Additional GCR Notes
That Didn't Make It Into The Manual
I. GCR Adjustment: What to do if it doesn't work.
This explains a little about the GCR, and tells how
to adjust a potentiometer or recable the system to
eliminate many common problems.
II. Things to also check.
These are practical tips from
people already using the GCR, about problems they've
run into and fixed. If you're having a problem, it's
likely that it will be discussed in here.
III. The GCR self-test software.
This tells you how to operate the GCR self-test software
from the hidden, "hi dan" backdoor in the "Spectre"
menus. This software will greatly help you track down
bugs by exercising the GCR's every function.
Note: Most of these other sections eventually refer to
the self-test software; this chapter tells you how to run
the self-tests.
IV. THE SELFTEST THAT EVERY GCR OWNER SHOULD DO.
This tells you an initial check you should run, verifying
that your GCR can read the Mac PD disk we shipped with
your unit, and also tells you about the very useful
programs on that Mac PD disk, things like virus checkers.
.pa
I:
GCR Adjustment:
What To Do If It Doesn't Work Right Off
Don't Panic!
First off, don't panic. We realize you've probably waited a
long time for the GCR, and it's irritating to have it not work
the first time. Rather than call us and complain, there are
several things you can do by yourself to check the unit out, and
probably fix it, all saving yourself an expensive long distance
phone call. Ok?
Also, the majority of GCR units work "out of the box", no
adjustment necessary. But should you have a system that needs
adjusting, this text will tell you how to do it.
First off, look for the obvious, okay? Trace out the disk
connections; all the wires go somewhere, right? You are plugged
into the right spots on the disk drives; looking from the top,
with the front of the drive towards you, the right hand
connector of the disk drive is TO the drive, the left hand
connector (by the switch) is FROM the drive. You must daisy chain
the drives in that order; the connectors are NOT reversible.
Note: the connectors on the GCR *are* reversible. But not on
any Atari drive; that's how Atari makes drive B into B.
Are your disk drives all on? Can you access both of them from
the normal Atari, in Atari mode? Try formatting a disk in both of
them.
Is the GCR plugged in? Is something plugged into the GCR? Is
the GCR straight-in to the cause, not leaning out at an angle?
That's easy to have happen.
Everything looks good, but the GCR is giving trouble, eh?
Okay, let's go fix it.
Adjusting The Potentiometer
Looking down on the GCR circuit board from the top, one chip
away from where you plug in the disk drives, is a small
potentiometer, also known as an adjustable resistor.
If It Works, Don't Adjust This
This resistor must be carefully adjusted IF your GCR doesn't
work properly and IF other fixes don't help. Don't adjust it if
you don't have to; we set it to a known good position at our
factory complex here in Gadgets, Colorado, and it will work with
most known Atari disk drives in that position. If you alter it,
it may be hard to get back to exactly where we had it.
I'm very hesitant to tell you to adjust this resistor without
explaining what it does. Maybe it's the old author-bug in me that
can't resist explaining things? But I feel it will make far more
sense to you if you know what you're adjusting, as opposed to
telling you "set it to position 1, run the program, set it to
position 3..." and so forth.
A potentiometer is a "variable resistor". The "variable" means
you can adjust it. The "resistor" means it's a device to impede
the flow of electricity through it.
This potentiometer is located on the "Write Data" wire going
from the GCR to your disk drives (and Atari, by the way).
Lesson One: THIS RESISTOR ONLY AFFECTS GCR WRITING. IT WILL IN
-NO- WAY AFFECT GCR READING.
Here's a picture of the process:
GCR writing data >----> variable resistor ----> disk drive(s)
The potentiometer is exactly like the volume control on a
radio (a "radio" is a primitive form of "compact disk", to you
programmers.) It controls how loudly the GCR talks to the disk
drives when writing to them.
If you set the potentiometer to fully-on, the GCR "talks at
full volume" to the disk drives; there is no resistance in its
way. If you set the potentiometer to "fully off", the GCR
"whispers", since a lot of the GCR's signal is limited by the
resistance.
Now, at first glance, you'd think that you'd always want the
GCR to yell at the disk drives, at full volume. That's what we
thought, too. Sadder and wiser, we bring you the news of a sad
truth: Reflections and Ringing.
Reflections on Ringing
Electrical wires have this thing about devices like the GCR,
which shout loudly and quickly. They tend to resist the shouting.
Literally, what happens is instead of the electricty just going
through, each pulse out of the GCR makes the wire "ring", exactly
like ringing a gong. Instead of your disk drive hearing a
"pulse", it hears a "Gonng-ong-ong-ong-ong-ongg". Some drives
interpret this as "pulsepulsepulsepulse", and hopelessly foul up
whatever data they're writing.
Part of this is caused by the Atari itself. Our pulse goes out
of the GCR, on the "write data" wire, into both the disk drives,
and back into the Atari itself. Once inside the Atari, it deftly
bounces back, a zillionth of a second later. This is known
technically as "reflection". There's little we can do about it
short of changing the ST's design; the ST just was never designed
to have someone else talking to its disk drives.
So, what can we do about it?
Well, first off, this isn't as bad as it sounds. Sometimes the
ringing makes no difference to your disk drive; your drive
listens to the ringing, says, "hunh, sounds like ringing to me",
and ignores it. This varies a lot from brand to brand; in
general, the *better* a drive at reacting to high speed
transients, the more likely it'll react to the ringing. For once,
quality causes problems. Since Atari stays away from the more
expensive mechanisms, ringing isn't that bad a problem.
Second, since this is a "resonant phenomena" (e.g., a gong
getting rung), changing things like where the GCR is in the
cabling scheme can make all the difference. You may be able to
halt the ringing completely. For instance, a long drive cable can
solve a problem. Putting the GCR at the very end of the cable
chain is generally a very good idea, but, if this doesn't work,
try putting it right in the middle, like this:
Atari ----- Drive 1 ------- Drive 2 ------ GCR
Atari ----- GCR ----- Drive 1 ------ Drive 2
Believe it or not, I've also seen this work:
Atari ----- Drive 1 ----- Drive 2 ----- GCR -----
(e.g., let a cable dangle off the end).
Anything that adds or subtracts to the cable length will
affect the frequency of the ringing, which in turn will affect
whether or not it screws you up. At some frequencies, the ringing
will tend to self-cancel; at others, it'll be worse.
If you've got a ringing problem, try changing things around as
per above and it might go away.
Also, if you add or remove a disk drive from the chain, its
ringing frequency will change; you're also adding/removing
"termination" on that drive, which makes a big difference.
So, if your GCR is giving you troubles (e.g., you write to the
disk and find what you wrote is hard to read back -- the Mac will
automatically detect this and let you know), the first thing to
try is rearranging the drive order.
The second thing to try is readjusting this potentiometer.
What we're trying to do with it is "soften the blow" that the GCR
makes to the ringable disk-write wire; if the blow is softer, it
doesn't gong at all, or rings much less loudly. The data still
gets to the drive okay; it just doesn't throw its weight around
on the wires so hard they ring.
There is a careful balance to be maintained here.
You can't set the potentiometer to zero. If you do that, your
ST may (MAY) work okay in GCR mode, but has a very big chance of
fouling up on regular old ST mode disk writes/formats.
You can't set the potentiometer too high, either. If you do,
too little signal will get out of the GCR to the drives, and the
drives will react erractically or not at all.
I've found that in general the adjustment range is from 1/4 to
3/4. In a certain few instances, going beyond that is called for,
but almost never.
What we're going to do is assume that your GCR is having
trouble, and walk you through the adjustment procedure. We can't
predict what your system will do, since ST's and their drives
vary so widely. Instead, we'll tell you what to do. The minute it
starts working, quit; you don't want to overadjust that
potentiometer, because it can wear out and become flaky.
Adjustment WalkThrough
The resistor adjustment is as follows: Counterclockwise all
the way is zero resistance. Clockwise all the way is maximum
resistance. We adjust for halfway here at the factory labyrinth.
