0.71 Bug entering draft group

0.71 Bug entering draft group

[Lee Willis]

When trying to enter my draft group I get the following ...

Signaling: (wrong-type-argument symbolp "nil")
  mm-decode-coding-region(1 1 "nil")
  mail-decode-encoded-word-string("")
  gnus-get-newsgroup-headers()
  gnus-select-newsgroup("nndraft:drafts" nil nil)
  gnus-summary-read-group-1("nndraft:drafts" nil t nil nil nil)
  gnus-summary-read-group("nndraft:drafts" nil t nil nil nil nil)
  gnus-group-read-group(nil t)
  gnus-group-select-group(nil)
  gnus-topic-select-group(nil)
* call-interactively(gnus-topic-select-group)

Lee.
-- 
I was doing object-oriented assembly at 1 year old ...  
For some reason my mom insists on calling it "Playing with blocks"

Re: 0.71 Bug entering draft group

[Jack Vinson]

>>>>> "LW" == Lee Willis <lee@gbdirect.co.uk> writes:

LW> When trying to enter my draft group I get the following ...
LW> Signaling: (wrong-type-argument symbolp "nil")
[snip]

I am getting the same error in pgnus 0.71 on an ntEmacs 20.3.1.  

Is this similar to the error ntemacs users were getting back in the 0.50's
due to the coding system?  I can't quite remember.

The summary buffer is created, but Gnus never writes anything to it.  The
message in my draft folder is from 0.69, if that means anything.

-- 
Jack Vinson <jvinson@chevax.ecs.umass.edu>    http://www.cis.upenn.edu/~vinson/
Zippy: I will establish the first SHOPPING MALL in NUTLEY, New Jersey...

Re: 0.71 Bug entering draft group

[Lee Willis]

Jack Vinson <jvinson@chevax.ecs.umass.edu> writes:

> >>>>> "LW" == Lee Willis <lee@gbdirect.co.uk> writes:
> 
> LW> When trying to enter my draft group I get the following ...
> LW> Signaling: (wrong-type-argument symbolp "nil")
> [snip]
> 
> I am getting the same error in pgnus 0.71 on an ntEmacs 20.3.1.  
> 
> The summary buffer is created, but Gnus never writes anything to it.  The
> message in my draft folder is from 0.69, if that means anything.

It doesn't work even if the messages are from 0.71 ...

Lee.
-- 
I was doing object-oriented assembly at 1 year old ...  
For some reason my mom insists on calling it "Playing with blocks"

Re: 0.71 Bug entering draft group

[Lars Magne Ingebrigtsen]

Lee Willis <lee@gbdirect.co.uk> writes:

> Signaling: (wrong-type-argument symbolp "nil")
>   mm-decode-coding-region(1 1 "nil")

Hm.  Have you set mail-parse-charset to the string "nil"?

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: 0.71 Bug entering draft group

[Lee Willis]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Lee Willis <lee@gbdirect.co.uk> writes:
> 
> > Signaling: (wrong-type-argument symbolp "nil")
> >   mm-decode-coding-region(1 1 "nil")
> 
> Hm.  Have you set mail-parse-charset to the string "nil"?

Erm well C-h v mail-parse-charset gives

mail-parse-charset's value is 
nil

Which seems to be correct according to the variables description.  This
kind of brings up something I've been meaning to ask for a while, when
pgnus undergoes radical changes where do we find out what we have to do
to keep our .gnus working? Changelog only? or is there something I'm
missing?

Lee.
-- 
I was doing object-oriented assembly at 1 year old ...  
For some reason my mom insists on calling it "Playing with blocks"

Re: 0.71 Bug entering draft group

[Jack Vinson]

>>>>> "LMI" == Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

LMI> Lee Willis <lee@gbdirect.co.uk> writes:
>> Signaling: (wrong-type-argument symbolp "nil")
>> mm-decode-coding-region(1 1 "nil")

LMI> Hm.  Have you set mail-parse-charset to the string "nil"?

My mail-parse-charset is nil.  Also note the documentation on this
variable (emacs 20.3).  Does xemacs expect the user to set this variable?

     mail-parse-charset's value is 
     nil

     Documentation:
     Default charset used by low-level libraries.
     This variable should never be set.  Instead, it should be bound by
     functions that wish to call mail-parse functions and let them know
     what the desired charset is to be.

-- 
Jack Vinson <jvinson@chevax.ecs.umass.edu>    http://www.cis.upenn.edu/~vinson/
Zippy: C'MON, everybody!!  I've flown in LESLIE GORE and two dozen KOSHER
 BUTCHERS!  They'll be doing intricate MILITARY MANEUVERS to the
 soundtrack from "OKLAHOMA"!!

