The Mouse project has matured to the point where many others will be involved in its evolution. This collection of facts should ensure that some of the important lore about the Mouse is neither lost nor distorted as time goes on.
The original specification was in the form of a piece of hardware, that is, an expensive species of Mouse which used lots of ball bearings, was susceptible to clogging from eraser fragments, cost around $400, slipped on a formica surface, and wore out after not too many hours of use (the brushes disappeared!). The specification was for a Mouse with the same sensitivity, having none of the problems just mentioned, and costing a maximum of $35. Because of the high degree of interaction between the teams developing Lisa and the Mouse, this was a quite adequate specification for a start. As time went on a document was drafted called "Mouse Specification" which described the device that people seemed to like.
The project manager was Bill Lapson. He stayed with the project as many others came and went, keeping the project moving as the organization about it went through drastic changes, in addition to being a technical contributor. Some real heroes were Dean Hovey, Jim Sachs, and Jim Yurchenco of Hovey/Kelley Design. Inspired to be part of the Apple team and hoping to be the possible manufacturers of the Mouse, these guys really used some imagination to produce a Mouse concept overcoming all the operational faults of previous Mice, and which had great promise of meeting cost objectives. Because of an urgent need for Mice for Lisa system development, Hovey/Kelley manufactured our historical Lucite Mice (only 50 were made, so hold on to those antiques). Several hundred P3 Mice were next manufactured by the Peripherals Division. These were Mice externally similar to the present Mice, but having very different innards. The heroes at this point were Bill Bull, Denise Bruno, and Esther Brunner who produced these interim Mice with virtually no fuss, delivering them on time. Chris Christophier helped with the tooling for the P3 Mice as well as the present production version.
We can attribute the present state of quasi-perfection to Paul Swearingen who (working with Bill Lapson) figured out a way to eliminate adjustments, what mechanical details affected signal timing, suggested ways of cutting costs, and Supported the production line run by ta ta! Jim Hsu, who did a fine job of planning and helping to de-bug the Mouse.
There are several names in the Lisa group which must be mentioned. The styling of the Mouse was done under the direction of Bill Dresselhaus with color and texture control by Bill Dresselhaus and Clive Twyman. Dresselhaus and Twyman should also be cited for the logo on the Mouse which is asymptotically approaching perfection. Functional performance of the Mouse was directed by Larry Tesler. Overall Lisa/Peripherals coordinator was Dave Evans who had a very substantial influence in keeping the project alive and in moderate health. A cast of anonymous secretaries performed the ergonometric (this is a clean word!) evaluation of the Mouse.
At this time a new kind of hero is involved with the Mouse. First is Peggy Hummel who is getting the Mouse released (until now all Mice have been kept in stock-room cages or tied to a computer!). This task has, without exaggeration, required the equivalent mental and emotional energy of designing the Mouse. Next is Bob Mihalus who has discovered that mice are too mortal. They lose their tails (cables break) and whiskers (springs wear out) and develop palsy (friction of the balls and shafts). Some live to be ancient but others die after a weekend beating.
Failed P3 mice should be scrapped... they've served their purpose as an interim product for in-house use. The principle failure modes, namely loose rollers and popped springs have been eliminated on the production version of the Mouse. P3 Mice can be distinguished from production Mice by looking for the black ball-retainer which is used only on the production Mouse.
Both Mice have the same ball tracking mechanism and PC board. The cases and connectors are different. Adaptation of the Lisa Mouse to Mac was done by the Mac group.
The Alps Corp independently developed a Mouse looking just like a Lisa Mouse but using mechanical brushes sliding on a disk with a radial pattern of conductors. Very clever, but too much friction and potential for wearing out as compared to the photo-optical Mouse now being produced in Peripherals. However, Alps claims they can produce the Mouse at extremely low cost ($3-10) so the concept should be given a fair evaluation and as much engineering as it takes to make it work right.
Shhhh! This may be confidential! The objective was to produce a Mouse costing under $35. Since Mice have never been produced in large volume, it is only possible to estimate that they'll cost $25-30, which includes about 10-15 minutes of assembly labor.
Until recently the plan was to manufacture Mice at the Accessory Products Division and start up a parallel manufacturing operation in Singapore. The present plan is to continue to manufacture Mice in the Peripherals Division. This may go on for three to six months while large scale testing of the Mouse is in progress. This will assure that any changes necessary will be executed by the group most familiar with the Mouse.
