On Sat, Feb 27, 2016 at 3:25 AM, Ross Berteig wrote:
> So I decided that rather than gripe about it, it would be a good use of my
> tech-writing hat to just write the missing document.
>
:-D
> So I can just fix repo_init by setting FOSSIL_HOME, right?
>
> Well, probably. In shell.c, we have fi
I got interested in why tester.tcl's repo_init command was failing to
keep the transient repositories used by test cases out of my fossil all
list, despite the fact that it deliberately sets $HOME to prevent that.
That led to the question of where are the environment variables used by
fossil d
