Weirdness with Agenda and "C-u r"

Weirdness with Agenda and "C-u r"

[Eric Abrahamsen]

Something seems to have gone off with the `org-tags-view' Agenda view:
if I hit "C-u r" to search with another string, I get this traceback:

Debugger entered--Lisp error: (invalid-function 4)
  (4)
  (org-tags-view (4) (if current-prefix-arg nil "publishing"))
  (let nil (org-tags-view (4) (if current-prefix-arg nil "publishing")))
  eval((let nil (org-tags-view (4) (if current-prefix-arg nil "publishing"))))
  org-let(nil (org-tags-view (4) (if current-prefix-arg nil "publishing")))
  org-agenda-redo((4))
  funcall-interactively(org-agenda-redo (4))
  call-interactively(org-agenda-redo nil nil)
  command-execute(org-agenda-redo)

I tried e-debugging `org-agenda-redo', but something magic happens when
it gets to the org-let, and no light was shed.

I seem to remember there have been some changes to the tag matching code
recently -- maybe that's it? Happy to bisect if necessary.

Eric

Re: Weirdness with Agenda and "C-u r"

[Marco Wahl]

Eric Abrahamsen <eric@ericabrahamsen.net> writes:

> Something seems to have gone off with the `org-tags-view' Agenda view:
> if I hit "C-u r" to search with another string, I get this traceback:
>
> Debugger entered--Lisp error: (invalid-function 4)
>   (4)
>   (org-tags-view (4) (if current-prefix-arg nil "publishing"))
>   (let nil (org-tags-view (4) (if current-prefix-arg nil "publishing")))
>   eval((let nil (org-tags-view (4) (if current-prefix-arg nil "publishing"))))
>   org-let(nil (org-tags-view (4) (if current-prefix-arg nil "publishing")))
>   org-agenda-redo((4))
>   funcall-interactively(org-agenda-redo (4))
>   call-interactively(org-agenda-redo nil nil)
>   command-execute(org-agenda-redo)
>
> I tried e-debugging `org-agenda-redo', but something magic happens when
> it gets to the org-let, and no light was shed.

> I seem to remember there have been some changes to the tag matching code
> recently -- maybe that's it? Happy to bisect if necessary.

FWIW I cannot reproduce the behavior you describe with commit 164555b4
having some files in `org-agenda-files' with some tags.

Concretely with file

#v+
* One                                                                  :tag1:

One.

* Two                                                                  :tag2:

Two.
#v-

in `org-agenda-files' and the actions

#v+
M-x org-tags-view tag1
C-u r tag2
#v-

the agenda

#v+
Headlines with TAGS match: tag2
Press ‘C-u r’ to search again with new search string
  tmp:        Two                                                         :tag2:
#v-

appears.


HTH,
-- 
Marco Wahl
GPG: 0x49010A040A3AE6F2

Re: Weirdness with Agenda and "C-u r"

[Eric Abrahamsen]

Marco Wahl <marcowahlsoft@gmail.com> writes:

> Eric Abrahamsen <eric@ericabrahamsen.net> writes:
>
>> Something seems to have gone off with the `org-tags-view' Agenda view:
>> if I hit "C-u r" to search with another string, I get this traceback:
>>
>> Debugger entered--Lisp error: (invalid-function 4)
>>   (4)
>>   (org-tags-view (4) (if current-prefix-arg nil "publishing"))
>>   (let nil (org-tags-view (4) (if current-prefix-arg nil "publishing")))
>>   eval((let nil (org-tags-view (4) (if current-prefix-arg nil "publishing"))))
>>   org-let(nil (org-tags-view (4) (if current-prefix-arg nil "publishing")))
>>   org-agenda-redo((4))
>>   funcall-interactively(org-agenda-redo (4))
>>   call-interactively(org-agenda-redo nil nil)
>>   command-execute(org-agenda-redo)
>>
>> I tried e-debugging `org-agenda-redo', but something magic happens when
>> it gets to the org-let, and no light was shed.
>
>> I seem to remember there have been some changes to the tag matching code
>> recently -- maybe that's it? Happy to bisect if necessary.
>
> FWIW I cannot reproduce the behavior you describe with commit 164555b4
> having some files in `org-agenda-files' with some tags.

Thanks for your reply. That's pretty weird! I just pulled and re-made,
and emacs -Q still does it, even with no Org agenda files defined. Just
a plain empty Agenda, then C-u r. I wish I knew what edebug was doing
with that org-let.