bug#1131: Fwd: [Aquamacs-bugs] problem with comint

bug#1131: Fwd: [Aquamacs-bugs] problem with comint

[David Reitter]

[-- Attachment #1: Type: text/plain, Size: 6159 bytes --]

See below.  I've reproduced this once with a current 22 branch CVS  
build (Carbon), as well as with 22.1 (with -nw), but can't reproduce  
today, so no backtrace.

According to Eric, it appears to still be an issue in Emacs 23  
("Emacs.app"), and there seems to be a very long (15 minute) timeout.



Begin forwarded message:

> From: "Eric Tytell" <tytell@umd.edu>
> Date: 8 October 2008 17:37:26 EDT
> To: aquamacs-bugs@aquamacs.org
> Subject: [Aquamacs-bugs] problem with comint
>
> Enter your bug report here.
>
> I had been using Aquamacs with Matlab, using the Matlab EmacsLink  
> interface,
> which allows debugging of Matlab scripts in Emacs.  Up until  
> yesterday, the
> interface was working fine.  Today, it crashes Aquamacs.  I've  
> tracked down
> the problem to this line (which seems to be unrelated to Matlab):
>
> (make-comint "MATLAB EEI" (cons (system-name) 5600))
>
> which (according to the docs for make-comint) is supposed to open a  
> TCP
> connection to the local computer on port 5600.  This displays "loading
> comint..." in the minibuffer, then the mouse cursor turns into the  
> spinning
> wheel and Aquamacs stops responding.  I have to kill it using "Force  
> Quit".
> I'm at a loss as to how to debug this further.  It's probably  
> something
> related to my system, but I don't know the details of how Aquamacs  
> handles
> TCP connections to figure anything more out.
>
> Any thoughts?
> Cheers,
> Eric Tytell
>
>
> In GNU Emacs 22.3.2 (i386-apple-darwin9.5.0, Carbon Version 1.6.0)
> of 2008-09-21 on plume.sr.unh.edu - Aquamacs Distribution 1.5
> Windowing system distributor `Apple Inc.', version 10.5.5
> configured using `configure  '--without-x' '--prefix=/usr/local''
>
> Important settings:
>  value of $LC_ALL: nil
>  value of $LC_COLLATE: nil
>  value of $LC_CTYPE: nil
>  value of $LC_MESSAGES: nil
>  value of $LC_MONETARY: nil
>  value of $LC_NUMERIC: nil
>  value of $LC_TIME: nil
>  value of $LANG: nil
>  locale-coding-system: iso-8859-1
>  default-enable-multibyte-characters: t
>
> Major mode: Text
>
> Minor modes in effect:
>  savehist-mode: t
>  smart-frame-positioning-mode: t
>  global-visual-line-mode: t
>  visual-line-mode: t
>  recentf-mode: t
>  encoded-kbd-mode: t
>  osx-key-mode: t
>  tabbar-mwheel-mode: t
>  tabbar-mode: t
>  show-paren-mode: t
>  delete-selection-mode: t
>  pc-selection-mode: t
>  cua-mode: t
>  tooltip-mode: t
>  tool-bar-mode: t
>  mouse-wheel-mode: t
>  menu-bar-mode: t
>  file-name-shadow-mode: t
>  global-font-lock-mode: t
>  font-lock-mode: t
>  blink-cursor-mode: t
>  unify-8859-on-encoding-mode: t
>  utf-translate-cjk-mode: t
>  auto-compression-mode: t
>  column-number-mode: t
>  line-number-mode: t
>  transient-mark-mode: t
>
> Recent input:
> <menu-bar> <help-menu> <bug-diagnosis> <start-vanilla-aquamacs>
> <down-mouse-1> <mouse-movement> <mouse-movement> <mouse-movement>
> <mouse-movement> <drag-mouse-1> A-c <menu-bar> <help-menu>
> <bug-diagnosis> <send-emacs-bug-report> <menu-bar>
> <help-menu> <bug-diagnosis> <send-emacs-bug-report>
> <down-mouse-1> <mouse-1> <down-mouse-1> <mouse-1> C-h
> f m a k e - c o <tab> m <tab> i n <tab> <return> C-x
> C-f C-g C-h f o p e n <tab> n e t <tab> <return> <down-mouse-1>
> <mouse-1> C-x b C-g C-h f m a k e - c o <tab> m <tab>
> i <tab> <return> <down> <down> <down> <down> <down>
> <down> <down> <down> <down> <up> <up> <down> <down>
> <down> <switch-frame> <menu-bar> <help-menu> <bug-diagnosis>
> <send-emacs-bug-report>
>
> Recent messages:
> Sending...
> Loading mailclient...done
> Loading browse-url...done
> Sending...done
> Truncate long lines disabled
> Mark set [2 times]
> Sending...done
> Making completion list... [2 times]
> Quit [2 times]
> Making completion list... [2 times]
>
> Command line: (/Applications/Aquamacs Emacs.app/Contents/MacOS/ 
> Aquamacs
> Emacs)
>
> PATH:
> /opt/local/bin:/opt/local/sbin:/Library/Frameworks/Python.framework/ 
> Versions/Current/bin:/Users/eric/bin:/usr/bin:/bin:/usr/sbin:/sbin:/ 
> usr/local/bin:/usr/X11/bin:/usr/bin:/bin:/usr/sbin:/sbin
>
> exec-path: (/opt/local/bin /opt/local/sbin
> /Library/Frameworks/Python.framework/Versions/Current/bin /Users/ 
> eric/bin
> /usr/bin /bin /usr/sbin /sbin /usr/local/bin /usr/X11/bin /usr/bin / 
> bin
> /usr/sbin /sbin /Applications/Aquamacs Emacs.app/Contents/MacOS/ 
> libexec
> /Applications/Aquamacs Emacs.app/Contents/MacOS/bin)