The best way to find out how the GCR is doing is with the
built-in testing software; it repeatedly formats, reads and
writes in both Spectre and GCR mode, and tests the whole layout
extensively. To get to this testing software, double click on
SPECTRE.PRG with the cartridge installed and hooked up. Type
hi dan
("hi (space) dan"), no return afterwards; if you make a
mistake, or nothing happens, then pull down any menu, click on
any selection, and try again. This is Dan's Famous Password.
You'll then have a menu of GCR fine-tuning items. Essentially,
this menu lets you exercise the GCR, ST, and disk drives without
going into Mac mode, and lets you do it repeatedly, to help
during adjustment.
What we want to do is format in ST mode, to make sure we
haven't screwed up ST mode (as with the potentiometer at 0), then
format in GCR mode, verify that format went okay, write to the
disk in GCR mode, and read back what we wrote in GCR mode.
We have several forms of this basic test; when your machine
starts passing this test repeatably, you're all set.
There's the first basic test, the "T" (capital T). This runs
your machine through the above cycle, and stops if there is a
problem.
There's the quicker version of that test, "t" (little t). This
runs your machine through the above test, but only does 10 tracks
on the disk, not 80, and is thus much faster, while basically
exercising the GCR / ST / Drives.
There is "C", a cyclic tester; it just does the T-testing
(full disk) over and over forever, keeping a score of what worked
and what failed. There is also "c", which does th quick test over
and over, keeping score.
What you want to do is run the "c" cycling quick test while
fiddling with the potentiometer to get yourself into the
ballpark. Once things start working, do the full cycling test
("C") and let it run for an hour or two, and see how the score is
after quite a few tries.
So, get a small screwdriver to adjust the potentiometer. You
should find it set at halfway. (There's a funny dent in one place
on the screwdriver slot that shows you where you are, but I don't
know how to describe it.)
Then, begin the "c" cycling quick test.
If your unit cannot even GCR format, try nudging the
adjustment down (counter clockwise) a little bit at a time,
letting it get back to GCR format each time. You may also have to
nudge it up past halfway. Take your time; you don't want to
overshoot and miss a good setting. Some drives can be really
touchy, I'll tell you.
Once you get it to GCR format, explore a bit the range it can
GCR format. It may do it for the full travel of the adjustment,
or just a tiny bit. Get some feel for how much adjustment range
you have.
Next, let it get into the read/write portions of the test.
Essentially, you want to be doing read/writes on a disk that was
formatted with the adjustment where it is now, so you'll have to
wait (this is why the quick test is handy).
Again, I can't predict at where your system will come alive,
or if you'll need this adjustment at all; the majority of ST
drives work just fine with the GCR as shipped.
Probably the most important thing to remember is that the
adjustment can be touchy, so go slow, and watch the results of
the cycling quick-test.
Once you get the thing pretty close, try the full test. This
takes awhile (that's why there's a quick-test). You may want to
nudge the adjustment up and down just a hair if you still find
any glitches.
Once you think you have it, let the machine go on full test
for a couple of hours. A properly working setup should have
perhaps 2 or 3 failures, most likely in "GCR format verify";
these are NORMAL, just soft errors during format. (Most drives
automatically retry a format once just because of these). There
shouldn't be any other errors.
If ST formatting (e.g., Spectre format) goes down, then your
potentiometer is too low, and the GCR is interfering with your
ST's ability to talk to the disk.
***********************************
II: Things To Also Check
Mega Internal
There have been reports that Mega ST's sometimes have
interference on their internal disk drive. What it boils down to
is a lot of radio noise inside the Mega's shielding being picked
up the internal drive and the wiring to that internal drive.
Mark Booth on GEnie constructed a shield out of aluminum foil,
and later, a soft-metal baking pan, and put it around the wires
at the back of the drive. This changed his internal drive from
flaky to rock-solid.
Mark reports that before he fixed it, his drive had problems
mostly starting at track $40 (that's track 64 to you and I, but
40 in hexadecimal, or computerese. The "Orwell's Monitor" display
of the GCR will show you the track number in hexadecimal, by the
way; if you read and write to a disk, and the GCR stalls at "TRK:
40", then this is likely the problem.
Monochrome Monitors
The monochrome monitor (SM314) is exceptionally noisy; it
emits a great deal of radio "noise".
Your disk drive has a coil in the disk read/write head which
is a perfect antenna for this sort of noise.
