headers
Brent Welch | 11 Sep 2002 21:17
X-Face
Favicon

Re: exmh bug?

Hmm - I'm cc'ing the exmh-workers list, because I really don't know
much about the various PGP interfaces.  I think there has been some
talk about "issues" with the latest version of gpg.

>>>Hacksaw said:
 > version 2.5 08/15/2002
 > Linux habitrail.home.fools-errant.com 2.4.7-10smp #1 SMP Thu Sep 6 
17:09:31
 > EDT 2001 i686 unknown
 > Tk 8.3 Tcl 8.3
 > 
 > It's not clear to me this is a bug with exmh per se, but it's something 
that

 > manifests through exmh, so I figured asking you might help me track it 
down.
 > 
 > When I receive a gpg encrypted message, and it asks me for a passphrase, 
it
 > first tries to ask me via the tty under which exmh is running. It tells 
me m
     y 
 > passphrase is incorrect every time, at which point exmh offers me the 
line i
     n 
 > the message about decrypting. I click the line and it offers me the 
dialog
 > box, and tells me the passphrase is correct, and shows me the decrypted 
 > message.
 >
(Continue reading)

Permalink | Reply | 3 comments |
headers
kchrist | 11 Sep 2002 21:48

Re: exmh bug?


The way to debug something like this is to turn on the log (Preferences 
-> Hacking Support -> Debug log enabled) and track the gpg commands 
being issued and the responses.

FWIW, using exmh 2.5 and gpg 1.0.7 I do not have problems sending 
encrypted messages to myself. When I select the message, an xterm 
window pops up asking for the passphrase. I don't recall exmh ever 
asking me for the passphrase from the tty that started exmh nor from a 
dialog box. (But then again, I'm not a heavy gpg user).

Kevin

In message <200209111917.PAA02912 <at> blackcomb.panasas.com>, Brent Welch writes:
> Hmm - I'm cc'ing the exmh-workers list, because I really don't know
> much about the various PGP interfaces.  I think there has been some
> talk about "issues" with the latest version of gpg.
> 
> >>>Hacksaw said:
>  > version 2.5 08/15/2002
>  > Linux habitrail.home.fools-errant.com 2.4.7-10smp #1 SMP Thu Sep 6 17:09:31
>  > EDT 2001 i686 unknown
>  > Tk 8.3 Tcl 8.3
>  > 
>  > It's not clear to me this is a bug with exmh per se, but it's something 
>  > that manifests through exmh, so I figured asking you might help me track
>  > it down.
>  > 
>  > When I receive a gpg encrypted message, and it asks me for a passphrase, 
>  > it first tries to ask me via the tty under which exmh is running. It
(Continue reading)

Permalink | Reply | 0 comments |
headers
Valdis.Kletnieks | 11 Sep 2002 21:52
X-Face
Picon
Favicon

Re: exmh bug?

On Wed, 11 Sep 2002 12:17:33 PDT, Brent Welch said:

> >>>Hacksaw said:

>  > When I receive a gpg encrypted message, and it asks me for a passphrase, 
> it
>  > first tries to ask me via the tty under which exmh is running. It tells 
>
Hmm.. I've seen the *opposite* issue - if I go to *SEND* a signed message,
sometimes Exmh will put up the dialog box, but fail to set keyboard focus
there, so no passphrase can be entered.  Of course, hitting 'return' doesnt
work so you need to click the 'OK' box, at which point it finds that the
passphrase that wasn't entered doesn't work, and asks again, this time with
proper focus set.

I suspect some variable/codepath is getting hosed for the focus, or possibly
some borkedness with --no-tty and/or --status-fd flags to gnupg.

/Valdis
Permalink | Reply | 0 comments |
headers
Kevin Cosgrove | 11 Sep 2002 21:53

Re: exmh bug?


Gotta wonder what the GPG config stuff in ~/.exmh/exmh-defaults looks 
like.  Also gotta wonder what the message headers in the offending 
message are saying to nmh/exmh.  My set-up works perfectly.  That is, 
I get a pop-up window to enter my passphrase into, and when I type it 
correctly, the message display changes from a prompt to click to 
decrypt to the message content.

TTFN....

On 11 September 2002 at 12:17, Brent Welch <welch <at> panasas.com> wrote:

Hmm - I'm cc'ing the exmh-workers list, because I really don't know
much about the various PGP interfaces.  I think there has been some
talk about "issues" with the latest version of gpg.

