Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!seismo!mcvax!unido!tub!net From: net@tub.UUCP Newsgroups: comp.sources.bugs Subject: Re: Bugfixes/Enhancements to "screen" - (nf) Message-ID: <52100003@tub.UUCP> Date: Sun, 26-Jul-87 16:21:00 EDT Article-I.D.: tub.52100003 Posted: Sun Jul 26 16:21:00 1987 Date-Received: Tue, 28-Jul-87 00:42:52 EDT References: <256@srs.UUCP> Lines: 46 Nf-ID: #R:srs:-25600:tub:52100003:000:2216 Nf-From: tub!net Jul 26 21:21:00 1987 > Here is an update for the "screen" program posted several months ago. I'm sorry to tell you this, but I have submitted a new release of screen to the moderator of comp.sources.unix three weeks ago; it is certainly not possible to apply your patches to the new release without modification. Incidentally, some of your enhancements have already been incorporated into this release... [I don't know why screen has not yet appeared in comp.sources.unix; sometime this week I'm going to write a letter to Rich Salz asking if he received my submission at all, or if there is some other problem.] > patch2: Fixes a small bug that would crash "screen" when run > under SunWindows (terminal type = sun). I have published a patch fixing this problem together with the other patch; I wonder why you have got only one of them. > 1) The ability to explicitly name each window I had also planned to implement a function to "label" windows; however, I was not quite sure how screen should obtain the name for a newly created window (via C-a C-c); I don't like the idea to directly prompt the user for input (i.e. read from the keyboard), because this causes all other windows to "hang" until the call to read() terminates. > 3) Doesn't redraw the current window if you happen > to change to the same one (via "^A[0-9]") The new release of screen doesn't redraw the display upon C-a C-a when there is only one window. However, it still doesn't detect C-a when window is already active... > 5) Makes an entry into "/etc/utmp" [...] > This code is stolen from the "grotwin" package [...] I don't like this patch very much. The utmp-code from "grotwin" is somewhat ugly and it doesn't work on 4.3 BSD systems with the new /etc/ttys format. The new release of screen contains code to properly handle /etc/utmp on 4.2 and (new-format /etc/ttys) 4.3 systems (using the gettyent(3) functions, if possible). It also contains the code that is necessary when you want install screen as set-uid with owner root (so that it is able to modify /etc/utmp if it is not writable for world). Oliver Laumann, Technical University of Berlin, Germany. ...!pyramid!tub!net or net@TUB.BITNET