How to handle a personal development branch with many changes of varying scope?
In the course of working on slow scattering glitch inference in my development branch I did not keep a properly cleaned git history or separation of feature branches. This meant that major features (the above) are mixed in with bugfixes or minor (hopefully uncontroversial) feature additions.
I suspected that the best way to handle this was to break up my changes into a number of feature branches, accomplishing goals of different scope. They are:
- !1261 (merged) a minor plotting bugfix
- !1263 (merged) a simple change to allow injection of non-cbc signals
- !1262 a change to allow passing functions which generate proposal cycles to the bilby_mcmc sampler
- !1264 (merged) to provide identity conversion/generation functions (so that the default is not CBC conversion functions, which fail for non-cbc models)
- and !1265 (closed) the main feature branch, which maybe should be a dependent package instead.
Is that the correct way to have approached this? If not please let me know and I can adapt accordingly.