[10717] in Athena Bugs
IBM RS/6k emacsclient problems
daemon@ATHENA.MIT.EDU (wchuang@Athena.MIT.EDU)
Thu Jul 29 11:51:46 1993
From: wchuang@Athena.MIT.EDU
Date: Thu, 29 Jul 93 11:51:35 -0400
To: bugs@MIT.EDU
Cc: vgeorge@Athena.MIT.EDU
Emacsclient on the IBM RS/6k's occasionally will return from Emacs, but
the calling program will not realize that it has returned. Here is a
log from a user's question in OLC (long post follows):
> Log Initiated for user Varghese George (vgeorge@ARJUN.MIT.EDU [0]).
> [Thu 29-Jul-93 9:52am]
>
> Topic: emacs
>
> Question:
>
> I run an emacs client for my mail. My ..mh_profile is as follows:
>
> Path: /mit/15.302/wrk/varghese/Mail
> #Path: Mail
> Editor:/usr/athena/bin/emacsclient
> Msg-Protect: 600
> Folder-Protect: 700
> Prompter-next: emacsclient
> rmmproc: /bin/athena/delete
> send: -alias aliases
> ali: -alias aliases
> post: -alias aliases
> whom: -alias aliases
>
> The problem is this. Sometimes (1 in 5 days) the following happens:
> - I "comp" an email
> - I hit C-x #, to return from emacs client to "comp"
> - Emacs client disappears, but "comp" stays dead without asking
> "what now?"
>
> None of the following helps:
> - restarting emacs server. (I have (server-start) in .emacs.)
> - restarting emacs itself
>
> Only rebooting the machine makes things ok.
>
> Is there another way?
>
> Machine info:
> Processor: IBM RS6000
> Display : Color IBM RS/6000
> Memory : 16384 K
> ___________________________________________________________
>
> --- Question grabbed by consultant tarah@MOMONEY.MIT.EDU [2].
> [Thu 29-Jul-93 9:56am]
>
> *** Reply from consultant tarah@MOMONEY.MIT.EDU [2].
> [Thu 29-Jul-93 9:57am]
> Hi there..
>
> When you comp an email message do you type this command at your athena%
> prompt.
>
> athena% comp
>
> ??
>
> -tara
> olc consultant
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 10:01am]
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 10:03am]
>
> Yup.
>
> Then, it will say:
> Waiting for Emacs...
>
> But even after I finish with emacs with C-x #, "Waiting for Emacs..."
> does not go away.
>
> *** Reply from consultant tarah@MOMONEY.MIT.EDU [2].
> [Thu 29-Jul-93 10:04am]
> Have you tried using:
>
> C-x c
>
> to exit from emacs. Would this work better?
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 10:05am]
>
> *** Reply from consultant tarah@MOMONEY.MIT.EDU [2].
> [Thu 29-Jul-93 10:06am]
> Sorry,.. that was
>
> C-x C-c
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 10:08am]
>
> I am trying it now ..
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 10:08am]
>
> *** Reply from consultant tarah@MOMONEY.MIT.EDU [2].
> [Thu 29-Jul-93 10:12am]
>
> Unfortunately, I have to leave but another consultant will be with you
> shortly if you have further problems. However if your question is
> resolved, please type 'done' at the olc> prompt to terminate the
> question. Thanks for using OLC!!
>
> -tara
> olc consultant
>
> --- Description changed by consultant tarah@MOMONEY.MIT.EDU [2].
> [Thu 29-Jul-93 10:12am]
> comp problems on dialup
>
> --- Question forwarded by tarah to pickup.
> [Thu 29-Jul-93 10:12am]
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 10:12am]
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 10:21am]
>
> Unfortunately, the following happens:
>
> C-x C-c, kills the emacs window, but the control comes back to "comp" and I can send the
> message. But emacs is gone, and my use of emacsclient is nullified.
>
> Is there a way not to knock out my emacs?
>
> --- Question grabbed by consultant wchuang@NAVIGATOR.MIT.EDU [2].
> [Thu 29-Jul-93 11:12am]
>
> *** Reply from consultant wchuang@NAVIGATOR.MIT.EDU [2].
> [Thu 29-Jul-93 11:13am]
> Hello, Varghese...
>
> Once you use emacsclient, and hit "Control-x #" in the Emacs window,
> go back to your xterm and hit Control-c.
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 11:20am]
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 11:22am]
>
> THAT just kills the "comp" with the response:
>
> =>comp
> Waiting for Emacs...^Cwhatnow: problems with edit--/mit/15.302/wrk/varghese/Mail/draft pr
> ese
> rved
>
> Message does not get sent.....
>
> *** Reply from consultant wchuang@NAVIGATOR.MIT.EDU [2].
> [Thu 29-Jul-93 11:26am]
> That's very odd (I occasionally get similar behavior, but Control-c
> works for me -- perhaps because I'm using a different program which
> is doing different things).
>
> I don't think there's a good work-around for this (however, you can
> always switch buffers back to the original text of the message, in
> Emacs, and then save that buffer)...
>
> I'd suggest sending e-mail to "bugs" about it.
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 11:26am]
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 11:34am]
>
> Which is the "different" program that you use?
>
> Also, is it possible to send a transcription of today's conversation with
> olc to bugs? It will kill me to reproduce it separately.
>
> *** Reply from consultant wchuang@NAVIGATOR.MIT.EDU [2].
> [Thu 29-Jul-93 11:36am]
> olcr (for responding to olc messages)
>
> You'd like me to send it?
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 11:37am]
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 11:38am]
>
> I'd really appreciate it....
>
> I'd also like to know why C-c in xterm works for you ...
>
> *** Reply from consultant wchuang@NAVIGATOR.MIT.EDU [2].
> [Thu 29-Jul-93 11:39am]
> It's not in the xterm, it's running emacsclient from olcr, in an xterm.
> olcr is probably a bit more robust in this case.
>
> --- User vgeorge read reply.
> [Thu 29-Jul-93 11:39am]
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 11:40am]
>
> Thanks ..
>
> *** Reply from user vgeorge@ARJUN.MIT.EDU [0].
> [Thu 29-Jul-93 11:40am]
>
> Please send a copy of the bug-report to me.
>
> --- User vgeorge is done with question.
> [Thu 29-Jul-93 11:41am]
o William Chuang
o CSS Student Consultant
