When technology is not the limit

The early days of computing were all about limits. Regardless of the era you pick (mainframe, minicomputer, PC, client-server, etc.) the systems were constrained and imposed hard limits on computations. CPUs were limited in speed. Memory was limited to small sizes. Disks for storage were expensive, so people used the smallest disk they could and stored as much as possible on cheaper tape.

These limitations showed through to applications.

Text editors could handle a small amount of text at one time. Some were limited to that amount and could handle only files of that size (or smaller). Other editors would "page out" a block of text and "page in" the next block, letting you work on one section of the text at a time, but the page operations worked only in the forward direction -- there was no "going back" to a previous block.

Compilers would allow for programs of only limited sizes (the limits dependent on the memory and storage available). Early FORTRAN compilers used only the first six characters of identifiers (variable names and function names) and ignored the remainder, so the variables DVALUES1 and DVALUES2 were considered to be the same variable.

In those days, programming required knowledge not only of the language but also of the system limitations. The constraints were a constant pressure, a ceiling that could not be exceeded. Such limitations drove much innovation; we were constantly yearning for more powerful instruction sets, larger memories, and more capacious and faster storage. Over time, we achieved those goals.

The history of the PC shows such growth. The original IBM PC was equipped with an 8088 CPU, a puny (by today's standards) processor that could not even handle floating-point numbers. While the processor could handle 1 MB of memory, the computer came equipped with only 64 KB of RAM and 64 KB of ROM. The display was a simple arrangement, with either high-resolution text only monochrome or low-resolution graphics in color.

Over the years, PCs acquired more powerful processors, larger address spaces, more memory, larger disk drives (well, larger capacities but smaller physical forms), and better displays.

We are at the point where a number of applications have been "solved", that is, they are not constrained by technology. Text editors can hold the entire document (up to several gigabytes) in memory and allow sophisticated editing commands. The limits on editors have been expanded such that we do not notice them.

Word processing, too, has been solved. Today's word processing systems can handle just about any function: wrapping text to column widths, accounting for typeface variations and kerning, indexing and auto-numbering, ... you name it.

Audio processing, e-mail, web browsing, ... all of these have enough technology to get the job done. We no longer look for a larger processor or more memory to solve our problems.

Which leads to an interesting conclusion: When our technology can handle our needs, an advance in technology will not help us.

A faster processor will not help our word processors. More memory will not help us with e-mail. (When one drives in suburbia on 30 MPH roads, a Honda Civic is sufficient, and a Porsche provides no benefits.)

I recognize that there are some applications that would benefit from faster processors and "more" technology. Big data (possibly, although cloud systems seems to be handling that). Factorization of numbers, for code-breaking. Artificial Intelligence (although that may be more a problem of algorithms and not raw hardware).

For the average user, today's PCs, Chromebooks, and tablets are good enough. They get the job done.

I think that this explains the longevity of Windows XP. It was a "good enough" operating system running on "good enough" hardware, supporting "good enough" applications.

Looking forward, people will have little incentive to switch from 64-bit processors to larger models (128-bit? super-scaled? variable-bit?) because they will offer little in the way of an improved experience.

The market pressure for larger systems will evaporate. What takes its place? What will drive innovation?

I see two things to spur innovation in the market: cost and security. People will look for systems with lower cost. Businesses especially are price-conscious and look to reduce expenses.

The other area is security. With more "security events" (data exposures, security breaches, and viruses) people are becoming more aware of the need for secure systems. Increased security (if there is a way to measure security) will be a selling point.

So instead of faster processors and more memory, look for cheaper systems and more secure (possibly not cheaper) offerings.