At various times there has been talk of "pulling the molds" and sending them to some other site, e.g., Singapore, APD, or some other vendor. Assuming there are good reasons for such a transfer, this should be done carefully. For example, the ribcage, a delicate and complex part, may not be moldable in an ordinary shop.
Changes!?!??? What kind of ------ changes? Indeed, large scale testing is likely to reveal the need for changes. For example, there is a suspected problem with cable breakage near the grommet. Possible causes are too much heat and pressure in the grommet molding process which damages the cable during manufacture. Another possible but unlikely cause is insufficient annealing of the wire after drawing. A likely cause is subjecting the cable to an unrealistically sharp bend. Some cures for a problem of this type are better control of the grommet molding process, adding another segment to the grommet, and using fatigue-resistant tinsel or copper-clad, stranded steel wire. Tinsel wire is used in telephone cords. It's been considered for the Mouse but was not used because it's stiffer and more difficult to terminate than the present stranded copper wire. Copper-clad wire wasn't used because it's custom made and may also be more difficult to terminate than stranded copper wire. Naturally, all of these cures are pursuable at this time.
An easily cured, real problem has just been found involving slippage of the idler spring on the idler wheel shaft which wears out the spring. This slippage is not supposed to occur. The cure is to put flats on the idler shaft ... almost certain to cure the problem without causing another. In fact, a small flat has been put on an idler wheel shaft and the result was to stop the unwanted slippage.
Another problem may be with signal timing. This problem affects production yield and should not affect the reliability of a tested Mouse. The important contributing factors here are consistency in the performance of the photo-optics and tolerances in the "detector block". As Mice are produced, the required performance limits for the photo-optics will emerge. This type of data can also be obtained by performing some methodical engineering tests. The present detector block tolerances are border-line. A second go at a detector block mold would result in more favorable signal timing. It would take about two months and $10,000 to make another mold, a safe course of action since an attempt to alter the present mold may destroy it.
Introducing the bounce resistance of a gerbil into the present Mouse may be the most difficult problem ... if it is a problem. The Mouse can leap off a desk like an Aztec warrior proving his courage, dangling a fraction of an inch off the floor from its cable. That's pretty good since many Mice will be required to perform this act one or more times during their life. A more serious problem ... if it is a problem, is that Mice, as they now are, Will probably never survive the Queen of SM, a direct drop on the floor from waist height. Shafts bend and cases crack. If it's really important, Mice can be made which will survive! First the shafts can be made of a stronger steel. Next, the bearing supports can be moved to provide better support. The case can be improved to resist breaking. A more extreme modification would involve stepped shafts, that is, heavy in the center where it contacts the ball and smaller at the ends.
The intensive life test program about to be carried out by Ed Houston of Corporate Services has a good chance of revealing weaknesses. I've tried to anticipate these weaknesses and their cures above. Yup! There may be more engineering.
On the other hand, why shouldn't there be more engineering on the mouse? If it is such an important part of the new computer systems, why not assign a bona-fide engineering team not only to maintain and sustain this products but to develop improved versions of Mice?
July 28, 1982
Bob Mihalus 15-A
Tim Smith 15-A
Paul Baker 2-R
Ken Okin 2-R
Wayne Rosing 2-R
Wasu Chaudari 15-A
Bob Montgomery 19-B
Jim Gisea 15-A
Dave Craft 15-A
Gary Marten 15-8
Paul Swearingen 20-E
John Vennard 20-A
Wolfgang Dirks 20-D
Jim Hsu 19-E
Denise Bruno 19-F
Chris Christophier 19-E
Walt Weber 20-0
Steve Jobs 16-G
Bill Bull 4-E
Alan Oakes 15-B
John Manousos 20-E
Bill Dresselhaus 15-B
Clive Twyman 15-B
Doug Pollack 9-B
Dave Evans 9-B
Denise Trepanier 20-C
Barry Yarkoni 20-C
Esther Brunner 19-F
Rod Holt 4-E
Tim Rother 19-A
Phil Fleschler 19-A
Joe Sheoela 19-H
Rilla Reynolds 9-N
Larry Johnston 19-D
Bruce Cheek 19-D