Re: 0.71 Bug entering draft group

[Lee Willis]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Lee Willis <lee@gbdirect.co.uk> writes:
> 
> > Signaling: (wrong-type-argument symbolp "nil")
> >   mm-decode-coding-region(1 1 "nil")
> 
> Hm.  Have you set mail-parse-charset to the string "nil"?

Hmm, just as a side note, where is mm-decode-coding-region actually
defined because I can't actually find it. Hmm. Mind I didn't look too
far, but a grep on *.el in pgnus-0.71/lisp didn't reveal any defun lines
...

Lee. 
-- 
I was doing object-oriented assembly at 1 year old ...  
For some reason my mom insists on calling it "Playing with blocks"

Re: 0.71 Bug entering draft group

[Justin Sheehy]

Lee Willis <lee@gbdirect.co.uk> writes:

> when pgnus undergoes radical changes where do we find out what we
> have to do to keep our .gnus working?

I haven't had to make any significant changes to my startup files in
order to keep pgnus working.  I have made a couple to take advantage
of new features, but that's it.

If one makes heavy changes to their ~/.gnus.el and also uses alpha
software, they should be prepared to read the code to find changes if
need be.

-- 
Justin Sheehy

In a cloud bones of steel.
  

Re: 0.71 Bug entering draft group

[Lee Willis]

Justin Sheehy <justin@linus.mitre.org> writes:

> Lee Willis <lee@gbdirect.co.uk> writes:
> 
> > when pgnus undergoes radical changes where do we find out what we
> > have to do to keep our .gnus working?
> 
> If one makes heavy changes to their ~/.gnus.el and also uses alpha
> software, they should be prepared to read the code to find changes if
> need be.

Oh I quite agree, and that's generally what I've done, and obviously
asked for help when I became stuck :) I was just wondering if there was
anywhere other than the Changelog and the conversations on here that I
should have been looking.  It wasn't a complaint about lack of support
or anything like that, merely a question in passing ...

Lee.
-- 
I was doing object-oriented assembly at 1 year old ...  
For some reason my mom insists on calling it "Playing with blocks"

Re: 0.71 Bug entering draft group

[Jack Vinson]

>>>>> "JV" == Jack Vinson <jvinson@chevax.ecs.umass.edu> writes:

>>>>> "LMI" == Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
LMI> Lee Willis <lee@gbdirect.co.uk> writes:
>>> Signaling: (wrong-type-argument symbolp "nil")
>>> mm-decode-coding-region(1 1 "nil")

LMI> Hm.  Have you set mail-parse-charset to the string "nil"?

JV> My mail-parse-charset is nil.  Also note the documentation on this
JV> variable (emacs 20.3).  Does xemacs expect the user to set this variable?

I have found the culprit.  At some point down the list of things that
happen before mm-decode-coding-region is called, gnus asks for information
about the group with gnus-get-info.

When this function is run on the "nndraft:drafts" group, here is the info I
get back:
("nndraft:drafts" 1
 ((1 . 4))
 nil
 (nndraft "")
 ((charset . "nil")
  (gnus-dummy
   (gnus-draft-mode))))

Note that charset is "nil," not nil or empty.  My other groups have no
charset in their group description.

Now that I understand this the solution is simple: go to the nndraft group,
do 'G E' and remove the charset entry from the group parameters.  Viola!
Problem solved.

-- 
Jack Vinson <jvinson@chevax.ecs.umass.edu>    http://www.cis.upenn.edu/~vinson/
Zippy: I'm thinking about DIGITAL READ-OUT systems and
 computer-generated IMAGE FORMATIONS..

Re: 0.71 Bug entering draft group

[Lars Magne Ingebrigtsen]

Lee Willis <lee@gbdirect.co.uk> writes:

> Hmm, just as a side note, where is mm-decode-coding-region actually
> defined because I can't actually find it. Hmm. Mind I didn't look too
> far, but a grep on *.el in pgnus-0.71/lisp didn't reveal any defun lines
> ...

It's defined by the `eval-and-compile' form in mm-util.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: 0.71 Bug entering draft group

[Lee Willis]

Jack Vinson <jvinson@chevax.ecs.umass.edu> writes:

> Now that I understand this the solution is simple: go to the nndraft group,
> do 'G E' and remove the charset entry from the group parameters.  Viola!
> Problem solved.

Yep, works for me too ...

Lee
-- 
I was doing object-oriented assembly at 1 year old ...  
For some reason my mom insists on calling it "Playing with blocks"