GENIE/Generator versioning scheme

GENIE/Generator releases are numbered with 3 dot separated integers as i.j.k, where the integers i,j,k correspond to the major, minor and revision number respectively. The corresponding SVN tag for version i.j.k is R-i_j_k (for example, GENIE version 2.0.2 is tagged as R-2_0_2).

GENIE/Generator version number semantics

• Versions with even minor number (eg 2.4.*) correspond to stable, fully validated physics/production releases.
• Versions with odd minor number (eg 2.3.*, 2.5.*) correspond to public `beta'/ candidate releases.
• Tagged versions always have an even revision number (eg 2.2.2 is a revision of the 2.2.0 physics / production release whereas 2.3.2, 2.3.4, 2.3.6,... are `candidate' releases for the 2.4.0 physics / production release).

GENIE/Generator codenames

The major production-quality releases of the GENIE/Generator product are code-named after modern extinct or endangered species (series of production releases: Auk, Bear (polar), Cheetah, Dugong,... ).

GENIE/Generator release qualifiers

The GENIE releases are marked as:

• pro :
  Validated production-quality versions recommended for physics studies.
• old :
  Older production versions that have been greatly superseeded by newer versions. These versions become unsupported. The typical lifetime of `physics' versions is of the order of 18-24 months.
• rc :
  Release candidates. Please use for testing new models and features, but avoid using for physics studies.
• special :
  Special releases prepared for a particular study, or an event such as a GENIE tutorial or a summer school. Avoid using outside the intended context.

GENIE/Generator comprehensive model configurations

Starting from v3.0.0, with the deployment of multiple comprehensive configurations (CMCs) and tunes, and the addition of a `--tune' argument in several apps, GENIE adopted the following uniform naming convention for its model configurations and tunes:

A comprehensive model configuration is identified by a 7-character string in the form

Gdd_MMv

where

1. G is a capital letter that identifies the author of the tune. The default value is G (for the GENIE collaboration).
2. dd is a number describing the year during which the model configuration was developed.
3. MM is a number (00, 01, 02, ...) identifying a model configuration branch.
4. v is a character (a, b, c, ...) enumerating variations / offshoots of a model configuration branch.

Note on test configurations:
Test comprehensive model configurations being developed, validated and tuned for deployment in future releases may be created temporarily and appear in public releases. This would allow users to experiment with early versions of these comprehensive models, making sure that these experimental models and their features are properly configured by GENIE. These configurations may not necessarily follow the above naming convention and should be identified by the use of an initial GTEST. Users are warned that test comprehensive model configurations may not be fully validated or tuned, and that their very definition may subject to changes before they emerge as an officially supported tune. We welcome feedback on these early snapshots of future model configurations but, for as long as they are classified as test configurations, support will be limited and we may not prioritise bug-fix revisions to address issues in test configurations.

GENIE/Generator tunes

Once a comprehensive model configuration is defined, a number of different tunes may be produced. Each distinct choice of a) fit datasets and dataset weighting scheme, b) parameters of interest and nuisance parameters, and c) prior constraints, leads to a different tune. A tune is identified by the model configuration name, and additional information enumerating the parameters and datasets. This is a 14-character string in the form

Gdd_MMv_PP_xxx

where

1. Gdd_MMv describes the model configuration (see above).
2. PP is a number identifying the set of tuned parameters. This parameter set is defined uniquely only in the context of a particular model configuration. The number 00 indicates that a model configuration has not been tuned by GENIE.
3. xxx is a number that identifies the dataset used for the model configuration tuning. This may include a unique set weights associated with each component dataset. The number 00 indicates that a model configuration has not been tuned by GENIE.