🚀 go-pugleaf

The way the Finder initially calculated disk space available was not a bug,
but a "feature" (some of you have heard this before!)  Steve Jobs said that
people wanted to see 400K available on blank disks, and that I should make
it so that was the case in the Finder.  Since the original single-sided
floppies held about 410,000 bytes, dividing by 1024 left exactly 400K;
subtracting a few K for the DeskTop file made it less than 400K free.
Dividing by 1000 instead of 1024 was basically a marketing decision, and an
apparently defendable one:  most naive users understood a K to be 1000, not
some power of two near 1000.

I imagine that Apple would have gotten more phone calls from naive users who
wanted to know why they didn't have 400K available on a disk than from
computer-sophisticates who wanted to know why the Finder was "wrong".
Debatable, to be sure.

On something entirely different:  given the latest version of LightspeedC,
why would anyone want to use MPW?  MacApp is the only reason I can think of,
but someday LSC will have support for objects.  (MacApp is a very nice
package, by the way, and I would probably be using it if I could.)  It seems
to me that MPW will never be able to be as fast as LSC, but that LSC could
well end up with all the features of MPW that you wanted (e.g. tight code
optimization, maybe as a compiler option, and support for the
object-oriented programming style).

-- 
Bruce Horn, Carnegie Mellon CSD
uucp: ...!seismo!cmucspt!cmu-cs-vlsi!blh
ARPA: blh@vlsi.cs.cmu.edu