From:
                                                           10/17/97 11:00 PM
Subject: MAX Digest - 16 Oct 1997 to 17 Oct
1997To: Recipients of MAX digests 

There is one message totalling 32 lines in this issue.

Topics of the day:

  1. sprintf abuse

----------------------------------------------------------------------

Date:    Fri, 17 Oct 1997 10:11:25 -0700
From:    Peter Elsea 
Subject: sprintf abuse

I remember when little computers had 8 bit busses, and you could physically
store a byte at a time. The Mac and PC came along with 16 bit busses, and
while some compilers jollied us along with 8 bit bytes as a data type, we
were limited to even addresses, and wasted a lot of storage without knowing
it. Harbison and Steele say "hardware dependent" K & R say "implementation
dependent". Naturally it's the compiler you have to deal with, and the
compiler's quirks come from the hardware (and the marketing department).  I
seem to remember something about a proposed universal character set- that
would have to be 32 bit if it included Chinese.

The symbol/k ratio is inconsistent- the table grows in chunks, and the
memory available is an aproximation anyway. I just cranked in 50000,
occasionally stopping to enter a bunch by hand The first 30 knocked 8K off
the Memory, and sometimes I could get 120 in and the memory would drop 1K.
The entire 50,000 took 1083k, so figure 46,000 per meg.

There was no noticable degradation in performance with 50,000 symbols
stored, but this is on an 8500 running at 130mhz. On those steam driven
machines Nick favors symbol fetching may slow down a bit.

Peter Elsea
Electronic Music Studios
University of California, Santa Cruz
http://arts.ucsc.edu/EMS/Music/index.html
 elsea@cats.ucsc.edu

------------------------------

End of MAX Digest - 16 Oct 1997 to 17 Oct 1997
**********************************************