Discussion:
Diagnosing error after sending
Sebastian Hofer
2014-05-07 14:13:46 UTC
Permalink
Dear all,

I recently updated wanderlust (+ apel/semi/flim) from el-get. I am now
getting an

apply: Wrong type argument: number-or-marker-p, nil

error after sending an email. The email is sent correctly, but not moved to
the Sent folder. Can someone advise me how to diagnose the problem?

TIA
Sebastian
Juliusz Chroboczek
2014-05-07 21:18:52 UTC
Permalink
Post by Sebastian Hofer
Can someone advise me how to diagnose the problem?
M-x set-variable RET debug-on-error RET t RET

-- Juliusz
Sebastian Hofer
2014-05-07 21:49:38 UTC
Permalink
On Wed, May 7, 2014 at 11:18 PM, Juliusz Chroboczek <
Post by Juliusz Chroboczek
Post by Sebastian Hofer
Can someone advise me how to diagnose the problem?
M-x set-variable RET debug-on-error RET t RET
-- Juliusz
Hi Juliusz,

thanks for your reply. I actually already did, but replied only to Erik
instead of the whole list. The backtrace is attached but doesn't help me
much. I'm happy for any hints.

Best
Sebastian
Erik Hetzner
2014-05-08 01:36:51 UTC
Permalink
Hi all,

Sebastian sent me a backtrace off-list. I have seen the same error
with a maildir folder used as an fcc and the el-get package. The
problem seems to be with this line in elmo-map.el:

(luna-define-method elmo-folder-next-message-number ((folder elmo-map-folder))
(1+ (elmo-location-map-max-number folder)))

(elmo-location-map-max-number folder) is returning something other
than an integer (probably nil) which throws an error.

Unfortunately I can’t reproduce this with my current, non el-get
install of WL. Maybe somebody else can figure out if this is an el-get
problem or a WL problem.

best, Eri
Post by Juliusz Chroboczek
Post by Sebastian Hofer
Can someone advise me how to diagnose the problem?
M-x set-variable RET debug-on-error RET t RET
-- Juliusz
--
Sent from my free software system <http://fsf.org/>.
David Maus
2014-05-09 04:37:42 UTC
Permalink
Hi,

At Wed, 7 May 2014 16:13:46 +0200,
[1 <text/plain; UTF-8 (7bit)>]
[2 <text/html; UTF-8 (quoted-printable)>]
Dear all,
I recently updated wanderlust (+ apel/semi/flim) from el-get. I am
now getting an
Did you update WL after 2014-05-04 and where does your el-get get the
source?

This good be an error introduced by the merge of heimkehr and master
last sunday.

Best,
-- David
apply: Wrong type argument: number-or-marker-p, nil
error after sending an email. The email is sent correctly, but not moved to the Sent folder. Can someone advise me how to diagnose the problem?
TIA
Sebastian
Sebastian Hofer
2014-05-17 15:26:07 UTC
Permalink
Post by David Maus
Did you update WL after 2014-05-04 and where does your el-get get the
source?
This good be an error introduced by the merge of heimkehr and master
last sunday.
Best,
-- David
Hi David,

sorry for the long wait. I indeed updated after that date. el-get gets
wanderlust from github.com/wanderlust/wanderlust.git. Funny enough, the
error disappeared and reappeared in the last week without me doing anything
to the config. I'm very puzzled by this.
Also I'm irregularly getting the following error:
Error running timer `wl-biff-event-handler': (wrong-type-argument stringp
[elmo-maildir-folder [0 0 0 0 0 0 0] maildir ".univie/INBOX" "." nil nil
nil nil t nil t x-ctext nil nil nil "/home/sebhofer/Mail/univie/INBOX" nil
nil nil])
Don't know if this is connected or not.

Cheers
Sebastian
David Maus
2014-06-09 06:10:15 UTC
Permalink
Hi,

The error when sending messages should be fixed now.

https://github.com/wanderlust/wanderlust/commit/e9efa4eb6694d9760061a8d6530dc2789dd05e62

The problem was that WL did not make sure the target folder of an
Fcc-operation was actually opened.

I'll try to reproduce the problem with wl-biff; this looks similar.

Best,
-- David

At Sat, 17 May 2014 17:26:07 +0200,
[1 <text/plain; UTF-8 (7bit)>]
[2 <text/html; UTF-8 (quoted-printable)>]
Did you update WL after 2014-05-04 and where does your el-get get the
source?
This good be an error introduced by the merge of heimkehr and master
last sunday.
Best,
  -- David
Hi David,
sorry for the long wait. I indeed updated after that date. el-get gets wanderlust from github.com/wanderlust/wanderlust.git. Funny enough, the error disappeared and
reappeared in the last week without me doing anything to the config. I'm very puzzled by this.
Error running timer `wl-biff-event-handler': (wrong-type-argument stringp [elmo-maildir-folder [0 0 0 0 0 0 0] maildir ".univie/INBOX" "." nil nil nil nil t nil t
x-ctext nil nil nil "/home/sebhofer/Mail/univie/INBOX" nil nil nil])
Don't know if this is connected or not.
Cheers
Sebastian
Sebastian Hofer
2014-06-10 08:48:42 UTC
Permalink
Hi David,

thanks, it seems this fixed it! I can't be 100% sure though because the
problem was not always reproducible, it went away several times w/o me
changing anything.