Begin forwarded message:

> From: "Eric Tytell" <tytell@umd.edu>
> Date: 9 October 2008 10:06:51 EDT
> To: "aquamacs-bugs@aquamacs.org Bugs" <aquamacs-bugs@aquamacs.org>
> Subject: Re: [Aquamacs-bugs] problem with comint
>
> Hi David -
>  In the documentation on my system for make-comint, it states that  
> it can
> take two types of "program" arguments: either an executable program,  
> or a
> cons pair, like the one I passed, with the form (HOST . SERVICE),  
> which will
> open a TCP connection via open-network-stream.
>
> After a little more fiddling yesterday, I found that it doesn't crash
> Aquamacs -- after about 15min, the lisp debugger kicks in and  
> returns an
> error.  Here's the full stack trace:
>
> Debugger entered--Lisp error: (file-error "make client process failed"
> "connection refused" :name "MATLAB EEI" :buffer #<buffer *MATLAB  
> EEI*> :host
> "hedgehog.local" :service 5600)
>  make-network-process(:name "MATLAB EEI" :buffer #<buffer *MATLAB  
> EEI*>
> :host "hedgehog.local" :service 5600)
>  open-network-stream("MATLAB EEI" #<buffer *MATLAB EEI*>  
> "hedgehog.local"
> 5600)
>  comint-exec(#<buffer *MATLAB EEI*> "MATLAB EEI" ("hedgehog.local" .  
> 5600)
> nil nil)
>  make-comint-in-buffer("MATLAB EEI" nil ("hedgehog.local" . 5600) nil)
>  apply(make-comint-in-buffer "MATLAB EEI" nil ("hedgehog.local" .  
> 5600) nil
> nil)
>  make-comint("MATLAB EEI" ("hedgehog.local" . 5600))
>  eval((make-comint "MATLAB EEI" (cons (system-name) 5600)))
>
> In the end, I think it's not a problem with Aquamacs, specifically --
> Emacs.app has precisely the same issue.  It's most likely something  
> about my
> system.
>
> But any help you can provide in debugging this issue would be  
> excellent.
>
> Cheers,
> Eric



[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2193 bytes --]

bug#1131: Fwd: [Aquamacs-bugs] problem with comint

[Adrian Robert]

Hello,

Does this issue still exist in emacs-23?  Can it be reproduced in any  
simple manner without access to Matlab?  Does it exist on an X11  
build of Emacs on Mac OS X?

If I try just

(make-comint "MATLAB EEI" (cons (system-name) 5600))

on Emacs.app on my system, with nothing listening on 5600 (but not a  
packet-dropping firewall either), I quickly get a lisp error about  
'(file-error "make client process failed" ...'.

thanks,
Adrian

bug#1131: [Aquamacs-bugs] problem with comint

[David Reitter]

[-- Attachment #1: Type: text/plain, Size: 897 bytes --]

Can't reproduce right now neither in 22 nor in 23, using "nc -l 5600"  
and the make-comint command.  Just using make-comint with a port that  
doesn't answer doesn't time out, but brings up an error message.

