Gerhard Fiedler wrote:
> It seems you didn't understand at least some of what I wrote. I never
> called anyone "inept".

I don't have a copy of your message and the archive server hasn't caught up
to that message yet, but you said one explanation for lack of unicode was
inept programmers.

> The thing is that sooner
> or later you will need M3 screws, so it /may/ be a smarter move to
> get that done earlier -- before you spend the next $1000 in
> imperial/UTS tools.

Probably not though.  It's not going to be a sudden all or nothing
switchover.  This will be driven by necessity one little piece at a time.

> I don't know -- you tell me. I didn't say that anybody does anything
> stupid.

Actually you did ("inept") and with your general tone of "why don't these
idiots just drop what they're doing and switch to metric already".

> It seems some people in the USA
> think that because during the 50ies the USA became the leading nation
> in science and industry, people in Europe ate bananas from the trees
> before that

Nah, not me.  I didn't think banana trees grew in Europe. ;-)

>> But in reality if it's not ASCII it's either not safe or too much of
>> a hassle.
>
> See, this is one of the problems that only seem to exist in the USA.
> The rest of the world does not have a problem with non-ASCII
> characters and use them on a daily basis.

That's because the rest of the world uses all those silly umlauts and such
;-)

Remember, ASCII stands for American Standard Code for Information
Interchange.  It's not surprising it suits us pretty well.  Early computers
were primarily developed in England and the US.  Umlauts, accents, and 27
different Icelandic characters weren't on the priority list when ASCII was
invented.  I can understand why others might not like ASCII, but it works
fine for me.

>> Unicode is a big pain in the butt because it takes 16 bits per
>> character.
>
> I'm not sure what you know about Unicode, but this is wrong.

It's been a few years since I looked into it, but the Windows API has (at
least used to) have one definition for a normal character data type and
another to support unicode.  If I remember right, it was called WCHAR (for
wide character I think).  Whatever it was called was 16 bits wide while
normal characters were 8 bits wide.  Maybe that has changed in the last few
years (I'd be surprised), but it was definitely true at one time.

>> the really really good reason to switch to unicode is just missing.
>
> Why are you so defensive here? Did I say anything about anybody
> switching to Unicode?

Yes.  You attributed lack of unicode support to inept programmers.

> That was written as a response to a message to Bill. He apparently
> doesn't use IE on Windows. I'm sure you don't like sites that only
> work on Mozilla on MacOSX, right? :)

Possibly, but I don't have to worry about that since no serious web designer
is not going to not test his web page with MSIE.

> It was difficult for every single one of the companies and
> people in Europe when they did it fifty years ago. There were people
> with tools in their basements in Europe before 1950, just as in the
> USA. There were engineers that grew up with different measurement
> systems, just like in the USA.
> There were huge amounts of company history and blueprints in other
> units, just as in the USA. The only difference is that they did it
> back then, unlike their counterparts in the USA.

That doesn't make a synchronized sudden switch the right answer, nor does it
prove that it was the right answer for Europe back then.

> Here's the fallacy. It's more expensive to keep the dual system, and
> the amount wasted on that adds up every year it will take to move
> over. It's as simple as that. The conversion cost may show up in the
> budget of a company as "$1.000.000 spent once on metrication
> efforts", but the cost of keeping the dual system in place won't show
> up as "$100.000 per year spent on multi-measurement system support".
>
> So yes, the short-sighted advantage is with staying with what works
> and is in place. Judging management by quarterly results reinforces
> this. But
> since the move is inevitable (and I'm not sure you agree, but you
> seem to), it's simply a matter of continuing a waste of money to
> delay the move, in many situations.

This is the part of your argument that pisses me off.  You are basically
saying that everyone who doesn't switchover wholesale now is stupid.
There's a lot more to it than "simply a matter of continuing to waste
money".  It really comes down to two things, available resources and return
on investment.  If you don't have the $1M to switch over, then it's not an
option whether you'd like to or not.  But let's say the funds are available.
It then becomes a question of "how can I spend this $1M to get the best
return?".  In your example, the ROI is about 10 years.  That's very long.
If that $1M can instead be spent on something that yields a ROI in 18
months, then that's a much better investment.  The world is never perfect,
but you have to intelligently chose what to address and what to live with.
In many cases the cost to switching to metric is substantial and the payback
small.  Without careful analisys you can't say whether switching makes sense
or not.

This is why the switchover is going to be slowly phased in, as has already
been done in getting to where we are, which is over half way in my opinion.
It will be driven, as it has been, by many individual decisions about supply
and demand and market pressure.  And that's exactly how it should be.  Let
the capitalist system do it's job.

> (Just as the additional effort you spent -- on your side and on your
> associate's side who you taught that lesson -- by teaching him that mF
> means millifarad. This was a cost factor, and you could have avoided
> it by not using mF and sticking to uF, as Bill suggests. You didn't,
> you spent
> the effort, and you did it because you thought that's the right thing
> to do and that it's cheaper in the long run.

No, it probably wasn't.  However my cost was small (the supplier had to fix
it, not me), so I could afford to base the decision on ideology instead of
practicality.  Also remember that it wasn't planned that it work out the way
it did.

> The other thing is: if the US automotive (and a few other) industries
> think that it's cheaper to go metric (they definitely had to convert
> /a lot/ of existing documentation and change stock), it may not be so
> far off to think that it'd be cheaper for many others, too, and that
> some of the main
> reasons they didn't do it are just not thinking about it, and
> sometimes mental (and other) laziness.

There you go again.  Now people who don't switch because you think they
should are "lazy".  You are once again ignoring the many good reasons not to
switch.  Companies will largely do what's to their advantage.  I can
understand why it's to the advantage of large heavy-iron companies to switch
to metric, but that doesn't automatically make it a good idea for everyone
else, at least not yet.  That will come as the switch trickles down.  Some
have enough to gain now so they switch.  That may make it attractive for
their suppliers to switch when it previously wasn't.  Now the tradeoff is
tipped a little more for their suppliers, and so on.  That doesn't make
those who haven't switch lazy or stupid.


******************************************************************
Embed Inc, Littleton Massachusetts, (978) 742-9014.  #1 PIC
consultant in 2004 program year.  http://www.embedinc.com/products
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist