whydoubt at gmail.com
Thu Apr 20 02:55:44 CEST 2017
I have long considered working on blame functionality for cgit, but
recently decided to pursue it. From looking at the history of this list, it
appears that it was last brought up August 2015. The proposal as I
understand it was to spawn off a call to the git command. I do not care for
that solution, but there are still two remaining options that I see.
Option 1: take git's builtin/blame.c, pare it down to the essentials, and
build ui-blame.c around that.
Option 2: work with git upstream to pull portions of builtin/blame.c into
libgit.a, rework builtin/blame.c to use that, and build ui-blame.c to use
that as well.
Having looked at builtin/blame.c for a bit, I think option 2 would be quite
difficult to do cleanly. While option 1 is not ideal, I think it is the
most reasonable. However, I wanted to get some feedback before I went to
the effort only to be told, "only option 2 will do".
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the CGit