Perhaps it is a DNS problem; longer timeouts can happen there.

Comint generally works for me.

Unless we have a reproducible report, let's deprioritize this.


On 6 Mar 2009, at 14:36, Adrian Robert wrote:

> Hello,
>
> Does this issue still exist in emacs-23?  Can it be reproduced in  
> any simple manner without access to Matlab?  Does it exist on an X11  
> build of Emacs on Mac OS X?
>
> If I try just
>
> (make-comint "MATLAB EEI" (cons (system-name) 5600))
>
> on Emacs.app on my system, with nothing listening on 5600 (but not a  
> packet-dropping firewall either), I quickly get a lisp error about  
> '(file-error "make client process failed" ...'.
>
> thanks,
> Adrian
>


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2193 bytes --]

bug#1131: [Aquamacs-bugs] problem with comint

[Eric Tytell]

[-- Attachment #1: Type: text/plain, Size: 1566 bytes --]

Hi all -
  This has not been giving me any problems recently, but I'm not sure that
the original bug is fixed.  I found a work-around, by changing the line to
read

(make-comint "MATLAB EEI" (cons 'local 5600))

That one now works consistently, while the other

(make-comint "MATLAB EEI" (cons (system-name) 5600))

failed when I was connected via wi-fi (and had a DHCP assigned TCP-IP
address), but not when I was connected via ethernet (and had a static
TCP-IP) address.  I don't know if it would still be a problem the original
way it was written.

Eric


On Fri, Mar 6, 2009 at 2:44 PM, David Reitter <david.reitter@gmail.com>wrote:

> Can't reproduce right now neither in 22 nor in 23, using "nc -l 5600" and
> the make-comint command.  Just using make-comint with a port that doesn't
> answer doesn't time out, but brings up an error message.
>
> Perhaps it is a DNS problem; longer timeouts can happen there.
>
> Comint generally works for me.
>
> Unless we have a reproducible report, let's deprioritize this.
>
>
>
> On 6 Mar 2009, at 14:36, Adrian Robert wrote:
>
>  Hello,
>>
>> Does this issue still exist in emacs-23?  Can it be reproduced in any
>> simple manner without access to Matlab?  Does it exist on an X11 build of
>> Emacs on Mac OS X?
>>
>> If I try just
>>
>> (make-comint "MATLAB EEI" (cons (system-name) 5600))
>>
>> on Emacs.app on my system, with nothing listening on 5600 (but not a
>> packet-dropping firewall either), I quickly get a lisp error about
>> '(file-error "make client process failed" ...'.
>>
>> thanks,
>> Adrian
>>
>>
>

[-- Attachment #2: Type: text/html, Size: 2224 bytes --]

bug#1131: [Aquamacs-bugs] problem with comint

[David Reitter]

[-- Attachment #1: Type: text/plain, Size: 446 bytes --]

On 6 Mar 2009, at 15:43, Eric Tytell wrote:
>
> That one now works consistently, while the other
>
> (make-comint "MATLAB EEI" (cons (system-name) 5600))
>
> failed when I was connected via wi-fi (and had a DHCP assigned TCP- 
> IP address), but not when I was connected via ethernet (and had a  
> static TCP-IP) address.  I don't know if it would still be a problem  
> the original way it was written.

I think that speaks for my DNS theory.


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2193 bytes --]

bug#1131: [Aquamacs-bugs] problem with comint

[Adrian Robert]

On Mar 6, 2009, at 11:58 PM, David Reitter wrote:

> On 6 Mar 2009, at 15:43, Eric Tytell wrote:
>>
>> That one now works consistently, while the other
>>
>> (make-comint "MATLAB EEI" (cons (system-name) 5600))
>>
>> failed when I was connected via wi-fi (and had a DHCP assigned TCP- 
>> IP address), but not when I was connected via ethernet (and had a  
>> static TCP-IP) address.  I don't know if it would still be a  
>> problem the original way it was written.
>
> I think that speaks for my DNS theory.

Maybe, though in that case I'm not sure if I'd be willing to call it  
a bug in emacs, unless Ctrl-G wouldn't abort it.  (It should now in  
Emacs.app ;).  Anyway I can't reproduce it putting in whatever bogus  
name instead of (system-name).

I'm inclined to close the bug.