"Ping Yin" [off-list ref] writes:
On Sat, Apr 12, 2008 at 6:30 AM, Junio C Hamano [off-list ref] wrote:
quoted
...
Same comment applies to cmd_update() and cmd_status(). I would strongly
suspect that status may want to ignore missing name/url and show the usual
diff, as it does not even have to require the submodule to interact with
any remote repository at all. The user may be privately using the
submodule and does not even need to push it out nor pull it from
elsewhere, and in such a case, .gitmodules may not even be populated with
an entry for that submodule, ever, not just as a "right in the middle of
adding" status.
This patch just does refactor, i do this in my sixth patch "Don't die
when command fails for one submodule"
Fair enough.
As usual, the preferred order is to first refector, fix without
enhancement as necessary, then finally build on more solidified base.