Strange nov line (N)

Strange nov line (N)

[Shigeki Uno]

Good day.

 I lost Gnus-5.8.4 cvs version because of OS trouble a week ago. :-<
So I have been trying to get Gnus-5.8.5 to work but it doesn't.
 
 Fetching news groups via internet and retrieving mails from local
mail spool(/var/mail/login-name) went well but problem occured. 

 When it comes to hit "space key" to enter summary mode, "Strange nov
line (N)" message shows up in mini buffer and each news groups and
mail groups turn to "0" in articles count. And I can't enter into
summary mode. Of cource I can't read both news and mails... :-( 

This *annoys* me.

What is the reason of this phenomenon ? How do I fix it ?

Gnus-5.8.5 cvs version as of April.22. Maybe.
gnus-secondary-select-methods is nnml.
(Of cource, I did "M-x nnml-generate-nov-databases" at first.)

Thanks in advance.

---
Shigeki Uno <mailto:shigeki@mediawars.ne.jp>

Re: Strange nov line (N)

[Shenghuo ZHU]

>>>>> "Shigeki" == Shigeki Uno <shigeki@mediawars.ne.jp> writes:

Shigeki> Good day.

Shigeki>  When it comes to hit "space key" to enter summary mode,
Shigeki> "Strange nov line (N)" message shows up in mini buffer and
Shigeki> each news groups and mail groups turn to "0" in articles
Shigeki> count. And I can't enter into summary mode. Of cource I can't
Shigeki> read both news and mails... :-(

Shigeki> This *annoys* me.

Shigeki> What is the reason of this phenomenon ? How do I fix it ?

This might be coding-system-related problem. Which version of Emacs
are you using? Is MULE enable or not?  Do your .overview files contain
^M or other non-ASCII characters?

-- 
Shenghuo

Re: Strange nov line (N)

[Shigeki Uno]

Shenghuo ZHU <zsh@cs.rochester.edu> wrote
in <5baeikfuj4.fsf@heart.cs.rochester.edu>

> >>>>> "Shigeki" == Shigeki Uno <shigeki@mediawars.ne.jp> writes:
> 
> Shigeki> Good day.
> 
> Shigeki>  When it comes to hit "space key" to enter summary mode,
> Shigeki> "Strange nov line (N)" message shows up in mini buffer and
> Shigeki> each news groups and mail groups turn to "0" in articles
> Shigeki> count. And I can't enter into summary mode. Of cource I can't
> Shigeki> read both news and mails... :-(
> 
> Shigeki> This *annoys* me.
> 
> Shigeki> What is the reason of this phenomenon ? How do I fix it ?
> 
> This might be coding-system-related problem. Which version of Emacs
> are you using? Is MULE enable or not?  Do your .overview files contain
> ^M or other non-ASCII characters?

Emacs 20.3 / Mule 4.0. And MULE is ,of cource, enable. Do I need to upgrade to
Emacs 20.6 ? I used to be a user of 20.6. 

"iso-2022-jp" exists but Gnus worked fine before even if that existed. 
I also checked other mail groups which was written in english only but result 
is the same...  My God !
 
---
Shigeki Uno

Re: Strange nov line (N)

[Shenghuo ZHU]

>>>>> "Shigeki" == Shigeki Uno <shigeki@mediawars.ne.jp> writes:

Shigeki> Emacs 20.3 / Mule 4.0. And MULE is ,of cource, enable. Do I
Shigeki> need to upgrade to Emacs 20.6 ? I used to be a user of 20.6.

Shigeki> "iso-2022-jp" exists but Gnus worked fine before even if that
Shigeki> existed.  I also checked other mail groups which was written
Shigeki> in english only but result is the same...  My God !
 
Could you eval (load-library "gnus-sum.el") and 
(setq debug-on-signal t), then try to enter a group, 
and send the backtrace?

-- 
Shenghuo

Re: Strange nov line (N)

[Shigeki Uno]

Shenghuo ZHU <zsh@cs.rochester.edu> wrote
in <5bhfcs4h1k.fsf@whiskey.cs.rochester.edu>

> Could you eval (load-library "gnus-sum.el") and 
> (setq debug-on-signal t), then try to enter a group, 
> and send the backtrace?

