lfd@cbnewsm.ATT.COM (leland.f.derbenwick) (12/22/89)
I'm going to try to keep out of the flame portions of this... The complaining about multi-button mice seems to be based mostly on two things: inconsistent use of the buttons and bad physical design. The inconsistency is not the fault of the mice, it's the fault of the software that uses the mice. (Imagine if Mac applications were totally inconsistent in what single and double clicks, shift-clicks, etc., meant. One of the things Apple did _very_ right was imposing rigid standards on application interfaces.) The physical design problem leads people to have to look down to find out what mouse buttons their fingers are on. Any mouse where your fingers don't naturally fall onto the right buttons was not designed well. Or maybe is too big or small for your hand--perhaps mice should come in sizes? Right now, I use an AT&T 630MTG terminal with what I consider the best mouse physical design: a "Digimouse" made in Switzerland that is shaped like the top third of a sphere, with the right and left poles chopped off. Your hand cups over the top, in an easy, natural grip. The buttons are on the bottom of the front edge, and you press them in towards you (almost a squeezing motion). Instead of shoving the mouse around with your fingertips, which seems to be the way flat mice get used, you use mainly your wrist, which isolates the button-pressing and mouse-moving actions to different sets of muscles. NOTE: The mouse design I've described was that used on the original Bell Labs "Blit" and the AT&T 5620 terminals. For those who prefer a flat- style mouse (and it _is_ largely a matter of taste and what's comfortable for _your_ muscles), the 630 is also available with that. (The flat mouse is also reputed to be quite a bit tougher than the "Swiss mouse", which doesn't take well at all to being slammed down on the desk if your program crashes!) In terms of consistency, almost all applications that run on the 630 use the forefinger for selection (point-and-click actions), the middle finger for application-specific menus, and the ring finger for terminal- manager (includes window control) menus. The exceptions generally take over the ring finger within the window for high-level control, and have the middle finger for mode-specific menus. (Note "index", "ring", and "middle"--if you're left-handed, you can tell the terminal that, and it maps the buttons accordingly.) Games, on the other hand, use the buttons for whatever their designers' little hearts desired! :-( It's not perfect, but it's consistent enough that frequent users don't seem to get lost. I've also used the Mac interface, a two-button PC mouse, and a 3-button DEC MicroVAX mouse. The Mac seems ideally designed for the occasional user, but given that I had to spend most of my work hours with these systems and have used a good 3-button interface, the Mac feels to me like driving in snow with an automatic transmission when I'm used to a standard: it works, but I've got to drive a lot slower. The PC mouse was a good compromise--still pretty trivial, but a lot faster to use than one button. I _hated_ the original DEC interface, because of the flat mouse design, inconsistencies between the window manager interface and the main application I was using (ported from another system), and how many mouse-strokes it took to do certain common actions. But Digital made their window manager interface heavily configurable, and so I was able fairly quickly to customize one that was fast for what I usually did, and matched the application enough not to confuse me. Having done that, the 3-button mouse was again a win. (I wish AT&T's interface were as customizable, though there's less I'd change.) Lots of stuff deleted in the following, but I wanted to reply to a couple specifics: Tim Maroney (tim@hoptoad.uucp) = ">" and "> >>" Peter da Silva (peter@ficc.uu.net) = "> >" > ... If multiple-button > mice had iconic or textual labels on the buttons, they would be easier > to use. Not _on_ the buttons (the mouse I prefer, I can't even _see_ the buttons most of the time from where I sit). But I've seen some applications whose designers knew that, for an efficient user interface, the buttons were going to be used in non-standard ways. So they put 3 button-shapes at the bottom of the window with labels for what the mouse-buttons mean. It makes a huge difference, Tim--you're right about that. > >> Part of the real-world metaphor approach to improving learning curves. > >> When working with a file on a real desktop, one almost always brings it > >> to the top of the stack first. > > > >Bt not always. And there are two seperate actions involved. On my desktop, > >I sometimes slide something out from the bottom of a pile a few inches. I > >can't do thet on the Mac. > > If all you want to do is look at it, you can do that on the Mac as > well. I suggest that the number of time someone wants to make marks on > a sheet of paper which is not on the top of its stack on the desktop is > vanishingly small. Uncommon, but not vanishingly small. Uncommon enough that it shouldn't be the default--I like being able to make a window active and bring it to the top with a single point-and-click. But I routinely use multiple windows where the window I am typing in is partially obscured by other windows. The (usually smaller) windows on top are giving continuous status displays; I am acting in one window but need to be aware of what effects my actions are having on other aspects of the system. (Context is system testing a large, complex multi-machine software/hardware system.) Working in a partially obscured window is also convenient when editing a test script to correspond to both the documentation and the code (stored on machines other than the test machine): small windows on top showing the relevant portions of doc and code, and the large, partially obscured window where I'm doing the editing. Don't knock it till you've tried it! Of course, if I had a 3-foot screen with laser-printer resolution, then I could just have tiled windows instead of overlapping ones. But while I'm waiting for that... > >Not at all. There are *other* aspects of the Apple UI that are nice. > >Having scroll bars that vary in size according to the percentage of the > >objects displayed is nice. > > Mac scroll bars do not vary in size except with the size of the window; > that is, they fill the left or bottom edge of the window regardless of > its size. Perhaps you are referring to the variable position of the > thumb? I think Peter means that the portion of the bar that is highlighted varies in size with what percentage of the total window is displayed. I agree that this is a nice feature. > >But the single-button mouse, and the kludges necessary to use it, isn't. > > I don't think they're kludges. I trust you're aware that there is a > sort of "conversion shock" in switching from one interface to another; > one has always taken it for granted that the previously learned > interface was the "right way", so this new way must be the wrong way, > because of the overhead it imposes on one's mind. It is always > frustrating to have to learn new ways of doing things. I have to > wonder if this kind of shock isn't a factor here. It probably is a _major_ factor. But, having used several interfaces, I still find the 3-button the best for intensive use. If I were only an occasional computer user, the 1-button would win hands down (indeed, the Mac interface in general would). If I used a computer frequently but not constantly, I _think_ I'd go for the 2-button commonly available for IBM-compatible PCs. Given that I live with the things, 3 buttons, please! -- Speaking strictly for myself, -- Lee Derbenwick, AT&T Bell Laboratories, Warren, NJ -- lfd@cbnewsm.ATT.COM or <wherever>!att!cbnewsm!lfd