If you look at the product Steve Jobs was most remembered for, the creation of the Mac, you'll find that in 1984, there were no user serviceable parts inside, no upgrade options and the case was fixed with very unusual bolts, two of which needed a tool that only Apple had.
Not to belabour the point:
The original Macintosh was slammed out the door, hot on correcting the many mistakes to emerge with Lisa. The Macintosh was a gambit on what Jobs expected from how he thought future users would ultimately use home computers. He was, by and large, correct, but that doesn’t mean the Macintosh 128K/512K was a
refined product.
The 128K was a literal proof-of-concept presented for sale for the well-heeled and the intrepid. It was a proof-of-concept available publicly the way the Public Beta of OS X was.
That it wasn’t designed with modular elements out of the box reflected more the challenge of getting a working product out the door in a timely manner with at least several of the major foibles from the Lisa more or less ironed out.
That the Macintosh Plus would arrive just two years later is the realization of a product whose prototype issues from the 128K/512K were ironed out enough to start planning ahead.
In those two years, the Macintosh moved from the short-lived, extremely limited MFS (no real nesting of directories) to the modernized HFS (introduced with the Plus). As well, the original Macintosh was supported only through System 3.2, whereas the Macintosh Plus, the refinement of those prototype teething experiences, débuted with System 3.0 and could run all the way through System 7.5.5.
As plans for the Macintosh Plus were already well underway (and would need to be, given the lead time on ordering the fabrication and manufacture of internal components) whilst Jobs was still at Apple during his first go-round, this wasn’t an instance of Sculley being dropped in following his prior guidance on the Apple II line, in which (magically) a semi-modular, upgradeable Macintosh came about in January 1986 — once that wily buccaneer, Jobs, had his sword taken from him in May 1985 (when he was reassigned to New Product Development and booted from Macintosh), to then walk the plank in September 1985. Arrr. 🏴☠️
(Sorry, I re-watched
The Pirate Movie not too long ago.)
'Computing as an appliance, where the user bought it, booted it, used it, and never tinkered inside was Steve's plan, and Apple pretty much stuck to that all the time he was with the company. Even when he had returned, the machines which could be tinkered with such as the Aluminum PowerBooks, the G4 iMac, G4 Mac mini, and the early Intel iMacs prior to his death, were all either sealed shut, or were difficult and complicated to disassemble.
He wanted the Macintosh experience to be simple enough for even
Harold and Linda to take out of the box, to plug in, and to get started without having to pull out the two or three, 350-page spiral-bound user manuals which shipped with the IBM PC of that moment.
The all-in-one package lent to that objective and served as an industrial design punctuation mark.
As for the PowerPC Macs after Jobs returned, the core principle of “simple enough” came through with the iMac — basically a re-stating of the original declaration made by the Macintosh. It came through with the first iBook.
Even so, latitude for upgrades and means to replace faulty parts continued not because the Spindler era ghost clinked its chains as a ghost does, but because it was straightforward; cost-effective (as, by that point, many components were commodity parts which appeared in computers generally, such as PCI buses, USB, PATA/ATAPI, and DIMMs/SO-DIMMs); and legible to a seasoned generation of computer users who didn’t really exist in such quanity in 1984, who knew things could be repaired, replaced, or upgraded.
That is: enough people in the world were using computers by the time of iMac and were aware of upgrades (like upgrading RAM) that using off-the-shelf components was, again, the simplest, most cost-effective way to deliver on the end product Jobs wanted out the door in a timely manner.
What followed with Apple hardware, as things tightened internally and parts got harder to reach (disclosure: I have done a fair bit of disassembly across every Mac between the iBook clamshell through the rMBPs of 2015), reflected something else Jobs was after, in unison with his chief industrial designer, Jony Ive: aesthetic elegance as a companion selling point unique to the Apple brand. The idea was to attract more new customers to the impression of “simplicity” — the same core tenet which went into Jobs’s objective for the Macintosh. Ive turned to Dieter Rams for inspiration.
To make that happen
and to make that happen fresh and anew with each major product redesign, this meant upselling aesthetic features like “thinness”. With thinness came compromises the company could now begin to afford to ask for in their product designs: devising proprietary components and sockets; relying more heavily on adhesives to keep everything together; and, increasingly, locking out the repair market from replacing those components. (This latter one really only accelerated once Jobs handed the keys to that guy named Tim.)
Steve did not do 'diversification' at all. Which was, in my view, one of his great strengths. He knew exactly what he wanted the user to have, and that is exactly what they got, whether he was right or wrong.
Live by the sword, die by the sword. He was fortunate to be given a second lease on life on account of having NeXT as a bargaining chip, and from that second chance, he oversaw several memorable hits.
They weren’t
all hits.
There were the hiccups. Some were speed bumps, whilst others were flops:
- the G4 Cube;
- the Core Solo/Duo & 32-bit EFI Intel models (with comically bad iGPUs);
- the 3rd gen iPod nano;
- the Yikes! Power Macintosh G4 (oh, do I have stories);
- the Blue Dalmatian and Flower Power iMacs;
- the brittle components on Titanium PowerBook G4s;
- the griddle-hot, pre-unibody MacBook Pros;
- the roll-out of OS X Cheetah;
- the hockey puck mouse (which, real talk: I did like the ridged button revision);
- the Apple Display Connector;
- the fairly high failure rates of Power Mac G5s still under warranty coverage;
- the GPU shorting out in ice and snow iBooks due to bad internal board design;
- and so on.
It doesn't do him justice to remember only the bits we want him to be known for.
No, it doesn’t. The dude had his many, many, maaaany faults — something the brand-faithful have a tendency to downplay, dismiss or, worse, forget. He was an exemplar of “all your faves are problematic” and “never meet your heroes”.