Updated to 25a481d8, still getting the same error and backtrace when
generating agendas where the memq block evaluates non-nil. :/
On Thu, Feb 7, 2013 at 1:33 AM, Bastien <b...@altern.org> wrote:
> Hi Kyle,
>
> Kyle Machulis <k...@nonpolynomial.com> writes:
>
>> As of commit 8517be79b5c1fe165d23ea65ad70a282e9c595bb (bisected to
>> find this), agenda compilation died with the following backtrace for
>> me:
>>
>> Debugger entered--Lisp error: (error "Bad timestamp `'
>> Error was: (Not a standard Org-mode time string: )")
>> signal(error ("Bad timestamp `'\nError was: (Not a standard Org-mode
>> time string: )"))
>> error("Bad timestamp `%s'%s\nError was: %s" "" "" ("Not a standard
>> Org-mode time string: "))
>> byte-code(...)
>> org-time-string-to-absolute("")
>> byte-code(...)
>> org-agenda-get-timestamps(nil)
>> org-agenda-get-day-entries("~/emacs_org/mozilla/mozilla.org" (2 6
>> 2013) :deadline :scheduled :timestamp :sexp)
>> apply(org-agenda-get-day-entries "~/emacs_org/mozilla/mozilla.org"
>> (2 6 2013) (:deadline :scheduled :timestamp :sexp))
>> byte-code(...)
>> byte-code(...)
>> org-agenda-list(nil)
>> call-interactively(org-agenda-list)
>> byte-code(...)
>> org-agenda(nil)
>> call-interactively(org-agenda nil nil)
>>
>> It's dying when it hits a headline as follows:
>>
>> * Meeting 13:30
>> , <%%(memq (calendar-day-of-week date) '(3))>
>>
>> I'm guessing this has something to do with the agenda sorting updates
>> in the commit not liking the fact that the sexp evaluates to true. If
>> the sexp evaluates nil for the day the agenda is trying to parse,
>> nothing happens. If it evaluates to something other than nil (t if I
>> use (= 3 (calendar-day-of-week date)), 3 if I use memq on a wednesday,
>> etc...), I get the above error.
>
> Thanks for reporting this. Yes, diary sexp are special and can't
> really be handled here. I fixed this in master, please confirm this
> is okay.
>
> --
> Bastien
