David Bremner <david@tethera.net> writes: >> diff --git a/test/T310-emacs.sh b/test/T310-emacs.sh >> index 9d0df187..b89a100e 100755 >> --- a/test/T310-emacs.sh >> +++ b/test/T310-emacs.sh >> @@ -1030,7 +1030,7 @@ End of search results. >> YYY/notmuch_fail exited with status 1 (see *Notmuch errors* for more details) >> === ERROR === >> YYY/notmuch_fail exited with status 1 >> -command: YYY/notmuch_fail search --format\=sexp --format-version\=5 --sort\=newest-first tag\:inbox >> +command: YYY/notmuch_fail search --format\=sexp --format-version\=5 --sort\=newest-first --exclude\=false tag\:inbox >> exit status: 1" >> > > I don't understand why --exclude is defaulting to 'false' here. Isn't > this a change in behaviour? You're correct. This is because the new options notmuch-search-hide-excluded and it defaults to t, but when we call (notmuch-search) the value used is nil because we didn't supply one for the search call. If you call notmuch-search interactively then this isn't an issue, the interactive spec will substitute the value of the hide-excluded option. This is why all the tests that use the hello buffer to select the inbox and open a search in it correctly use the notmuch-search-hide-excluded option. This test does not do this. We could workaround this by changing the semantics of the variable. Make it notmuch-search-show-excluded and default to nil so nothing has to be changed in existing functions that call notmuch-search. Or we could switch to using cl-defun and have the default value substituted in the parameter list. I've never used cl-defun with interactive functions but long term I think that's the saner solution. -- Mohsin Kaleem _______________________________________________ notmuch mailing list -- notmuch@notmuchmail.org To unsubscribe send an email to notmuch-leave@notmuchmail.org