IRWIN@PUCC.BITNET (Irwin Tillman) (12/04/86)
At Princeton we are building a campus-wide network which will support
a variety of different devices. We are in search of a good
escape sequence which we can reserve for the use of the internet.
Users currently type ESCAPE DELETE to get the network, but this sequence
is used by some software. We are considering Control-X Control-A, and
would like to hear from others who have faced this situation. What
sequence did you choose? Does anyone have a favorite piece of software
or hardware that already makes use of the sequence Control-X Control-A?
Irwin Tillman BITNET: IRWIN@PUCC
Princeton University UUCP: allegra!psuvax1!PUCC.BITNET!IRWIN
ARPA: IRWIN%PUCC.BITNET@wiscvm.wisc.edurobert@gitpyr.gatech.EDU (Robert Viduya) (12/05/86)
>IRWIN@PUCC.BITNET (IRWIN@PUCC.BITNET, <1420@PUCC.BITNET>): > At Princeton we are building a campus-wide network which will support > a variety of different devices. We are in search of a good > escape sequence which we can reserve for the use of the internet. > Users currently type ESCAPE DELETE to get the network, but this sequence > is used by some software. We are considering Control-X Control-A, and > would like to hear from others who have faced this situation. What > sequence did you choose? Does anyone have a favorite piece of software > or hardware that already makes use of the sequence Control-X Control-A? Well, the Kermit protocol uses Control-A as a Start-Of-Packet indicator and specifies that any character may precede it. At GaTech, our campus-wide network uses the sequence Control-O Control-L Control-L as a hold (i.e., pause the virtual circuit and drop into a network control prompt) and Control-O Control-K Control-L as a disconnect (hangup). I'm not sure why those particular sequences were chosen, but I do know that a three character sequence was chosen over a two character sequences mainly to decrease the chances of that sequence from occurring in various communication protocols. One could easily argue that a four character or five character sequence would decrease the chances even more, but one has to draw a line somewhere. Anyway, the user can set the sequence to anything he wants to at any time, although network software limits it to a maximum of four characters. Alternatively, each port can be configured to use a BREAK instead of a sequence of characters as either hold or disconnect (I've got BREAK set as my hold sequence and no disconnect sequence). In my opinion, BREAK seems to be the best choice mainly because it is an "out-of-band" signal (similar to RTS/CTS being "out-of-band" flow control signals). Most host based software treat BREAK as being "fragile" and provide ways of signalling a BREAK through other characters. robert -- Robert Viduya robert@pyr.ocs.gatech.edu Office of Computing Services (404) 894-4660 Georgia Institute of Technology Atlanta, Georgia 30332
mouse@mcgill-vision.UUCP (der Mouse) (12/13/86)
In article <1420@PUCC.BITNET>, IRWIN@PUCC.BITNET (Irwin Tillman) writes: > At Princeton we are [...] in search of a good escape sequence which > we can reserve for the use of the internet. Users currently type > ESCAPE DELETE to get the network, but this sequence is used by some > software. We are considering Control-X Control-A, [...] Does anyone > have a favorite piece of software or hardware that already makes use > of the sequence Control-X Control-A? Emacs. I put list-buffers on ^X^A. ESC-DEL sounds like a reasonable choice to me. I might even have chosen it if I were writing the code. What software is this that picked ESC-DEL? I guess it just goes to show that there's *nothing* that nobody uses. You can get unlikely, of course, but that's all you can do. der Mouse USA: {ihnp4,decvax,akgua,utzoo,etc}!utcsri!mcgill-vision!mouse think!mosart!mcgill-vision!mouse Europe: mcvax!decvax!utcsri!mcgill-vision!mouse ARPAnet: think!mosart!mcgill-vision!mouse@harvard.harvard.edu [USA NSA food: terrorist, cryptography, DES, drugs, CIA, secret, decode]
rpw3@amdcad.UUCP (Rob Warnock) (12/14/86)
I believe the Rand "e" editor (or some versions of it) use
Cntl-X Cntl-A as a prefix for some commands. Emacs may also
have such commands. In any case, Cntl-X is a very popular
prefix character in the world of editors, and Cntl-A is
probably used a low.
Also, many protocols (not just Kermit, as mentioned) use
Cntl-A as a lead-in, since it's ASCII <SOH> or Start-of-Header.
Cntl-X is <CAN> or Cancel. I wouldn't be at all surprised to see
a packet aborted with <CAN> and a new one started with <SOH>.
I'd avoid that particular sequence.
Rob Warnock
Systems Architecture Consultant
UUCP: {amdcad,fortune,sun}!redwood!rpw3
DDD: (415)572-2607
USPS: 627 26th Ave, San Mateo, CA 94403ben@catnip.UUCP (12/14/86)
In article <2765@gitpyr.gatech.EDU> robert@gitpyr.UUCP (Robert Viduya) writes: >>IRWIN@PUCC.BITNET (IRWIN@PUCC.BITNET, <1420@PUCC.BITNET>): >> At Princeton we are building a campus-wide network which will support >> a variety of different devices. We are in search of a good >> escape sequence which we can reserve for the use of the internet. >> Users currently type ESCAPE DELETE to get the network, but this sequence >> is used by some software. We are considering Control-X Control-A, and >> would like to hear from others who have faced this situation. What >> sequence did you choose? Does anyone have a favorite piece of software >> or hardware that already makes use of the sequence Control-X Control-A? > >Well, the Kermit protocol uses Control-A as a Start-Of-Packet indicator and >specifies that any character may precede it. I don't think this is right. I thought that the Control-A was the *only* control character Kermit would use. Further, I thought Kermit was designed for use in a hostile environment, i.e., one where characters coming from the serial line were seven bits long, and where control codes such as X-ON, X-OFF and del did nasty things. As for a good escape sequence, I am partial to the one used on our AT&T ISN. / BREAK. I believe that the BREAK has to be typed within a second or so after the /. This escape sequence seems to be transparent to all the software we have tried. -- Ben Broder {ihnp4,decvax} !hjuxa!catnip!ben {houxm,clyde}/
andersa@kuling.UUCP (Anders Andersson) (12/18/86)
In article <569@mcgill-vision.UUCP> mouse@mcgill-vision.UUCP writes: >Emacs. I put list-buffers on ^X^A. > >ESC-DEL sounds like a reasonable choice to me. I might even have >chosen it if I were writing the code. What software is this that >picked ESC-DEL? Emacs. I put rubout-word on ESC-DEL. It's even "standard", isn't it? I nowadays consider an out-of-band BREAK (or anything involving BREAK) as the best choice for a primary back-to-terminal-server command. This may cause trouble if your terminal has a long way to go to the server (for instance via some other host computer or communication devices), but such connections are ugly anyway. As second best, a prefix character should be chosen from those seldom used (touch wood!) ^\, ^], ^^ or ^_. I would explicitly avoid ESC and ^X because of their use in various applications, most notably Emacs. -- Anders Andersson, Dept. of Computer Systems, Uppsala University, Sweden Phone: +46 18 183170 UUCP: andersa@kuling.UUCP (...!{seismo,mcvax}!enea!kuling!andersa)