Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!caip!rutgers!nike!ucbcad!ucbvax!RICE.EDU!rbbb
From: rbbb@RICE.EDU (David Chase)
Newsgroups: net.emacs
Subject: Re: GNU Emacs hanging ...
Message-ID: <770.rbbb.titan@Rice>
Date: Sat, 18-Oct-86 21:32:26 EDT
Article-I.D.: Rice.770.rbbb.titan
Posted: Sat Oct 18 21:32:26 1986
Date-Received: Tue, 21-Oct-86 21:06:42 EDT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The ARPA Internet
Lines: 36

If you cannot wait for the fixes from Sun, you can avoid this problem by
adding two sleeps to your code.  The delay (1 second) is annoying, but it
absolutely avoids the problem (I know; it has not happened to me even once
since I did this, and that was at the beginning of the summer).

The places to sleep:

1) when setting the terminal back to its original (non-emacs) state,
insert a sleep AFTER you have sent the terminal-altering escape sequences,
BEFORE you perform whatever stty's are necessary to set the terminal in
its non-emacs state.

2) when setting the terminal into its emacs state, insert a sleep AFTER
you perform any stty's, BEFORE you send any sequences to the terminal.

Thus, you get something like:

enter_emacs() {
   stty(...)
   sleep(1)
   terminal_init()
   }

leave_emacs() {
   terminal_restore()
   sleep(1)
   stty(...)
   }

The one second sleep is a bit heavy-handed, but it was easy, does the job,
and isn't too offensive to users (especially after emacs has hung them up
once or twice).

This (hanging) happens a LOT if you are on a Sun 3 running Gosling emacs.

David