--- Bill Hooper <[EMAIL PROTECTED]> wrote:
> On 2008 Mar 13 , at 9:52 PM, Ziser, Jesse wrote:
> > ... I got a spec for a piece of hardware-interface software I'm
> > supposed to write. It said that
> > several of the ... quantities are ... in what it called "centi-
> > Amps" ...
>
> Theat's one of the reasons to try to persuade people not to use
> centimetres (but use millimetres instead).
Well, yes, the spec could have said (more verbosely) "tens of milliamps"
instead of "centiamps",
if you really think that would be better (I'm not sure if I do). But if you're
saying the
measurement itself should be transmitted in milliamps, I have to disagree.
As discussed earlier, the actual unit in which the value is sent to the
computer is an engineering
decision and could not necessarily be in mA itself. Bear in mind that this is
one machine talking
to another. The size of the increments used is not arbitrary. It is carefully
chosen to maximize
the useful range and precision of measurement given the bandwidth constraints
of the system. In
fact, it isn't even guaranteed to be decimal; many of these systems end up
transmitting things in
something like "4ths of a degree" or "17ths of a volt". This fact is not
fixable and not
necessarily a mistake.
> The centimetre is virtually the ONLY currently used SI unit that is
> incorporates the "centi-" prefix. True, it is still recognized as
> legitimate SI, but we can simplify SI still further by avoiding (and
> eventually eliminating) the "centi-" prefix (along with deci, deka and
> hecto).
In some environments, sure. In an engineering environment, I'm not sure that
would simplify
things, and it might even complicate them. Engineers (probably unlike most
people) are more
efficient when they can rely on a large, continuously evolving set of efficient
and precise jargon
to make big, complicated sentences shorter.
All my co-workers can be assumed to know the full set of metric prefixes
(they'll look it up if
they don't), and if they need to discuss a concept that has no name (i.e.
something that naturally
happens in units of 72 millimeters for some reason) they will actually make up
a temporary term to
simplify communication ("OK, this word salad is getting hard to follow. Let's
start calling these
lengths 'lambdas'"). I'm all for simplifying the units used to communicate
formally and to
interact with the ordinary public. But for informal engineering discussions
and the like, those
terms probably still have value.
This actually reminds me of a discussion about "code factoring" and another one
about "functional
blocks". The concept of where to simplify communication and by how much can be
analyzed at a
pretty high level and applies to multiple disciplines, ranging from the social
to the technical.
It's a topic I'm interested in, actually, and I have to resist leaping into an
in-depth discussion
here.
> And if it were to be used it would be called "centiamperes" (cA) not
> "centi-Amps".
That was exactly what irked me. Although I think "centiamps" (minus the "ere")
is a reasonable
shortened form for practical tech writing, the hyphen and capital don't seem to
serve any purpose
but to demonstrate the author's unfamiliarity with metric.
____________________________________________________________________________________
Never miss a thing. Make Yahoo your home page.
http://www.yahoo.com/r/hs