Hence, the closer your monitor is to your drive, even with
shielding, the more likely it is the disk will pick up the
monitor's noise, and start to lose the (relative) whisper of
noise coming from the disk in the barrage of noise from the
monitor.
I have found my internal Mega drive works fine -- if I keep
the monitor more than 6 inches away. My internal drive is
shielded and so forth; it's just the mono monitor kicks out so
much noise that the shielding can't keep it all out. (No
shielding is perfect!)
You might want to try putting your machine into continual test
mode (c) then moving your mono monitor closer and farther from
your drive. If there's a shielding problem, it will be
immediately obvious; your drive will get massive read/write
errors, or fail completely, when the mono monitor is close.
The solution is basically up to you. Moving the mono monitor
away helps; shielding the drive helps. I'd say that for most
people moving the monitor is far easier.
Again, this is a known problem. It also applies to external
disk drives.
RPM
The GCR is exceedingly tolerant of off-RPM drives. For
instance, we have Mac disks that the Mac can't read, but the GCR
can, that were apparently made on a bad Mac drive.
HOWEVER, if your drive is way whacked-out, you'll have
problems eventually, not only with the GCR, but talking with
other ST's.
We include a speed tester program for you to check; Atari disk
drives should be right at 300 RPM, with the tolerable range 1%,
or 3 RPM, thus: 297 - 303 RPM is just fine.
You will find that over about 308 RPM you can't even format 10
sectored disks anymore (e.g., Spectre format).
Again, if it works, don't worry about it. If your RPM is far
off, say, 290 RPM, you'll want to get this fixed; it will affect
your ability to exchange disks with the rest of the world.
Alignment
Is your disk drive just plain out of alignment? There's all
sorts of esoteric adjustments involved here. If your ST can't
talk to other ST's, it's likely to be out of alignment; if so, it
cna't read Mac disks either, which are supposed to be aligned to
the standards for 3 1/2" floppies.
Your dealer can test this.
Another not-as-certain-but-prolly-good-enough test is to use
the GCR's diagnostics menu (the "hi dan" menu) to Read All
Sectors on the supplied Mac PD disk. That disk is known good
alignment; if your GCR can't read it, you've probably got
problems.
Cables?
Are your cables good? Some Atari cables made in the
aftermarket are incredibly shoddy -- cold solder joints and the
whole bit. You may want to try swapping cables with a
friend/dealer and see if that helps. Also note that just changing
cable length will affect the resonant frequency of the cable, and
may solve your problems right there. Again, if this solves it,
don't try to fix it anymore.
------------------------------
III. The Great And Powerful GCR Test Software
GCR selftest has a number of features.
You Must:
* 1) Use the NEW versions of DRVR128.PRG and LAUNCH.PRG. You
know and I know that DRVR128 wasn't used for much before version
2.0; now it is. Things won't work without it.
2) Have a GCR with ROMs plugged in. Otherwise the software has
no way of knowing whether or not you should be allowed into the
GCR self test software.
So, go into the Spectre program, type "hi dan" (where DOES he
come up with these open-sesame passwords?). If you blow it, pull
a menu and click something, and start over. If you press RETURN
you will likely launch into Spectre mode, So don't.
SelfTest Functions:
Going through the functions in the GCR Self Test:
(From bottom to top, since it makes more sense that way)
"q" quits and returns you to Spectre. The MENU part of the
screen WILL NOT replot. This is not a bug, this is a feature to
help you test your memory. (Actually we don't know how to get the
silly thing to replot.)
Note: WE RECOMMEND REBOOTING if you're going back to ST mode.
There could be low memory tables screwed up that we missed fixing
back up; getting the GCR to work involves massive coronary bypass
surgery on the ST, and undoing that is always expensive.
Next, "r" reads a single sector, GCR only, from
track/sector/side/drive you pick. It doesn't read a Spectre
sector, although it might say it does -- you can read a
Spectre sector with any ST disk editor, like TinyTool. "r" reads
it, then lets you know if it could read it, and dumps it out in
hex.
Now ALL numbers you enter must be in hex, because of the I/O
package I use for getting and putting digits.
