William ChopsWestfield wrote:

>> you want to be understood, you should use the proper units.
> =

> So you should NEVER use a micro prefix in electronic media unless you're
> certain both your software and the audience's software can properly
> represent and present a greek mu?   That doesn't mean using a mu that
> some random software vendor stuck in a an unused slot of a common font,
> either.  It's got to be something standardized. (unicode is probably OK.
>  How widely implemented IS unicode?  =


I don't know about your work environment, but every current decent online
program understands Unicode. Its more common versions, in addition to a
number of 8-bit ISO fonts (not vendor-specific fonts), are in pretty
widespread use. The charset element of the Content-Type header is also
well-defined and understood by all decent and remotely current and
standard-conform browsers and email/newsreaders.

The Windows NT family of OSes understands Unicode, and so do most
applications running on it. I'm sure most decent Linux/Unix programs
understand it. Mainframe programs from the 60ies or 70ies or DOS WordStar
don't... Ever since I left Win 3.1 behind I've never had problems
exchanging texts with accented characters, or the Greek lowercase mu.
Again, this problem -- if it is really one -- seems to be restricted to the
USA (not sure about other English-speaking countries). =


The common ISO fonts (ISO 8859) are well-defined. Even if some program
should not have support for the specific encoding being used, it is
possible to translate that easily into Unicode (if it's in some sort of
generic text format like XML or HTML).

So far my newsreader tells me that this message uses the us-ascii charset
(default), because I haven't yet used anything beyond 7-bit ASCII. Now I'll
insert that dreaded Greek lower case mu "=B5" and the charset switched to
iso-8859-1 (could have been utf-8 or whatever; that's just the preference
of my newsreader, and it would switch to utf-8 if I inserted a character
that's not in iso-8859-1). This message will get sent out with the
Content-Type header set to something like "text/plain; charset=3DISO-8859-1=
".
Every email and news client should be able to understand and display this
correctly; it's one of the more common standardized codepages. (It's not a
vendor standard, it's an ISO standard, and the Content-Type header is
clearly defined for ages in an RFC.) And if yours doesn't, you should
switch to one that does... IMO of course :)

And even if you have to stick to 7-bit ASCII for some reason, 'm' for
'micro' is simply a bad choice. I'm not sure what really is to discuss with
that... 'm' is 'milli'. That's one of the few universally known
conventions. Using 'm' for 'micro' is asking for communication trouble.


>> The one thing I wonder about this is why it's only in the US
>> that this seems to be a problem.
> =

> Probably because the US was generating reams of printed schematics back
> before whatever metrification we've done, with no mu's on our
> typewriters, back when "m" was no more ambiguous than "u" is now... =


I'm not so sure this argument makes sense. Where did the convention that
'micro' means 10^-6 come from? I'm pretty sure it came from the same place
the convention that 'milli' means 10^-3 came from. So whoever knew about
micro knew about milli (or could and should have known), and also that 'm'
was the agreed-upon abbreviation for 'milli'. Which made it pretty obvious
IMO that 'm' was, even at that time, a bad choice as abbreviation for
'micro'. If you cherish unambiguity, that is. =


Also, the standards I'm talking about have been around for fifty years and
are officially adopted in the USA for at least thirty years. Still sticking
to what has been done 50 years ago... Most did switch to using
semiconductors in electronics by now, even though they did not show up in
the reams of schematics from "back then", from 50 years ago :)

Gerhard

-- =

http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist