From mboxrd@z Thu Jan 1 00:00:00 1970 Path: main.gmane.org!not-for-mail From: Miles Bader Newsgroups: gmane.emacs.devel Subject: Re: German tutorial fix Date: 20 May 2002 01:10:58 +0900 Sender: emacs-devel-admin@gnu.org Message-ID: <87u1p4ruct.fsf@tc-1-100.kawasaki.gol.ne.jp> References: <87y9emho8s.fsf@lexx.delysid.org> <200205160722.g4G7M8X18181@aztec.santafe.edu> <3277-Thu16May2002155214+0300-eliz@is.elta.co.il> <200205171928.g4HJSdw20390@aztec.santafe.edu> <87znyywagx.fsf@pot.cnuce.cnr.it> <87adqyz047.fsf@tc-1-100.kawasaki.gol.ne.jp> <200205190530.g4J5UNa22924@aztec.santafe.edu> <87n0uwuwkc.fsf@emacswiki.org> <87offctgcb.fsf@tc-1-100.kawasaki.gol.ne.jp> <87elg8upk7.fsf@emacswiki.org> Reply-To: Miles Bader NNTP-Posting-Host: localhost.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: main.gmane.org 1021824808 8314 127.0.0.1 (19 May 2002 16:13:28 GMT) X-Complaints-To: usenet@main.gmane.org NNTP-Posting-Date: Sun, 19 May 2002 16:13:28 +0000 (UTC) Cc: emacs-devel@gnu.org Return-path: Original-Received: from quimby.gnus.org ([80.91.224.244]) by main.gmane.org with esmtp (Exim 3.33 #1 (Debian)) id 179TJI-00029z-00 for ; Sun, 19 May 2002 18:13:28 +0200 Original-Received: from fencepost.gnu.org ([199.232.76.164]) by quimby.gnus.org with esmtp (Exim 3.12 #1 (Debian)) id 179TWp-00049M-00 for ; Sun, 19 May 2002 18:27:27 +0200 Original-Received: from localhost ([127.0.0.1] helo=fencepost.gnu.org) by fencepost.gnu.org with esmtp (Exim 3.34 #1 (Debian)) id 179TJQ-000253-00; Sun, 19 May 2002 12:13:37 -0400 Original-Received: from smtp01.fields.gol.com ([203.216.5.131]) by fencepost.gnu.org with smtp (Exim 3.34 #1 (Debian)) id 179THa-00022D-00; Sun, 19 May 2002 12:11:42 -0400 Original-Received: from tc-2-252.kawasaki.gol.ne.jp ([203.216.25.252] helo=tc-1-100.kawasaki.gol.ne.jp) by smtp01.fields.gol.com with esmtp (Magnetic Fields) id 179THX-0005SD-00; Mon, 20 May 2002 01:11:39 +0900 Original-Received: by tc-1-100.kawasaki.gol.ne.jp (Postfix, from userid 1000) id 96E7030B5; Mon, 20 May 2002 01:10:58 +0900 (JST) Original-To: Alex Schroeder System-Type: i686-pc-linux-gnu In-Reply-To: <87elg8upk7.fsf@emacswiki.org> Original-Lines: 98 X-Abuse-Complaints: abuse@gol.com Errors-To: emacs-devel-admin@gnu.org X-BeenThere: emacs-devel@gnu.org X-Mailman-Version: 2.0.9 Precedence: bulk List-Help: List-Post: List-Subscribe: , List-Id: Emacs development discussions. List-Unsubscribe: , List-Archive: Xref: main.gmane.org gmane.emacs.devel:4126 X-Report-Spam: http://spam.gmane.org/gmane.emacs.devel:4126 Alex Schroeder writes: > I bet it also depends on keyboard layout and customizations outside of > Emacs. Well of course, but emacs' entire command set is predicated on certain constraints, like the control key being easy to type. > Skimming the TOC and checking some chapters selectively, I think here > is what he might say: [I know I shouldn't reply to this, because it's really beside the point, but well...] > 1. Habit formation -- sometimes you use the arrow keys, sometimes C-f > to move point. That is bad for habit formation. That is no doubt true. However, the claim made was not that the control-keys are _easier_ to learn (they clearly are not, if the user already knows about arrow keys), but that they are more efficient to use. There's also an argument that the habits we want people to form are in fact the control-key methods of doing stuff. > 2. GOMS keystroke level model -- arrow keys might involve hand > movement similar to moving from the keyboard to the mouse, thus you > have one H element in the analysis, and a K for the press, and > mentally preparing M. C-f has mentally preparing, and two > keypresses. The timing he gives for the simplified analysis would > be M = 1.35s, K = 0.2s, H = 0.2s, thus the two are exactly > equivalent as far as the GOMS model is concerned. Hmmm, that would be a good clue that the GOMS model (whatever that is) has a flaw. Moving your hands from the home row to the arrow keys (and back again) is _very_ costly on most PC keyboards -- not only is the actual movement expensive, but (1) quite often you have to adjust your body position to accomodate the arm movement [they're _that_ far away], and (2) you screw up the positioning of your hands on the home row (which doesn't happen at all with the control-key cursor movement keys; they're in fact almost exactly like typing). Thus, if I understand what you wrote above, he's seriously underestimating H. I would also suggest that he's _overestimating_ M in many cases -- the sort of simple editing commands that often happen while typing in text are not the sort of thing you actually think about, they're almost automatic. I don't doubt that he's based the above numbers on actual measurements, although such things are almost certainly highly dependendent on who's being measured. > 3. Hick's Law -- since you now have two equivalent methods of moving > point, this not only hampers habit formation, it also imposes a > cognitive burden when you have to choose between the two. See point 1. In any case, everything is a tradeoff -- while having many arbitrary and capricious methods of performing every command is certainly confusing, in some cases the benefits outweigh the problems. In the current case, we're talking about probably the _most used_ commands in emacs (well, actually I suppose that would be `self-insert-command' :-), which is exactly the case where one wants to think about efficiency, even if it takes a slight bit of extra work by the user to remember. [after all, if we really took this `Law' seriously, we'd have to get rid of the keybindings, and M-x, and just use the menus for everything...] > My claim is just that 1. C-f is not obviously better, and 2. conflicting > opinions exist. So why use it as an argument, if we have far better > arguments at hand? For example stupid terminals. The danger is that people will read such this and say `I have arrow keys, so I don't need to bother with all these strange and odd key-bindings -- they're only for sad users stuck on 1970's hardware.' If we believe (as I obviously do) that the non-arrow-key bindings are _better_ in some situations, and also offer other less tangible benefits (such as their integration into the emacs `scheme' for keybindings), then we really ought to also point out `Hey, these bindings may seem unnecessary, but really, you ought to try them for a while, you may be glad you did.' Even if they later decide that they really rather prefer the arrow keys, just having some experience with the traditional cursor-movement keys will probably aid them in understanding other emacs' keybindings, and may make learning emacs easier in general. > I have it, RMS had it, iirc, Ben Wing had it, JWZ had it, James > Gosling had it, ... health might be just as important as typing speed. Sure, but I suspect that most of the emacs command set is toast if you really care to avoid RSI (and are arrow keys really any better, other than by slowing you down? They require hand movement and keypressing, just like any other key...). -Miles -- Somebody has to do something, and it's just incredibly pathetic that it has to be us. -- Jerry Garcia