Cheers
Sebastian
Post by David Maus
Hi,
The error when sending messages should be fixed now.
https://github.com/wanderlust/wanderlust/commit/e9efa4eb6694d9760061a8d6530dc2789dd05e62
The problem was that WL did not make sure the target folder of an
Fcc-operation was actually opened.
I'll try to reproduce the problem with wl-biff; this looks similar.
Best,
-- David
At Sat, 17 May 2014 17:26:07 +0200,
[1 <text/plain; UTF-8 (7bit)>]
[2 <text/html; UTF-8 (quoted-printable)>]
Did you update WL after 2014-05-04 and where does your el-get get the
source?
This good be an error introduced by the merge of heimkehr and master
last sunday.
Best,
-- David
Hi David,
sorry for the long wait. I indeed updated after that date. el-get gets
wanderlust from github.com/wanderlust/wanderlust.git. Funny enough, the
error disappeared and
reappeared in the last week without me doing anything to the config. I'm
very puzzled by this.
Error running timer `wl-biff-event-handler': (wrong-type-argument
stringp [elmo-maildir-folder [0 0 0 0 0 0 0] maildir ".univie/INBOX" "."
nil nil nil nil t nil t
x-ctext nil nil nil "/home/sebhofer/Mail/univie/INBOX" nil nil nil])
Don't know if this is connected or not.
Cheers
Sebastian
Daniel E. Doherty
2014-07-09 15:27:52 UTC
Permalink
All,

I got the same error using the following MELPA repository:

(add-to-list 'package-archives
'("e6h" . "http://www.e6h.org/packages/"))

which looks to be Erik's work. I'm using it because I got too much cross-talk
between using both el-get and melpa, and this gave me the chance to go all
melpa without losing Wanderlust.

Is the fix going to get propagated there as well?

At Tue, 10 Jun 2014 10:48:42 +0200,
[1 <text/plain; UTF-8 (7bit)>]
[2 <text/html; UTF-8 (quoted-printable)>]
Hi David,
thanks, it seems this fixed it! I can't be 100% sure though because the problem was not
always reproducible, it went away several times w/o me changing anything.
Cheers
Sebastian
Hi,
The error when sending messages should be fixed now.
https://github.com/wanderlust/wanderlust/commit/e9efa4eb6694d9760061a8d6530dc2789dd05e62
The problem was that WL did not make sure the target folder of an
Fcc-operation was actually opened.
I'll try to reproduce the problem with wl-biff; this looks similar.
Best,
 -- David
At Sat, 17 May 2014 17:26:07 +0200,
[1  <text/plain; UTF-8 (7bit)>]
[2  <text/html; UTF-8 (quoted-printable)>]
    Did you update WL after 2014-05-04 and where does your el-get get the
    source?
    This good be an error introduced by the merge of heimkehr and master
    last sunday.
    Best,
      -- David
Hi David,
sorry for the long wait. I indeed updated after that date. el-get gets wanderlust
from github.com/wanderlust/wanderlust.git. Funny enough, the error disappeared and
reappeared in the last week without me doing anything to the config. I'm very
puzzled by this.
Error running timer `wl-biff-event-handler': (wrong-type-argument stringp
[elmo-maildir-folder [0 0 0 0 0 0 0] maildir ".univie/INBOX" "." nil nil nil nil t nil
t
x-ctext nil nil nil "/home/sebhofer/Mail/univie/INBOX" nil nil nil])
Don't know if this is connected or not.
Cheers
Sebastian
Erik Hetzner
2014-07-09 16:16:41 UTC
Permalink
Hi Daniel,

Yes, I was waiting to update my package repo until some of the issues
that people were having with the changes settled down. It seems like
now is the time to upgrade the repo.

As an aside, I am having some trouble trying to figure out how to
clean up the APEL/SEMI/FLIM repo divergence (between the github and
git.chise.org repos, both of which were converted independently, and
slightly differently, from the old CVS repos, and which have diverged
somewhat since then). If anybody has any suggestions I would
appreciate them.

