFLAGS
   in the src module.

 * Bugfix (commit ac8e38e): impose EWSB before calculating the
   spectrum.

 * Set minimum required SARAH version to 4.0.4, because it implements
   the full two-loop VEV beta functions from arXiv:1310.7629 .

 * Add stand-alone examples to illustrate how to use FlexibleSUSY's
   classes and libraries independently of FlexibleSUSY's build system.

 * Add tower example to illustrate how to glue multiple models to form
   a stack of effective field theories.

 * Add customized-betas example to illustrate how to replace an
   auto-generated C++ component by something of an alternative origin.

 * Implement leading two-loop MSSM and NMSSM tadpoles from Slavich
   (used in the EWSB conditions).

 * Implement leading two-loop MSSM and NMSSM CP-even and CP-odd Higgs
   self-energy contributions from Slavich.

 * Allow to constrain the boundary condition scale via the model file
   variables {Low,SUSY,High}ScaleMinimum and
   {Low,SUSY,High}ScaleMaximum.

 * Allow explicite setting (and disabling) of the Yukawa couplings in
   the constraints.

 * Enable/disable multi-threading at the configure level

 * lower required g++ version to 4.4.7

 * Enable source code export without the meta code via `make
   install-src'.

 * Add FFLite module as a thread-safe alternative to LoopTools

 * Create helper function to find the LSP.

 * Allow to select beta-function loop order in the SLHA input file.

 * Allow disable/enable threshold corrections in the SLHA input file.

 * Rename pole mass calculation precision option and set them in the
   model file.

FlexibleSUSY-0.5.3 [January, 21 2014]

 * Bugfix (commit 44903c): correcting malformed print out in
   config/list_sarah_model_files.sh in case model files do not exist

 * Bugfix (commit 3aae11): Prevent hard-coding of the running Weinberg
   angle in terms of the gauge couplings

 * Bugfix (commit ce4a73): Generalize calculation of gauge couplings
   at the low-scale (fixes #1)

 * Vertices are saved in a file to avoid repeating same calculation.

FlexibleSUSY-0.5.2 [January 14, 2014]

 * Bugfix (commit 58f8f9): Convert beta functions which are identical
   zero to the data type of the corresponding parameter.

 * Bugfix (commit e5f937): Correcting check of SARAH patch level
   against minimum required patch level.

 * Bugfix (commit e2d43b): Adapting free phases of fermion fields if
   mass is less than zero.

 * Bugfix (commit e777e1): Converting indices to C convention in
   tree-level EWSB equations.

 * Set minimum required SARAH version to 4.0.3, because it includes a
   bug fix in the index structure of the charged Higgs self-energies.

 * Allow setting of single matrix/ vector elements in the constraints.

 * Model files are now in the directory model_files/
   (instead of templates/)

 * The command line arguments of the createmodel script changed.
   Please see ./createmodel --help for more details.

 * Add support for the LHInput[p] command in constraints, which reads
   the parameter `p' from the SLHA input file.

 * Constrain time used to simplify the beta functions (default: 120
   seconds per beta function).  To change the time constraint, set
   FlexibleSUSY`FSSimplifyBetaFunctionsTimeConstraint .

 * Avoid swapping by distributing the calculation of the two-scale
   beta functions among multiple .cpp files.

 * Introduce separate meta code stamp (triggers running of the meta
   code) with name models/<model-name>/00_DELETE_ME_TO_RERUN_METACODE

FlexibleSUSY-0.5.1 [November 23, 2013]

 * Handle parameters of type vector in the beta functions.

FlexibleSUSY-0.5 [November 18, 2013]

 * Store particle masses as Eigen::Array and mixing matrices as
   Eigen::Matrix.