Gui release notes

These versions are listed from newest to oldest.

=Versions 1.x=

==Version 1.1.1==

===Known bugs and issues===

In all versions, it is possible to save an empty session over an existing session.  Frequent backups of the working directory should be made.

This version is not backward-compatible with the beta versions; users must re-define the experimental
definitions.  Level and matrix data CAN be exported from beta versions and imported to this version.  All
later versions are expected to preserve backward compatibility with version 1.1.
Plots of collision functions cannot be generated for all states.

===Description===

See version 1.0.

====Bug-fixes====

A bug that sometimes prevented correlated errors from being read from the Gosia output has been fixed.

===Recommendations===

To migrate from a beta version to version 1.1.1, export the level scheme and matrix, then import them into a
new session in this version.

==Version 1.1==

===Known bugs and issues===

In all versions, it is possible to save an empty session over an existing session.  Frequent backups of the working directory should be made.

This version is not backward-compatible with the beta versions; users must re-define the experimental
definitions.  Level and matrix data CAN be exported from beta versions and imported to this version.  All
later versions are expected to preserve backward compatibility with version 1.1.
Plots of collision functions cannot be generated for all states.


The "fully-automatic" loading of yield data reports Ge detector numbers that are 1 larger than the correct number.  This is misleading but does not cause any accuracy errors.


===Description===

See version 1.0

====Bug-fixes====

A bug in adding interband matrix elements is fixed.  Version 1.0 did not include all allowed couplings.

===Recommendations===

To migrate from a beta version to version 1.1, export the level scheme and matrix, then import them into a
new session in this version.

==Version 1.0==

===Known bugs and issues===

In all versions, it is possible to save an empty session over an existing session.  Frequent backups of the working directory should be made.

This version is not backward-compatible with the beta versions; users must re-define the experimental
definitions.  Level and matrix data CAN be exported from beta versions and imported to this version.  All
later versions are expected to preserve backward compatibility with version 1.0.
Plots of collision functions cannot be generated for all states.

The "fully-automatic" loading of yield data reports Ge detector numbers that are 1 larger than the correct number.  This is misleading but does not cause any accuracy errors.

===Description===

Faster setup of calculations.

Defining complicated detector arrays can be done in one operation.

A library of detector types and arrays is included.

Loading of all experimental yield data can be done in one operation.

Simulations now use the standard definition of Ge efficiency.

Simulated yield tables are clearer.

Stopping power and exit energy are calculated by Elast automatically.

The Rochester SRIM server can be called for more accurate stopping power data.

Thick-target experiments are handled accurately, including backscatter detection from the surface.

The correction to the gamma angular distribution now includes the motion of the emitting nucleus.

Deorientation coefficients and lifetimes can be calulated in a few seconds.

====Bug-fixes====

Merging, deleting and reordering of bands are allowed at any time.

Any number of bands can be merged.  (See restrictions regarding duplicate spin states and different-parity
states in one band.)

Malfunctions during plotting of yield data are fixed. The overall normalization for all data sets is now
used in plots comparing measured and predicted yields.

The reported chi-squared values are properly-weighted on plots and in reports.  (Chi-squared and
normalization was done correctly in fits in all versions.)

It is no longer necessary to reload yield data after changing the level scheme, unless new states are
added.

===Recommendations===

To migrate from a beta version to version 1.0, export the level scheme and matrix, then import them into a
new session in this version.

=Beta Versions=

==Version 2.0.8.beta==

===Known bugs and issues===

The GUI allows only simple changes to the level scheme after the matrix and experiment have been defined (adding and deleting of bands).  The release candidate (expected September 2011) has the ability to merge and reorder bands at any time.

If the level scheme is changed after calculated or experimental yield data are in memory, the user must reload the experimental data and generate new calculated yields.  This has been fixed in the release candidate version (expected September 2011).

===Description===

This version uses a more accurate estimate of a typical Ge detector's ''intrinsic'' efficiency (based on Gammasphere's efficiency) to give approximate "raw" count rates.  The total counts (for 100% efficienct detectors) were found to agree with perturbation calculations for weakly-coupled two-state systems, where the perturbation method is accurate.

Refer to the recommendations section if you are working with simulated yield data.

This version preserves the same backward compatibility with older session files as did version 2.0.6.beta.

This is expected to be the last beta release before a significantly upgraded release version expected in July.

