In article ,
"Ed Huntress" wrote:

"Joseph Gwinn" wrote in message
...
In article ,
"Ed Huntress" wrote:


I was once a big advocate for metrics but decades of questioning its
advantages in the marketplace has led me to realize we're doing the thing
that provides the best economic result. The cost of converting would be
far
greater, I believe, than the slight friction it adds to trade. If that
changes, we'll finish converting to metrics, but not until it pays to do
so.

I don't have an opinion on english versus metric in products, but I will
tell one small war story from the software realm:

Some years ago, on a project to build a very large UHF early-warning
radar (these are ten-story buildings with antenna patches at least 50
feet in diameter, with ranges measured in the thousands of miles into
space), the issue of units came up early in the development. I
suggested use of the SI system in the software, and this was accepted.
The rationale was that having a consistent-physics-based system of units
would save money by reducing errors in the software (~600,000 lines of
C++). The exception was that many kinds of data needed by the software
comes in legacy units of measure, so in these cases we convert to SI
internally as soon as possible, so the bulk of the software uses only SI
units.

That sounds perfectly sensible, and I would be alarmed if you told us the
decision was to use anything else. In fact, I find it remarkable that any
science or the loftier realms of engineering use anything else, after most
sciences switched in the 1960s or even earlier, in some cases.

Apparently there are some stragglers.

We were starting from scratch, so we had the luxury of choice. If we
had been upgrading something without replacement, no way could we have
switched.

Joe Gwinn