I am in the process of refactoring JumboConverters and would like guidance on the best way to fork and merge JC. At present I shall assume that I can, if necessary, control patches to JC (although currently several people have write access).
The current public repo is:
https://bitbucket.org/wwmm/jumbo-converters/overview (abbreviated to WWMM-JC)
and I have cloned / forked this at
https://bitbucket.org/petermr/jumbo-converters/overview (PMR-JC)
I have been making changes and committing to PMR-JC.
At some stage I shall wish to recommit these to WWMM-JC and I don't want this to get too far out of sync. So do I assume that:
* I pull WWMM-JC at regular intervals (? weekly)
* transfer the WWMM-JC changes to my local PMR-JC and merge
* recommit PMR-JC to bitbucket.
Then at some stage:
* announce there will be a merge of PMR-JC with WWMM-JC
* diff the tests in some way (so people can see what new functionality there is?)
* invite people to link against PMR-JC
* if no howls of protest then merge PMR-JC with WWMM-JC
Comments welcome to stop me getting this wrong...
--
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069