Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!steinmetz!davidsen
From: davidsen@steinmetz.steinmetz.UUCP (William E. Davidsen Jr)
Newsgroups: comp.emacs,comp.unix.xenix
Subject: Warning! Warning! MicroEMACS 3.9
Message-ID: <6871@steinmetz.steinmetz.UUCP>
Date: Fri, 31-Jul-87 20:14:56 EDT
Article-I.D.: steinmet.6871
Posted: Fri Jul 31 20:14:56 1987
Date-Received: Sun, 2-Aug-87 10:10:46 EDT
Organization: GE Corp. R & D, Schenectady,NY
Lines: 28
Keywords: memacs, warning, bug
Xref: mnetor comp.emacs:1539 comp.unix.xenix:543

I have noted what appears to be a bug in the new version 3.9 of EMACS
when run under System V. If I read the code right it is *polling* the
keyboard(!) for character input. After running the editor with "time", I
noted that if I started it, did nothing, and exited after 30 sec, there
would be about 23-25 sec system time.

Because this system is gotten at a low level, it seems to confuse the
dispatcher and slows the rest of the system far more than a process
using CPU in the user program would. I also saw some uucp links fail due
to timeout, which may be related. After some limit of CPU is reached,
the process seems to be "niced" and then runs at a priority just greater
than the idle loop.

I would suggest (a) see if this happens on your SysV system. I tried it
on two completely unlike systems, but your system may not have a problem
with it. (b) run the v3.8i "termio.c" with the rest of 3.9. This brought
the CPU back where I would expect it to be. I'm not sure what may break
in this configuration. There is a comment (in main.c) saying "added
Baron O. A. Grey's SYSTEM V typeahead() code...I hope this works... if
you check this out, drop me a line." I shall.

Hope this prevents some trouble, as the rest of the release is pretty
neat.

-- 
	bill davidsen		(wedu@ge-crd.arpa)
  {chinet | philabs | sesimo}!steinmetz!crdos1!davidsen
"Stupidity, like virtue, is its own reward" -me