On 2/29/2016 4:07 PM, Ross Berteig wrote:
On 2/29/2016 3:50 PM, Joe Mistachkin wrote:
I think the "fix" here is to have both places check all the variables
in a well-defined order, e.g. FOSSIL_USER should [still] be first,
followed by the variables for POSIX, and finally those for Windows.
Sounds sensible to me. Setting variables that "belong" to the other
platform is sometimes a recipe for confusion, but doing fewer things in
platform-specific ways is easier to document and easier to understand.
I shuffled the order in db_create_default_users() to check the two Unix
names ahead of the Windows name, and to check all three names on all
platforms, which makes it match the order in user_select().
I also added a new test-usernames command that prints all of the sources
of possible usernames, ending with a call to user_select() and printing
what that selected.
I rewrote the section on usernames in my new Environment Variables and
Global Options document to read better and to now fully agree with how
fossil discovers a username for use with the local repository and when
creating new repositories.
I think that document is now mostly true, but certainly could use some
editorial review.
I propose to merge my ross-doc-env branch to trunk later tomorrow so
that more eyes can land on the documentation, unless there's something
grossly wrong that needs fixing first.
--
Ross Berteig r...@cheshireeng.com
Cheshire Engineering Corp. http://www.CheshireEng.com/
+1 626 303 1602
_______________________________________________
fossil-dev mailing list
fossil-dev@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/fossil-dev