"w" writes zeroes to any sector you like. (Remember you'll
blow away the disk's data doing this.) It is more for testing the
ability of the GCR to write to the disk at all, WITHOUT damaging
the disk, than to write any specific data. There used to be big
problems where writing to the disk would kill the next sector on
the disk; "w" was meant to catch these. It isn't used much
anymore.
"R" (capital R) reads ALL sectors on the disk; you can select
either Spectre or GCR mode. This is a quick verify that all
sectors are good, and is remarkably exactly like the verify in
the main Spectre menu, except that Read-All keeps going after an
error, and Verify quits after first-error.
You can thus find the error on a just-failed-format disk with
this tool.
NOTE: For some weird reason, sometimes you try formatting a
disk once, it fails on verify, you try again, it works. This is
just because disk drives are imperfect, like their programmers.
Don't let this bother you.
Read-All plots a line-feed (screen goes up) and a "-" whenever
it loses reading a sector. A really bad disk will look like a
canyon. You'll see what it looks like first time you run it.
"W" (capital W) writes ALL sectors to a disk. This of course
blows away all data on the disk. The idea is to test that you can
write to the whole disk and not damage the disk structure.
"f" formats a disk, into either GCR or Spectre format. This is
the Very Same formatter in Spectre main menu. It can really only
fail in the most horrid of conditions, where it formats a track,
then can read nothing of what it formatted (error code = fedcba98
in this case). It'll likely "format" a really bad disk, since it
is not doing a verify.
Note that a disk formatted with "f" is totally blank. The
necessary stuff to make it into a Spectre or Mac disk (directory,
volumne info, and such) is not there.
"v" verifies the disk. This is a read-all-sectors, except
that on the first bad sector, verify stops -- it's telling you
the disk didn't verify.
"a" to adjust Oscarlater doesn't work, and will spazz out your
machine. It was for something else and we forgot to take it out.
Oops.
The Quick and Long Tests
The "t" and "T" tests do several operations in a stream,
stopping if something fails. The difference between them is that
the "T" test tests every single spot on the disk, taking quite
some times; the "t" tests 1/8th of the disk, going a lot faster,
but basically testing all the GCR functions. Both are useful in
certain cases.
There are 80 "tracks" on a disk. "T"est tests all 80. "t"est
tests every 8th, or 10 total tracks.
"t" for quick-test does this to every 8th track across the
disk:
1) Formats disk Spectre mode and checks track to make sure
this worked. This "wipes the disk clean" of any previous GCR
format; one problem with testers is not getting rid of previous
results.
2) Formats tracks GCR mode and read-checks these tracks.
3) Writes to every sector on the tracks, GCR mode, and goes
back and read-checks those sectors. This ensures writing to the
disk works, the oscillator is set, the sun is in its place, and
will rise tomorrow morning, and stuff like that. It
Hence, if your unit passes this entire test, it's entirely
functional.
We do recommend repeating the tests many times, since
intermittent bugs might only show up every once in a while. So we
included automatic-repeat versions of the test, called Cyclic
Testing (since they cycle over and over). "c" does the quick test
over and over; "C" does the full test over and over.
"c" for cycle through quick test, "C" for cycle full test:
This is an EXTREMELY USEFUL option. What it does is the above
quick tests over and over and over until you press RESET or the
system dies. It keeps running statistics on just how many
success/fails you got during each step. This is a wonderful thing
to let run all afternoon while playing golf or swimming or
whatever you do all afternoon.
You can diagnose bad drives this way, bad disks, bad karma,
and "I'm bad" albums too.
This tests all GCR functions and just hammers the living heck
out of the hardware; if anything at all is broken, you'll know
all about it from this test. Let it run awhile.
On our machines, we usually got around 20 passes on the test,
4 fails during GCR-format-verify pass (a simple "soft hit" during
format), and maybe 1 Spectre mode format failure on the first
startup due to GEM screwing things up.
Hence, don't be worried about a FEW GCR-format-verify
failures; this is normal for disk formats.
I would not recommend running this all night; that might
overheat and fail your drive or the disk.
