I'd vote against the switch to git.  Simple things, like svn update [remote directory] seem too complicated. As for slightly involved tasks like code bisection to locate the past commit that broke the test, well....I've never been able to figure this one out so I am pretty lost trying to debug DIALS.  I do not wish to make things difficult for the maintenance of cctbx.xfel.

Nick

Nicholas K. Sauter, Ph. D.
Computer Staff Scientist, Molecular Biophysics and Integrated Bioimaging Division
Lawrence Berkeley National Laboratory
1 Cyclotron Rd., Bldg. 33R0345
Berkeley, CA 94720
(510) 486-5713

On Mon, Jun 13, 2016 at 8:52 AM, Luc Bourhis <luc_j_bourhis@mac.com> wrote:

On 13 Jun 2016, at 17:09, Graeme.Winter@Diamond.ac.uk wrote:

… is there any enthusiasm out there for moving cctbx to git? The more I work with git the more I wish this existed properly for cctbx.
 
I’ve raised this a bunch of times and always it is “not right now” however “a good idea” – this’ll be the last time I ask…
 
Could we move to git, please? I note the bootstrap script for Windows now has a git dependency even!

+1

… and time to share some thoughts! It would be great if we could get the cctbx up and running by just cloning, configuring and building without needing to download anything else.

That could be done by adding SCons, Boost, and ccp4io as  submodules in the cctbx repository. And I reckon ccp4io_adaptbx should be added as a subtree since it is developed by the same people as the cctbx.

Just an idea to think about as I haven’t not worked out the details at all.

Best wishes,

Luc



_______________________________________________
cctbxbb mailing list
cctbxbb@phenix-online.org
http://phenix-online.org/mailman/listinfo/cctbxbb