[net.micro.cbm] 1541 @replace Bug.

dillon@ucbvax.ARPA (The Sherif "Matt D.") (01/12/85)

	There does exist an @replace bug, But it's very obscure.  It is
not caused by illigal scratching or uninitialized disks, as it used to happen
to me before I stopped using it.  Of course many terrible things happen due
to other errors.

	A word of warning... If you plan to use ROM routines in the 1541 BEWARE
Considering that there are already 5 ROM generations of the 1541, and 
probably will be more, The addresses you use inside the 1541 will change.  You
cannot realisticly expect your program to work on other C64's with 1541's.

	The only thing you really can rely on in terms of the 1541 is the I/O. 
A program to read data from a track which does not use any ROM routines works
on ALL generations, as well as the 2031.  But, you must remember several things
when reading raw data directly from the disk:
	The data is encoded... You have to decode it as well as worry about
	 sync marks.
	The data density changes depending on the track your on

	Unless your a hacker, and already have delved into the 1541's ROMS,
	your never going to understand the technical aspects of the 1541.


The best sector transfer rate I've been able to get from the 2031, and the
best sector transfer rate Bryce has been able to get from the 1541 using
his FLASH! has been a skew factor of 4 ... About a track a second in terms
of program loading time.  This was accomplished by downloading a 3-page
program to THe 2031 (1541) which did everything manually and did not touch
ROM routines, which then transfered the data non-standardly to the host 
computer (PET/64).

					Matt.