Note this also tests the GCR's ability to handle warming up,
which CONSIDERABLY changes our hardware, tests its ability to
handle varying RPMs as your disk drive warms up, and tests its
reaction to varying RPMs as the result of disk media heating.
We have tested from 295 to 305 RPM and found All Is Well with
this formatter; above 305 is all-bets-off, since that's getting
into 10-sector-mode-dies-land, and you're not in Kansas anymore,
Dorothy.
Factory spec on drives is 300 +/- 1% or 3 RPM, but we know
better then to rely on that.
Macs are able to read/write to disks formatted within this
range, we have found. However, we don't guarentee that
READ/WRITING to the disk, EXCEPT GCR FORMATTING, will work
outside this range. (GCR FORMAT is exceptionally tolerant, good
looking, has cute eyes and a sincere smile. Read/Write are
intolerant, ugly, wear hard contacts that make for red eyes, and
have missing teeth. In other words, they just can't compensate as
well as Format for exceptionally screwed up drives.)
-----------------------------------------------------------
A few notes on retries
When you're in Mac mode, if you notice your disk drive taking
longer than usual to do something, you're probably going through
"retries".
We have quite a bit of experience in how to help a drive read
a marginal disk -- or help a marginal drive read a perfectly good
disk.
Thus, when you get a GCR-READ failure, we don't immediately
pull the fire alarm and die. Instead, we go through a careful set
of retries designed to give you every chance of getting data off
that disk.
You'll notice a lag time when you put a blank disk into the
machine. This is the GCR furiously going through its list of
things to do during retry; it's doing its best to read the blank
disk, because for all it knows, it's a good Mac disk.
Here's what the GCR does:
1) Hysteresis correction. On most disk drives, if you land on
a track coming from one direction, and land on the same track
coming from the other direction, you don't land in exactly the
same place. Disk drive manufacturers wish you did, but often you
don't. This is called Hysteresis.
Hence, the first thing we do is go off-track, then land again
on the track you want from the other direction, because it's
entirely possible the machine that wrote this disk had hysteresis
problems.
2) If that doesn't work, we do a full "recal", or pull the
head back to track 0, then go back to your track. This ensures
we've hit the track from both directions for hysteresis fixes,
and also ensures we're not off some tracks. For instance, on some
drives, opening and closing the door will jiggle the head off
track, and you'll get a disk error next time you use it.
"Recalibrating" the disk makes sure you're on the correct track
number.
3) If that doesn't work, we begin adjusting the speed we read
or write data at to (hopefully) more closely match what's on the
disk. Remember, that disk was formatted on one machine, and may
have had data written to it by many other machines, over time;
that means lots of variation in the disk's data rate going just
from sector to sector.
Hence we adjust our circuit up and down several steps, re-
reading the disk each time, seeing if we're improving the
situation.
4) Finally, if all this fails, we conclude the disk is
unreadable. If you have a GCR, you're given the option of
formatting the disk into GCR (Mac) mode, single or double sided.
In our long experience with the GCR, and in particular, when
we've deliberalte fed it known-bad disks, we've found that this
error recovery is very effective. It will greatly slow down the
GCR if the GCR has to resort to these tricks to read the disk,
but that's a far better thing than losing the data!
In particular, if you run into a disk that's painfully slow to
work with, copy it to a new, freshly formatted disk, and see if
that doesn't cure the problem.
Note: Macintosh II formatted diskettes are different than
Mac/Mac 512/Mac Plus/Mac SE diskettes in their "interleave
pattern". Reading from all these disks will be at the same speed;
writing to a Mac II format disk can be slow, depending on several
factors. If you have one disk that's inexplicably slow, that
might be the cause. We have a solution for this, but must wedge
it into a fairly tight space.
Disks formatted by GCR stick with the older-Mac pattern, not
the Mac II pattern.
NOTE: Turning on the "Orwell's Disk Monitor" (shift up arrow)
WILL slow down disk writing terrifically! The time spent updating
that display causes you to "miss the next sector" and slows
things down; instead of 6 writes happening per spin, only 1 does.
NOTE: The "cache" is shut down for GCR operations. This is
because we WANT all GCR writes to be physically re-read from the
disk, so that at write-time (say, when saving a file), you find
out right then if the file is re-readable. With the cache on,
this would not have happened. Spectre mode still uses the cache,
because its disk writes are so bulletproofed.