best, Erik

At Wed, 09 Jul 2014 10:27:52 -0500,
Post by Daniel E. Doherty
All,
(add-to-list 'package-archives
'("e6h" . "http://www.e6h.org/packages/"))
which looks to be Erik's work. I'm using it because I got too much cross-talk
between using both el-get and melpa, and this gave me the chance to go all
melpa without losing Wanderlust.
Is the fix going to get propagated there as well?
At Tue, 10 Jun 2014 10:48:42 +0200,
[1 <text/plain; UTF-8 (7bit)>]
[2 <text/html; UTF-8 (quoted-printable)>]
Hi David,
thanks, it seems this fixed it! I can't be 100% sure though because the problem was not
always reproducible, it went away several times w/o me changing anything.
Cheers
Sebastian
Hi,
The error when sending messages should be fixed now.
https://github.com/wanderlust/wanderlust/commit/e9efa4eb6694d9760061a8d6530dc2789dd05e62
The problem was that WL did not make sure the target folder of an
Fcc-operation was actually opened.
I'll try to reproduce the problem with wl-biff; this looks similar.
Best,
 -- David
At Sat, 17 May 2014 17:26:07 +0200,
[1  <text/plain; UTF-8 (7bit)>]
[2  <text/html; UTF-8 (quoted-printable)>]
    Did you update WL after 2014-05-04 and where does your el-get get the
    source?
    This good be an error introduced by the merge of heimkehr and master
    last sunday.
    Best,
      -- David
Hi David,
sorry for the long wait. I indeed updated after that date. el-get gets wanderlust
from github.com/wanderlust/wanderlust.git. Funny enough, the error disappeared and
reappeared in the last week without me doing anything to the config. I'm very
puzzled by this.
Error running timer `wl-biff-event-handler': (wrong-type-argument stringp
[elmo-maildir-folder [0 0 0 0 0 0 0] maildir ".univie/INBOX" "." nil nil nil nil t nil
t
x-ctext nil nil nil "/home/sebhofer/Mail/univie/INBOX" nil nil nil])
Don't know if this is connected or not.
Cheers
Sebastian
--
Sent from my free software system <http://fsf.org/>.
David Maus
2014-07-20 07:14:40 UTC
Permalink
Hi Erik,
Hi all,

At Wed, 09 Jul 2014 09:16:41 -0700,
Post by Erik Hetzner
As an aside, I am having some trouble trying to figure out how to
clean up the APEL/SEMI/FLIM repo divergence (between the github and
git.chise.org repos, both of which were converted independently, and
slightly differently, from the old CVS repos, and which have diverged
somewhat since then). If anybody has any suggestions I would
appreciate them.
The SEMI/FLIM/APEL-situation is really unfortunate. I came up with the
following options:

A. Make our versions of SEMI/APEL/FLIM part of the Wanderlust
A1. Add our repositories as git submodule

-or-

A2. Import the files into the WL repository
(Bonus: Import the files' history, too)

B. Publish our version with a distinct name

E.g. apel-wl, semi-wl, and flim-wl and make the wl package depend on
those packages.

I discarded:

C. Base the MELP-packages on chise.org and maintain a patch that
updates those version to our current version (similar to what
Debian does).

=> Too much effort

D. Base the MELPA on the Debian packages

=> Too much effort

And I am not sure about

E. Officially fork APEL/SEMI/FLIM

I think that an official fork is justified: The official development
of the packages has stalled, the real development happens in our
repositories.

Or, maybe is are combinations:

F1. Fork SEMI and remove all dependencies on APEL and FLIM
F2. Include SEMI into WL and remove the dependencies on APEL and FLIM

Best,
-- David
Post by Erik Hetzner
best, Erik
At Wed, 09 Jul 2014 10:27:52 -0500,
Post by Daniel E. Doherty
All,
(add-to-list 'package-archives
'("e6h" . "http://www.e6h.org/packages/"))
which looks to be Erik's work. I'm using it because I got too much cross-talk
between using both el-get and melpa, and this gave me the chance to go all
melpa without losing Wanderlust.
Is the fix going to get propagated there as well?
At Tue, 10 Jun 2014 10:48:42 +0200,
[1 <text/plain; UTF-8 (7bit)>]
[2 <text/html; UTF-8 (quoted-printable)>]
Hi David,
thanks, it seems this fixed it! I can't be 100% sure though because the problem was not
always reproducible, it went away several times w/o me changing anything.
Cheers
Sebastian
Hi,
The error when sending messages should be fixed now.
https://github.com/wanderlust/wanderlust/commit/e9efa4eb6694d9760061a8d6530dc2789dd05e62
The problem was that WL did not make sure the target folder of an
Fcc-operation was actually opened.
I'll try to reproduce the problem with wl-biff; this looks similar.
Best,
 -- David
At Sat, 17 May 2014 17:26:07 +0200,
[1  <text/plain; UTF-8 (7bit)>]
[2  <text/html; UTF-8 (quoted-printable)>]
    Did you update WL after 2014-05-04 and where does your el-get get the
    source?
    This good be an error introduced by the merge of heimkehr and master
    last sunday.
    Best,
      -- David
Hi David,
sorry for the long wait. I indeed updated after that date. el-get gets wanderlust
from github.com/wanderlust/wanderlust.git. Funny enough, the error disappeared and
reappeared in the last week without me doing anything to the config. I'm very
puzzled by this.
Error running timer `wl-biff-event-handler': (wrong-type-argument stringp
[elmo-maildir-folder [0 0 0 0 0 0 0] maildir ".univie/INBOX" "." nil nil nil nil t nil
t
x-ctext nil nil nil "/home/sebhofer/Mail/univie/INBOX" nil nil nil])
Don't know if this is connected or not.
Cheers
Sebastian
--
Sent from my free software system <http://fsf.org/>.
Loading...