Repository transition

From Einstein Toolkit Documentation
Revision as of 14:21, 24 March 2014 by Rhaas (talk | contribs) (add comment on splitting X and TestX)
Jump to: navigation, search

Arrangements should go into one git repository each, except where noted:

  • EinsteinAnalysis
  • EinsteinBase
  • EinsteinEOS
  • EinsteinEvolve
  • EinsteinInitialData
  • EinsteinUtils
  • incoming
  • manifest
  • pyGWAnalysis
  • tools
  • VizTools/CarpetHDF5
  • VizTools/CarpetN5
  • VizTools/DataVaultXVSutils
  • CactusBase
  • CactusElliptic
  • CactusIO
  • CactusNumerical
  • CactusUtils
  • TAT
  • flesh
  • PITTNullCode
  • CactusConnect
  • EinsteinInitialData/GRHydro_InitData
  • CactusExamples:
  • CactusTest (GRHydro_InitData goes here, as Test_GRHydro, RH: this is a license conflict. GRHydro_InitData is GPL)
  • CactusWave
  • CactusPUGH

possible conflicting locations

  • it would make sense to keep thorn X and its TestX thorn in the same repository. They should then not be split up into X in EinsteinEvolve and TestX in Cactus Test. This affects:
    • GRHydro and GRHydro_Init_Data
    • MoL and TestMoL
    • Carpet and its Test thorns (currently Carpet / CarpetExtra arrangements)

In particular for the case of GRHydro, GRHydro itself has had many more frequent changes requiring updates to test data than GRHydro_Init_Data so it makes sense to keep the test parfiles and test data with GRhydro rather than with GRHydro_Init_Data (the future Test_GRHydro).