***************************
IV:
The Self Test Every Owner Should Do
Okay, let's assume that you've installed the GCR as per the
manual. You can perform some tests on it to assure yourself that
All Is Well. Caution: in order to access the self-test, you will
have to have ROMs correctly installed in the GCR; an "empty" GCR
won't go into selftest mode.
Go Ahead ... Read Our Disk
We included a formatted, double-sided Macintosh disk in the
package, for the SOLE purpose of giving you a known good disk to
test your drive against. If your drive will read this disk, then
lots of things are good -- the drive alignment is ok, RPM is
probably close, cabling is basically there, and the GCR's read
circuit is operating. But if something is broken, you can "work
backwards" to figure out what, using this disk. For instance, you
can swap drives, cables, try a friend's GCR or drive or computer,
and so forth, because you know the disk is fundamentally sound.
Since we were including a Mac disk anyway, we filled it to the
gills with various exceptionally useful utilities. Let me briefly
mention here that Disinfectant is a superb anti-virus tool, and
if you don't think you need it, you probably do; the nVIR and
SCORES viruses are all over the place, even creeping into
commerical software releases and CD-ROMs where the manufacturer
should know better. We see viruses ALL the time in mailed-in
disks "that have some problem I can't figure out". Just because
the ST has been *relatively* virus free doesn't mean the common
Mac viruses will leave you alone.
Vaccine is a "CDEV" that will warn you when a virus tries to
do its thing (at least, the viruses known when Vaccine was
written -- but it's still effective against the worst offenders).
CDEV's require Finder 6.0 / System 4.2 at minimum.
We also included goodies like Term-Plus, a nice terminal
program, and some system status CDEVS.
Generally, this software is shareware -- send the requested
donation (see the program / documentation for amount and address)
if you use the software. It's okay to distribute it, and we're
not charging you anything for that disk (well, it costs us a
duplication & blank disk fee, but that's only a couple of
dollars).
There's two tests to do. First, you should be able to read in
all the files from this disk in Mac mode. Note: This is a double
sided disk; if you only have single sided drives, skip this test.
Now if your GCR is already in and working fine by the time you
read the manuals, which is how half of humanity works, then fine,
go ahead, read our disk. You should be able to read in the whole
thing, every file, with no problems at all. If your system will
do this, you're far ahead of the game. (Writing is still another
issue, but you'll have proven a lot of the stuff common to
writing works, like the drive.)
If your GCR hiccups ("File could not be read/copied"), then
it's time to do more testing to isolate the problem.
Second test: You should go into the previously mentioned GCR
diagnostics menu ("hi dan") and do a Read-All-Sectors, GCR mode,
double sided (single if you only have single sided drives), and
make sure the GCR can physically read the entire disk.
What you should see is your system reading with no problems.
If you do see a problem, it's time for some further testing to
isolate it.
For instance, if you yourself can format and then read-all a
disk with no problems, then probably your drive can read your
disks, but not ours. That's RPM or alignment problems, at a
guess.
I'd suggest becoming familiar with the GCR self-test menu and
running many tests with it.
=======================================
Last-Minute Information for Spectre GCR
Development of the Spectre GCR proceeded so quickly that we
left the manual behind. Hence, we'd like to mention a thing or two
you'd like to know about.
First off, PLEASE look over the manual for 2.0, even if you're
a professional-quality Spectre Hacker. Things have changed; there's
new options and so forth.
Foreign Laser Printers
* Foreign Laser Printers (SLM804 non-US): the reason Spectre
has a problem with them is the different paper size triggers a sensor
inside the SLM804, causing all sorts of problems.
For now, you can single-feed pages into the SLM804 and they'll
work fine in Spectre mode; this bypasses the paper sensor. The paper
tray has to be pulled out.
We found out the reason for this problem two days ago, and did
not have time to update the software appropriately to permanently
fix this problem.
System Configuration Change
Whenever you select "SAVE SETTINGS" from the Spectre menu, a file
named SPECTRE.CNF is written to the disk; this holds your current settings
for such things as memory size, hard disk configuration, and so on.
