I wish it had beyond gross appearances. The bitmap display and mouse pointing device were the most obvious inspirations.

The Alto was intended to amplify a wide range of users.

Some of the software on Alto — such as Smalltalk with its overlapping window GUI — were directly responsible for the overall look and feel of the Mac (and before that, the Lisa).

Photo of Alto screen showing Smaltalk-74 right before Smalltalk-76

The Alto — Chuck Thacker main designer — which first appeared in 1973, was a “parametric machine” of 16 virtual processors with 0-overhead-interleaving driven by microcode using an engine that was 5–6 times faster than main memory. The minimal display on the Alto was 808x606 (about 490,000 pixels).

The overall architecture of the Alto, with format of a microcode word below

11 years later, the Mac was a rather underpowered Motorola 68000, whose CPU was along the lines of a mini-computer (like a PDP-11 but not a Data General Nova). The Mac display was 512 by 342 (about 175K pixels, a little more than just 1/3 the pixels on the Alto 11 years earlier).

However, it is worth looking at the earlier Lisa, which was more of what people needed, but was about the price of an average US car (to me, a bargain for one’s information vehicle! but not to most consumers and businesses).

One of the significant design strengths of the Alto was that the reloadable microcode allowed customized efficient virtual machines for very high level languages to be built, and be modified to introduce optimizations as they were discovered to be needed.

The microcode also allowed new real-time display and sound primitives (such as bitblt, the universal screen painting primitive, 2.5D half-tone animation, and both 12 voice polytimbral real-time sampling and 8 voice polytimbral FM synthesis to be done several years after the Alto had been designed).

And most of what is usually HW in all the peripheral controllers was simulated by the microcode (including controllers for the display, disk, keyboards, mouse, handset, sound, Ethernet, and up to 1000 input lines). This both saved a lot of money to make these machines, and also provided much more flexibility.

Software is a very good idea! (But it also needs to be taken deeply into the “hardware”!)

Taking the 11 year difference into account, the various computers designed and built at Parc in the early 80s still had a microcode architecture but were now able to use several of the AMD “bit-slice” CPU components. This would have been a great choice for Apple to take up — it would have revolutionized personal computing — but most of the Apple technical people were very uncomfortable with bases that weren’t like mini-computers.

Both Intel and Motorola could have made chips that were more “parametric” with regard to their function. This never happened despite the fact that most versions of the chips they did make had various kinds of fixed immutable “micro” and “nano” coding internally.

Apple routinely blew chance after chance to have a dual strategy for the future by rejecting a number of opportunities to also have a line of workstations, which by Moore’s Law would be Apple’s future Macs 5–7 years out.