From owner-freebsd-hackers Wed Mar 13 02:58:50 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id CAA21621 for hackers-outgoing; Wed, 13 Mar 1996 02:58:50 -0800 (PST) Received: from nixpbe.pdb.sni.de (mail.sni.de [192.109.2.33]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id CAA21602 for ; Wed, 13 Mar 1996 02:58:24 -0800 (PST) Received: (from nerv@localhost) by nixpbe.pdb.sni.de (8.6.12/8.6.12) id LAA25407 for hackers@freebsd.org; Wed, 13 Mar 1996 11:57:42 +0100 Message-Id: <199603131057.LAA25407@nixpbe.pdb.sni.de> Subject: Re: using ddb to debug a double-panic? To: terry@lambert.org (Terry Lambert) Date: Wed, 13 Mar 96 11:54:08 MET From: Greg Lehey Cc: hackers@freebsd.org (Hackers; FreeBSD) In-Reply-To: <199603122105.OAA06519@phaeton.artisoft.com>; from "Terry Lambert" at Mar 12, 96 2:05 pm X-Mailer: xmail 2.4 (based on ELM 2.2 PL16) Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >> I've come in on the end of this. This horrible thing that passes for >> a mail system here has swallowed a day's worth of mail again, and >> didn't even regurgitate it. If anybody else has sent opinions, please >> resend them. > > Are you the originator of this thread? I don't think so. I just mentioned that I was planning to significantly expand ddb when I get away from this place. > Did you miss my "small essay > and bogus screen shot" for WINICE? Looks like it. You mentioned WINICE, and I asked for more details. I definitely didn't see anything that looked like a screen shot. > Do I need to resend it to you? Yes, please. >> Seriously, Bruce (or anybody else): what kind of editing? The main >> objection I have to vi-style editing in shells is that it is so >> ESC-intensive. I suspect it's also more difficult to program, though >> I don't suppose that's the real problem. If anybody has any >> alternatives to emacs-style bindings, please let me know a detailed >> description of how it should work. > > 4DOS/DOSKey/WinICE/FANSIConsole/DCL/TPU/EDT/tcsh/ksh(the real one)/etc. > all use: > > 1) Cursor up: previous command (repeat to the limits of the > command recall buffer). OK. Alternative: ^P > 2) Cursor down: next command (assuming a previous cursor up -- > otherwise default command input line). OK: Alternative: ^N > 3) Partial command, then cursor up/down: command previous/next > beginning with partial command as line prefix. I don't see the need for this. It's different from other implementations, and you can get this effect by going to the beginning of a line and typing in the text. > 4) Abort key (^C): discard edits, new input line. OK, I suppose. > 5) Backspace: delete character to left of cursor OK. > 6) Delete: delete character to right of cursor OK, I suppose. Alternative: ^D. > 7) Insert: toggle insertion/overstrike; default overstrike (on > VMS programs, can change default as a preference... on 2 of > the DOS programs, the previous in-ude mode is remembered -- > the toggle is global, not per entry line). I hadn't planned on insert mode at all. Does anybody else have any ideas on this? > 8) Home key: beginning of line OK, I suppose. Alternative: ^A > 9) End key: end of line OK, I suppose. Alternative: ^E In addition, I'd definitely want: 10) ^B/left arrow: character left 11) ^F/right arrow: character right 12) ^K: erase to end of line 13): ^Y: yank erased text In addition, I think I'd like to have: 14) Alt-D: delete word forward 15) Alt-F: forward word 16) Alt-B: Backward word > The one problem is the "backspace deletes left" for terminals where > the cursor left key emits "^H". In these situations, the BS key > becomes synonymous with the cursor left key and the "delete character > to left of cursor" function is lost. I think that, under these circumstances, I'd opt for the cursor left function being "lost" (i.e. only available via ^B). Recall that we're talking console only here, of course, so the difference is moot. > Extensions: scroll left/right in line for long lines instead of damaging > editability (ie: don't screw up like VMS did). I *hate* scroll. IMHO, the right (and more difficult) way is to wrap lines. There's also the question of how often a line of more than 80 characters will be needed. I'll send out a separate message on other features that are beginning to occur to me. If anybody has any thoughts about this subject, please reply. Greg