===Recommendations===

If you are working with simulated data, refer to the page [[Rachel_simulated_yield_output | Rachel simulated yield output]] for a description of the new output format and the meaning of terms.


==Version 2.0.7.beta==

===Known bugs and issues===

None

===Description===

This is a bug-fix version relative to 2.0.6.beta.  

# The simulated yield data are now updated properly.  
# The simulated data now include the proper Ge detector solid angle factor.  Simulated yield data are corrected by dividing the simulated raw counts by a factor of <math>\epsilon \Delta\Omega_\gamma</math>, where <math>\epsilon</math> is the approximate intrinsic efficiency (<math>0<\epsilon <1</math>) and <math>\Delta\Omega_\gamma</math> is the solid angle subtended by the Ge detector.  In previous Rachel versions, the absence of the solid angle factor would cause a systematic disagreement between Ge detectors only if they subtended different solid angles.

Refer to the recommendations section if you are working with simulated yield data.

This version preserves the same backward compatibility with older session files as did version 2.0.6.beta.

This is expected to be the last beta release before a significantly upgraded release version expected in July.

===Recommendations===

If you are not working with simulated data, no corrections are needed.

ONLY if simulated data are in use, re-run the simulated data calculation.  Answer "yes" to the prompt to re-integrate the yield data.


==Version 2.0.6.beta==

===Known bugs and issues===

When calculating simulated yields, the new simulated yield data in session (and saved session file) will not be updated unless the user answers "y" to the prompt "Re-integrate the predicted yields? [Y/n]:"  This may or may not affect subsequent fits to the simulated data, depending on the changes to the matrix, experimental parameters, detector arrangements, etc.  The new simulated data are written correctly to the terminal and to the gosia yield file, so fits would not be affected, but a large disagreement between simulated data and calculated yields could appear on plots.  ''Remedy:  always choose to re-integrated the yields when calculating simulated yield data.''

===Description===

This is a bug-fix version relative to 2.0.5.beta.  

Refer to the notes for "Version 2.0.5.beta" below for the details of the bug fix and the recommended solution to update your session files.

This version preserves the same backward compatibility with older session files as did version 2.0.5.beta.

This is expected to be the last beta release before a significantly upgraded release version expected in July.

===Recommendations===

After replacing an older version of Rachel with this version, re-integrate the calculated yields in memory by selecting "Integrated yields" and "Run gosia input" and clicking "Go."  This will update the calculated yields and overwrite any erroneous calculated yields that may be in memory from the bug.

==Version 2.0.5.beta==

===Known bugs and issues===

The bug in version 2.0.4.beta is still not completely fixed.  See below.  A new version will be released probably in mid-June 2011.  Refer to this post in the Gosia forum to see how to correct for the bug:

[http://www-user.pas.rochester.edu/~gosia/phpBB3/viewtopic.php?f=9&t=9 Forum post on correcting for the bug]

===Description===

This is a bug-fix version relative to 2.0.4.beta, and some user prompts and output information is clarified.  

The GUI now correctly informs you of the choices of kinematic solutions for inverse kinematics cases.  The safe energy is reported before asking for the initial beam energy.

Refer to the notes for "Version 2.0.4.beta" below for the details of the bug fix and the recommended solution to update your session files.

This version preserves the same backward compatibility with older session files as did version 2.0.4.beta.

This is expected to be the last beta release before a significantly upgraded release version expected in July.

==Version 2.0.4.beta==

===Known bugs and issues===

This version has one known bug:  if the user selects "View gosia input" or "Save gosia input" for some types of calculations, such as "Integrated yields," inaccurate point yields would be stored in memory after the operation.  This could be more or less apparent, depending on how accurate the point-scattering approximation is for the experiments defined.  The absolute cross sections then reported on plots could then be extremely inaccurate.

After running "Integrated yields" with "Run gosia input" selected, the calculated yields and absolute cross sections on plots would (again) be accurate.

This bug does not affect real yield data or simulated yield data.  This bug would not affect fitting of matrix elements, since the ''corrected'' yields used by Gosia are stored in an external file only.

====Remedies====

# Upgrade to version 2.0.5.beta.
# Re-run the "Integrated yields" calculation to store accurately-calculated yields in memory.  (Save the session.)  This will update all calculated data in memory and on plots.