Following is a sample I got. 
--- start of backtrace ---

Signaling: (coding-system-error iso-2022-jp-2)
  mm-decode-coding-region(1 45 iso-2022-jp-2)
  mail-decode-encoded-word-string("[netbsd,05423] Re: jp.netbsd.org still down?")
  gnus-get-newsgroup-headers()
  gnus-select-newsgroup("nnml:ML.netbsd-j" nil nil)
  gnus-summary-read-group-1("nnml:ML.netbsd-j" nil nil nil nil nil)
  gnus-summary-read-group("nnml:ML.netbsd-j" nil nil nil nil nil nil)
  gnus-group-read-group(nil)
* call-interactively(gnus-group-read-group)

--- end of backtrace ---

(I also can't enter groups written in english only...)

What's wrong ? 

---
Shigeki Uno

Re: Strange nov line (N)

[Shenghuo ZHU]

>>>>> "Shigeki" == Shigeki Uno <shigeki@mediawars.ne.jp> writes:

Shigeki> Following is a sample I got. 
Shigeki> --- start of backtrace ---

Shigeki> Signaling: (coding-system-error iso-2022-jp-2)
Shigeki>   mm-decode-coding-region(1 45 iso-2022-jp-2)
Shigeki>   mail-decode-encoded-word-string("[netbsd,05423] Re: jp.netbsd.org still down?")
Shigeki>   gnus-get-newsgroup-headers()
Shigeki>   gnus-select-newsgroup("nnml:ML.netbsd-j" nil nil)
Shigeki>   gnus-summary-read-group-1("nnml:ML.netbsd-j" nil nil nil nil nil)
Shigeki>   gnus-summary-read-group("nnml:ML.netbsd-j" nil nil nil nil nil nil)
Shigeki>   gnus-group-read-group(nil)
Shigeki> * call-interactively(gnus-group-read-group)

Shigeki> --- end of backtrace ---

On your system, it seems that iso-2022-jp-2 is not a coding system. To
make sure this, could you eval (mm-coding-system-p 'iso-2022-jp-2)?

If it is nil, I think you use a different version of mule as I do.

Shigeki> (I also can't enter groups written in english only...)
Shigeki> What's wrong ? 

What are those groups? Are groups starting with jp or japan?

-- 
Shenghuo

Re: Strange nov line (N)

[Katsumi Yamaoka]

>>>>> In <20000426.062925.68564712.shigeki@mediawars.ne.jp> 
>>>>>	Shigeki Uno <shigeki@mediawars.ne.jp> wrote:

Uno> X-Mailer: Mew version 1.95b34 on Emacs 20.3 / Mule 4.0 (HANANOEN)

Uno> Signaling: (coding-system-error iso-2022-jp-2)
Uno>   mm-decode-coding-region(1 45 iso-2022-jp-2)

This coding system is appeared in Emacs 20.4 and later.  It is better
to use recent Emacsen because people say Emacs 20.3 is buggy.
I have no confidence but the following code might work.

(define-coding-system-alias 'iso-2022-jp-2 'iso-2022-7bit-ss2)
-- 
Katsumi Yamaoka <yamaoka@jpl.org>

Solved. (was: Re: Strange nov line (N))

[Shigeki Uno]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

> >>>>> In <20000426.062925.68564712.shigeki@mediawars.ne.jp> 
> >>>>>	Shigeki Uno <shigeki@mediawars.ne.jp> wrote:
> 
> Uno> X-Mailer: Mew version 1.95b34 on Emacs 20.3 / Mule 4.0 (HANANOEN)
> 
> Uno> Signaling: (coding-system-error iso-2022-jp-2)
> Uno>   mm-decode-coding-region(1 45 iso-2022-jp-2)
> 
> This coding system is appeared in Emacs 20.4 and later.  It is better
> to use recent Emacsen because people say Emacs 20.3 is buggy.
> I have no confidence but the following code might work.
> 
> (define-coding-system-alias 'iso-2022-jp-2 'iso-2022-7bit-ss2)

This lisp code makes gnus-5.8.5 cvs version on emacs-20.3 enable to
enter summary buffers.

I deeply appreciate your help.

Thank you.

---
Shigeki Uno  

Re: Strange nov line (N)

[Shigeki Uno]

Shenghuo ZHU <zsh@cs.rochester.edu> writes:

[...]
 
> On your system, it seems that iso-2022-jp-2 is not a coding system. To
> make sure this, could you eval (mm-coding-system-p 'iso-2022-jp-2)?
> 
> If it is nil, I think you use a different version of mule as I do.

So!

> Shigeki> (I also can't enter groups written in english only...)
> Shigeki> What's wrong ? 
> 
> What are those groups? Are groups starting with jp or japan?

Not news groups. These're mail groups. Thanks to Yamaoka-san, The problem
which annoy me has just solved !  I'm saved. 

> Shenghuo

Thanks for your help.

---
Shigeki Uno