>>>Hacksaw said:
 > version 2.5 08/15/2002
 > Linux habitrail.home.fools-errant.com 2.4.7-10smp #1 SMP Thu Sep 6 17:09:31
 > EDT 2001 i686 unknown
 > Tk 8.3 Tcl 8.3
 > 
 > It's not clear to me this is a bug with exmh per se, but it's
 > something that manifests through exmh, so I figured asking you
 > might help me track it down.
 >
 > When I receive a gpg encrypted message, and it asks me for a
 > passphrase, it first tries to ask me via the tty under which
 > exmh is running. It tells me my passphrase is incorrect every
 > time, at which point exmh offers me the line in the message
(Continue reading)

Permalink | Reply | 0 comments |
headers
Kevin Cosgrove | 11 Sep 2002 22:26

Re: exmh bug?


Here's a message that works fine for me:

X-Mailer: exmh version 2.5 07/13/2001 with nmh-1.0.4
To: kevinc <at> doink.com
Subject: gpg test
Mime-Version: 1.0
Content-Type: text/x-pgp; format=mime; x-action=encrypt;
	x-recipients="D165C0CC, D165C0CC"
Content-Transfer-Encoding: 7bit
Date: Wed, 11 Sep 2002 13:18:58 -0700
From: Kevin Cosgrove <kevinc <at> dOink.COM>
Message-Id: <20020911201900.492892FEAB <at> joseph.doink.com>
X-Spam-Status: No, hits=-1.3 required=5.0 tests=AWL version=2.11

-----BEGIN PGP MESSAGE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Exmh version 2.5 10/15/1999

blah, blah, blah

-----END PGP MESSAGE-----

It could be that your content type of text/plain is messing you up.
Your config stuff looks substantially the same as mine.

Cheers....

On 11 September 2002 at 16:11, Hacksaw <hacksaw <at> hacksaw.org> wrote:
(Continue reading)

Permalink | Reply | 0 comments |
headers
Chris Garrigues | 21 Sep 2002 20:46

Re: Exmh && speed

> From:  Chris Garrigues <cwg-exmh <at> DeepEddy.Com>
> Date:  Wed, 28 Aug 2002 11:07:01 -0500
>
> Here's a fix that I think will make a real difference.
> 
> Ftoc_ShowSequences needs to be able to be called with an optional list of msgids 
> to update and if it's called that way it only removes or adds tags for those 
> messages.  Then in places like MsgChange, we only update the messages which have
> changed.
> 
> Also, a separate Ftoc_ShowSequence function which only updates the display of 
> one sequence should be written which also takes an optional list of msgids.
>   
> In a place like MsgChange, it would only need to update the cur sequence.
> 
> If nobody else gets to it, I'll do this when I get back.

I just checked this in.  Let me know if it helps.

Chris

--

-- 
Chris Garrigues                 http://www.DeepEddy.Com/~cwg/
virCIO                          http://www.virCIO.Com
716 Congress, Suite 200
Austin, TX  78701		+1 512 374 0500

  World War III:  The Wrong-Doers Vs. the Evil-Doers.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Scott Lipcon | 23 Sep 2002 04:58

traceback in new exmh

I just updated to the latest CVS - I had been running a build from June.  
Hitting the Flist button gives the following traceback:

syntax error in expression "int(17+1+(222-)*(19-17-2)/(224-))"
    while executing
"expr int($minLine+1+($msgid-$minMsg)*($maxLine-$minLine-2)/($maxMsg-$minMsg))"
    (procedure "Ftoc_FindMsg" line 57)
    invoked from within
"Ftoc_FindMsg $i"
    (procedure "Ftoc_ShowSequences" line 16)
    invoked from within
"Ftoc_ShowSequences $F"
    (procedure "ScanFolder" line 81)
    invoked from within
"ScanFolder inbox 0"
    invoked from within
"time [list ScanFolder $F $adjustDisplay"
    (procedure "Scan_Folder" line 2)
    invoked from within
"Scan_Folder $exmh(folder) $ftoc(showNew)"
    (procedure "Inc_PresortFinish" line 7)
    invoked from within
"Inc_PresortFinish"
    invoked from within
".fops.flist invoke"
    ("uplevel" body line 1)
    invoked from within
"uplevel #0 [list $w invoke]"
    (procedure "tkButtonUp" line 7)
    invoked from within
(Continue reading)

