Gosia Development Plan: Phase 1
The current Gosia code uses extensive overlaying of the arrays to maximize the speed and also to fit within the < 2MB of memory limitations of a CDC Cyber 1675 compute used when the code was developed at Rochester in 1980. This overlaying is unnecessary using modern computers which typically have over three order of magnitude more core memory and orders of magnitude greater speed. The complex intertwined coding makes upgrading of the current version of Gosia impractical. Therefore there was unanimous agreement that the highest priority is to partition the Gosia coding into its natural separate subroutines to simplify the structure and facilitate future development. It was decided that the Nigel Warr version of the code will be adopted for the future development to take advantage of the many coding improvements already implemented by Nigel. The Steering Committee members are tackling the following tasks:
1 ) Add the Leuven calibration and change the PIN diode option to address immeditate needs of the community. [Warr]
2) Continue adding comments to Gosia coding, and split use of the zeta array and other common blocks to simplify the coding. [Warr]
3) Increase the allowed number of mesh points, level energies and matrix elements. [Warr]
4) Develop a B-spline interpolation procedure to replace the troublesome Lagrange interpolation procedure which is used for interpolation during integration over energy and angle as well as for calculating internal conversion coefficients. This is essential before using an increased number of mesh points. [Napiorkowski]
5) Explore a more robust way to handle inverse kinematics for integration at extreme mass ratios. [Hayes]
6) Develop a separate intuitive input generator program for Gosia. [Napiorkowski, Pietak]
7) Develop a new Gosia website [Cline]
8) Upgrade the Gosia Users manual [Cline]
9) Implement a web based forum to facilitate communication among users. [Cline]
10) Prepare the first Gosia Newsletter. [Cline]
11) Restore Czosnyka version fo Gosia2 incorporating. [Warr]
12) Prepare a set of test benchmark cases for quality assurance testing of Gosia. [Hayes]
Stage 1 development will finish when the above have been completed and the code has been fully tested to ensure quality assurance. Then this phase 1 Gosia code will be released.

Gosia Development Plan: Phase 2
Once the phase 1 tasks are complete it will be possible to envision major upgrades to the capabilities of the gosia codes. Proposed longer term upgrades include:
1) Develop and evaluate a genetic algorithm procedure for minimization. [Napiorkowski, Pietak, Srebrny].
2) Expand Gosia to allow fit to the more sensitive particle-gamma-gamma data [Cline]
3) Allow more flexibility to fit user-defined parameters employed to generate the matrix elements. [Hayes]
4) Add input of electron data as well as include E0 decay
5) Remove cylindrical symmetry limitation for gamma detectors. [Cline]
6) Develop a new version of Gosia2.
Development of these options may occur once the phase 1 development is complete.