23 Feb
2012
23 Feb
'12
8:56 a.m.
On 2/22/12 2:19 PM, "Dave" <dave.g4ugm at gmail.com> wrote:
The very
first 6800 I bought came with a failing carry between the two
bytes
of the 16-bit registers. Very disconcerting as it was the first micro I
had
ever made. However the supplier replaced it no questions asked...
Dave Wade G4UGM
Illegitimi Non Carborundum
I hadn't thought about this in a long time: I once worked on what we would
now call an embedded system, that did video generation for cable TV
channels. The system was based on the 6800, and we learned that there
were particular product runs where you had to insert a NOP before a SEI
(SEt Interrupt mask) or the SEI would fail. It bothered me no end at the
time - it just seemed *wrong* to waste precious working store and
processor cycles!
The file system (on 8" floppies) was also sequential in allocation, so it
was 'best practice' to immediately purge a file you didn't need anymore.
If you forgot, then N new files later someone had to run a purge that
relocated blocks to deal with fragmentation (holes), very very slowly.
The assembler didn't spool its print output, either, so we ran full builds
as seldom as possible, preferring to patch in the EPROM programmer when we
could.
Ah, those were the days?. -- Ian