Permalink | Reply | 6 comments |
headers
Chris Garrigues | 23 Sep 2002 16:41

Re: traceback in new exmh

> From:  Scott Lipcon <slipcon <at> mercea.net>
> Date:  Sun, 22 Sep 2002 22:58:16 -0400
>
> I just updated to the latest CVS - I had been running a build from June.  
> Hitting the Flist button gives the following traceback:
> 
> syntax error in expression "int(17+1+(222-)*(19-17-2)/(224-))"
>     while executing
> "expr int($minLine+1+($msgid-$minMsg)*($maxLine-$minLine-2)/($maxMsg-$minMs
> g))"
>     (procedure "Ftoc_FindMsg" line 57)
>     invoked from within
> "Ftoc_FindMsg $i"
>     (procedure "Ftoc_ShowSequences" line 16)
>     invoked from within
> "Ftoc_ShowSequences $F"
>     (procedure "ScanFolder" line 81)
>     invoked from within
> "ScanFolder inbox 0"
>     invoked from within
> "time [list ScanFolder $F $adjustDisplay"
>     (procedure "Scan_Folder" line 2)
>     invoked from within
> "Scan_Folder $exmh(folder) $ftoc(showNew)"
>     (procedure "Inc_PresortFinish" line 7)
>     invoked from within
> "Inc_PresortFinish"
>     invoked from within
> ".fops.flist invoke"
>     ("uplevel" body line 1)
(Continue reading)

Permalink | Reply | 5 comments |
headers
Scott Lipcon | 23 Sep 2002 16:54

Re: traceback in new exmh

Oops - turns out i wasn't careful installing it, so the exmh(library)
variable was pointing at my old installation.  I'm surprised it worked
as well as it did.

In any case, I removed the old library directory, and edited exmh,
exmh-bg, exmh-strip, etc to point at the right one.  Now I'm getting
another traceback on startup:

can't read "mhPriv(pubseq,family,cur)": no such variable
    while executing
"MhSeq $folder $seq add $mhPriv(pubseq,$folder,$seq) [MhSeqExpand
$folder $msgids"
    (procedure "MhReadSeqs" line 73)
    invoked from within
"MhReadSeqs $folder seqs"
    (procedure "Mh_Sequences" line 2)
    invoked from within
"Mh_Sequences $folder"
    (procedure "Flist_UnseenUpdate" line 4)
    invoked from within
"Flist_UnseenUpdate $folder"
    (procedure "FolderChange" line 51)
    invoked from within
"FolderChange family {Msg_Show cur}"
    invoked from within
"time [list  FolderChange $folder $msgShowProc"
    (procedure "Folder_Change" line 3)
    invoked from within
"Folder_Change $exmh(folder)"
    (procedure "Exmh" line 101)
(Continue reading)

Permalink | Reply | 3 comments |
headers
Chris Garrigues | 23 Sep 2002 16:58

Re: traceback in new exmh

> From:  Chris Garrigues <cwg-bcc <at> DeepEddy.Com>
> Date:  Mon, 23 Sep 2002 09:41:22 -0500
>
> > From:  Scott Lipcon <slipcon <at> mercea.net>
> > Date:  Sun, 22 Sep 2002 22:58:16 -0400
> >
> > I just updated to the latest CVS - I had been running a build from June. 
>  
> > Hitting the Flist button gives the following traceback:
> > 
> > syntax error in expression "int(17+1+(222-)*(19-17-2)/(224-))"
> >     while executing
> > "expr int($minLine+1+($msgid-$minMsg)*($maxLine-$minLine-2)/($maxMsg-$minMs
> > g))"
> >     (procedure "Ftoc_FindMsg" line 57)
> >     invoked from within
> > "Ftoc_FindMsg $i"
> >     (procedure "Ftoc_ShowSequences" line 16)
> >     invoked from within
> > "Ftoc_ShowSequences $F"
> >     (procedure "ScanFolder" line 81)
> >     invoked from within
> > "ScanFolder inbox 0"
> >     invoked from within
> > "time [list ScanFolder $F $adjustDisplay"
> >     (procedure "Scan_Folder" line 2)
> >     invoked from within
> > "Scan_Folder $exmh(folder) $ftoc(showNew)"
> >     (procedure "Inc_PresortFinish" line 7)
> >     invoked from within
(Continue reading)

Permalink | Reply | 0 comments |

Gmane