If you make a system change (say, add a hard disk or RAM), or
change ST's (for a demo?), you will *have* to delete the old SPECTRE.CNF
file. Otherwise, when the Spectre starts up and reads the old SPECTRE.CNF
file, it will become damaged in memory and not function correctly. (You
can't just start up Spectre and do a new SAVE SETTINGS; you must delete
SPECTRE.CNF, start up Spectre, and do a new SAVE SETTINGS.)
You cannot just do a new "SAVE SETTINGS" to fix this.
Shutdown / Restart menu options
These are options from the Mac "Finder" "Special" menu.
The "Restart" option does not work. It puts the ST to sleep; only
a RESET or poweroff/poweron can wake it up. It won't wake back up, as a
Mac will. (The reason: it executes a 68000 RESET opcode, which puts all
the ST hardware to sleep.)
The "Shutdown" option may be working now. We're not really sure.
The question is if Shutdown successfully updates hard disk directories,
by "ejecting" them, before turning off the system. We don't have the answer
yet.
For now, it is safer to "throw away" all disk icons (this
merely ejects the disks; it doesn't throw away the data!), then turn
power off or RESET, than to use Shutdown.
Erase Disk menu option
We don't really recommend you use this.
If you use this off a hard disk, you will almost certainly damage
the partition; you may end up with a 400K sized MFS partition, regardless
of what it was before (15 megabytes?)
It is safer to use the "Format Floppy" option from the Spectre
front menu on floppies.
Erase Disk will probably fail on Spectre format floppies, and only
has a chance of working properly on GCR (Mac) floppies, but we still don't
guarantee it.
In-Line Formatting
When you put an unreadable disk into the Spectre, you
will eventually get a menu saying, "This disk is unreadable; would
you like to initialize it, or eject it?" You're given the option of
initializing it single or double sided.
This is "inline formatting".
This should now be working properly. However, it ONLY works in
GCR (Mac) disks -- because we haven't yet found a way to change the
menu to have a Spectre / Mac format selection.
Mac Plus Mode
There is an internal "flag" that tells the Mac whether or not it
is running on a 512kE machine or a true Mac Plus.
We default to 512kE; however, you can toggle to "Mac Plus Mode"
by pressing ALT-keypad-plus if you wish. ALT-keypad-minus reverts to 512kE.
One place this is fun to do is in the Key Caps desk accessory; you
will change to a Mac Plus keyboard.
Some software may only run in true Mac Plus mode, although most is
smart enough not to check this flag. We provide this option just in case.
6.0.3's up
System 6.0.3 now works fine. 6.0.4 is brand new and untested, but
it ought to work. 6.0.3 was a very minor bug indeed, as it turned out.
6.0.2 / HyperCard
This works now.
6.0.2 Sound
This doesn't work. The reason is Apple drastically changed sound
under 6.0.2., to reflect the brand new sound chips in the Mac II. We've
not yet figured out the reason it doesn't do sound in the older Mac Plus
manner we support. (... all the documentation *says* it should work ...)
Sound On/Off
Version 1.9F used to have big problems in leaving sound on too
long. This has been fixed. It now flips on and off automatically with no
known problems, within 1/15th second. The ESC key, that used to be
necessary, is pretty much useless now; the automatic on/off is
very effective.
DCFormat / Transverter / Speed
On the release disk you will also find DCFORMAT, Transverter,
and a speed tester, documented seperately. They're good utilities.
Disk Is 10-sector-format
Do not try to copy this disk with the "desktop drag" on older ST's
(e.g., dragging one disk icon to the other). Copy individual files
instead. We went to a 10-sector format to cram more on the disk (as you
can see, it's quite full even with 10-sector!) It's possible NeoDesk will
copy this format properly, and ANY decent disk backup program will do it;
it's not copy protected or anything, we just had to go 10-sectors for the
space.
FileMaker 4's Working
Need we say more? This was the last big MacApplication that failed . . .
Excel 2.2's Working
. . . except for Excel 2.2, which we also fixed.
Cybil And Doug
See our last newsletter. No new developments since then . . .
Farrah and Doug
. . . Except for this one.