Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!uvm-gen!pegram From: pegram@uvm-gen.UUCP (pegram r) Newsgroups: comp.arch Subject: Re: Patents and Architecture Message-ID: <1561@uvm-gen.UUCP> Date: 29 Jun 90 21:45:55 GMT References: <63035@sgi.sgi.com> Sender: nobody@uvm-gen.UUCP Organization: EMBA Computer Facility, Univ. of Vermont, Burlington. Lines: 148 From article <63035@sgi.sgi.com>, by karsh@trifolium.esd.sgi.com (Bruce Karsh): > In article <05A4JU1@xds13.ferranti.com> peter@ficc.ferranti.com (Peter da Silva) writes: >>In case you haven't noticed, the car market is incredibly standardised. > Yeah, a Volkswagen Beetle is just like a Mercedes Benz. > The car market isn't very standardized at all. Perhaps you mean that the > driver interface is standardized? Well let's see: Oh brother! Yes it is - especially compared to what it was at the turn of the century - and the exceptions only go to prove the point. Going down your list... > o The key goes in on either the dashboard or the steering column. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Usually to the right of the steering column. The one major exception being Saabs (at least up through the 900) where the key goes in by the gearshift and the car must be in reverse for it to start - like the new automatic park interlocks only funkier. There has been at least one documented failure of a theft of a 900 due to the ignorance of the theives. > o The turn signal is on either the left or the right. > o The steering column is on either the left or the right. (Standards > apply on a per-country basis except for mail-carriers. > o There are either 2 or 3 motion control pedals. No comment on the first two points, they are correct - and turn signals can be self cancelling and non self cancelling (older Citroens). The motion control pedals *are* standardized. The long vertical on the left is the accellerator, the one just right of it is the brake, and the third pedal over is the clutch, if one is needed. You do know when you are driving an automatic don't you? The first two pedals operate in the same manner, push on them to get the action desired, and push harder to increase the amount of action. This was not always the case. The first popular mass produced car (covering my a** here 8-), the Ford Model T, had three pedals on the floor, but none of them operated as described above. There was no floor mounted accellerator, it was a hand unit, behind the steering wheel and opposite a control for the spark advance. On the floor were the brake, and two pedals that governed the transmission. I don't remember exactly what they did (I've never even *seen* a fliver close up), but one changed you from high to low gear (by depressing it, I believe) and the other changed the direction of motion (i.e. put you into reverse, again when depressed) I haven't got this description complete or correct, but it is enough to show how different things can be and were. Another example of offbeat foot controls was on an old racing car (german?), I think, which had the accellerator in a logical place, in between the clutch and the brake. > o There are a variety of gear shift patterns. Automatics have made PRNDL a household word, overdrive and lockup transmissions just add to that standard. The variations in how the automatic selector operates are usually fairly clear - Chrysler even used push buttons! Stick shifts usually vary in the placement of reverse, only occaisionally is first moved out of the H and fifth moved in. > o The speedometers are moving needle, digital readouts, or heads-up > displays. > o The bright lights are either on the dashboard or on the steering > column or on yet another floor pedal. > o The emergency brakes are either beside the driver's seat or on yet > another floor pedal. > o The transmission is on the steering column or beside the driver's seat > or on the dashboard. All true. > o The shift patterns come in an endless variety. You said that already, and it's not particularly true. > o The horn button is on the center of the steering wheel, on the axis > of the steering column on an additional concentric ring inside the > steering wheel or on pressure sensitive pads along the inner edge of > the steering wheel. > o The headlights are on the dashboard to the left or the right of the ^^^^^^^^^^^^^^^^^ I hope not - one sure way to a medical problem - you're blind or you crash a lot at night. 8-) 8-) The switch for them is there though, sure enough. > steering column or on the turn signal handle. > o The safety belts are on the door, on the floor, or on a mechanical > motor driven gizmo which puts the shoulder belt on automatically. > These are examples of the lack of standardization of the safety related > components of an automobile. This has not detered drivers from switching > freely between various models of autos. The point is that there is a *limit* on the variability between cars, a generally tolerable one, caused by Adam Smith's invisible hand ((I *had* to say that 8-), read "economic forces of the market place"). IMHO we are at or *just* past the Model T stage. The differences between a Mac, a Xerox Star, Gem, Windows (or it's look alikes) and Amiga's intuition are tolerable and relatively easy to bridge, the difference between them and even the pop up menus on a Sun with Xwindow is too much to be easily bridged. The interface may or may not be optimal - accelerator, brake and clutch placement and action are not. The interface may not be very easy to understand the first time either, but the learning curve should be reasonably short and only have to be done once! The interface should also allow you to do what you need to do with reasonable efficiency once you have mastered it. Does anyone want to steer with a tiller these days? I wager that few will want to deal much with command lines in future either - and only then for the power that command lines offer for certain operations. If you have a *much* better idea for a user interface, you better come out with it _NOW_ or forever hold your peace - except perhaps on the net 8-). Most likely the changes to our current graphical interfaces will be evolutionary, such as making each window have its own menu bar or adding tear-off menus (both things currently out there, but not on Macs for example). > In the same way, computer architectures need not be standardized. Users > don't need every display to be identical among their computers, they need > for practically every display to be understandable. True enough, but who arbitrates on "understandableness"? Customers do, and they are likely to go with what they understand already. It's not just the designers (see below ) who like identical solutions for the basic operations. How many of us like to change text editors for example, hmm? > Since that's hard to do, many software designers have instead concluded that > identical-ness is the panacea for the usability problem. The auto is a > good example of why this is not so. Customers don't like identical-ness, they > like new and improved things. If auto makers all made exactly the same ^^^^^^^^^ exactly my point, *evolutionary* change does work. > interfaces, most would be put out of business by the few who did new things. > If computer designers stop making new and improved interfaces, they'll be > put out of business by designers who will innovate. > > So let's innovate. Let's give users a choice. Lets put the tired old > non-inovative companies out of business. > >>Your position seems to be that Ford and GM shouldn't have the steering wheel >>on the same side, and Chrysler should put the driver in the back seat. > > Ford, GM, and Chrysler all have models with the steering wheel on either > side. Should all cars have identical steering wheels. Should granny's car > and a low-rider's car both have the same steering wheel? Granny wont like > it and neither will the low-rider. > Bruce Karsh > karsh@sgi.com Bob Pegram (Internet: pegram@griffin.uvm.edu)