4 Mar
2005
4 Mar
'05
9:19 p.m.
These articles say that Kildall's CP/M easteregg appears in DOS 1.0
or so.
Has anyone here actually run the proof for themselves?
What CP/M easter egg? That ensuing discussion lacks any useful detail. 
9:54 p.m.
> What we're trying to figure out is the
keystrokes and the MSDOS version.
On Fri, 4 Mar 2005, Eric Smith wrote:
The story is apocryphal. No one has ever been able to
cite the
actual details because it never happened. If it had happened, it
would be possible to get a court transcript.
It is doubtful that it exists. The story is such an exact match
to the easter egg in TRS-DOS 2.x (that really DOES exist), that
the story is almost certainly a corruption of the TRS-DOS story,
unless Randy Cook got the idea from Gary Kildall?
There is also a [presumably apocryphal] story of the TRS-DOS
easter egg being used in court.
--
Grumpy Ol' Fred cisin at xenosoft.com
10:58 p.m.
From: "Eric Smith" <eric at brouhaha.com>
Sent: Friday, March 04, 2005 8:32 PM
John wrote:
The source to CP/M is freely available and well understood, it's not there.
It does make a good story though, even though it's not true ;-)
Randy
What we're trying to figure out is the
keystrokes and the MSDOS version.
The story is apocryphal. No one has ever been able to cite the
actual details because it never happened. If it had happened, it
would be possible to get a court transcript.
Eric
9:36 p.m.
On Fri, 4 Mar 2005, John Allain wrote:
For people who haven't read it, here's what it
said:
"IIRC, Killdall stood up in court and entered a keystrokes at a PC
running MSDOS and brought up an easter egg he had programmed
into CP/M years earlier, proving they had used his code."
What we're trying to figure out is the keystrokes and the MSDOS version.
Here's another account: http://dfarq.homeip.net/article.php?story=935
"Back in the mid-1990s, PC Magazine columnist John C. Dvorak wrote
something curious about this operating system. He said he knew of an
easter egg present in CP/M in the late 1970s that caused Kildall's name
and a copyright notice to be printed. Very early versions (presumably
before the 1.0 release) of DOS had this same easter egg."
I suspect that it might be very well be a corruption of the story
about Randy Cook's easter egg in TRS-DOS, that showed up in NewDOS
prior to NEWDOS80. That one really DID exist. RS, in TRS-DOS 2.3,
changed "Randy Cook" in the easter egg to "Tandy Corp". It was
invoked by running a boot file as if it were a program, with one
of the master paswords, and pressing 'J'? and 'N'? while it was loading.
--
Grumpy Ol' Fred cisin at xenosoft.com
9:49 p.m.
Fred wrote:
I suspect that it might be very well be a corruption
of the story
about Randy Cook's easter egg in TRS-DOS, that showed up in NewDOS
prior to NEWDOS80. That one really DID exist. RS, in TRS-DOS 2.3,
changed "Randy Cook" in the easter egg to "Tandy Corp". It was
invoked by running a boot file as if it were a program, with one
of the master paswords, and pressing 'J'? and 'N'? while it was loading.
That's much more believable. I worked at Apparat, but by the time I
started there they were already shipping NEWDOS/80.
The original NEWDOS (up through NEWDOS 2.1?) was a patched version
of TRSDOS. I don't know whether they originally were secretive about
that, but for at least the latter part of the time they sold it, they
told customers that they must have a copy of TRSDOS in order to buy
NEWDOS. Since Radio Shack would sell you a TRSDOS floppy (only, no
manual) for under $15, it wasn't a big issue.
NEWDOS/80 V1 was a complete rewrite from the ground up, and added many
new capabilities. It was available in separate editions for the
Model I and Model III. V2 added even more nice features.
The author of NEWDOS/80 (and the earlier patches to TRSDOS) was a
long-time IBM mainframe programmer. Based on research I did a few
years ago, I believe he was also one of the programmers of the
software for the Apollo Guidance Computer. He was adamant that we
not give out his name as he wanted to protect his privacy.
Eric
5 Mar
5 Mar
7:54 p.m.
I suspect that it might be very well be a corruption
of the story
about Randy Cook's easter egg in TRS-DOS, that showed up in NewDOS
prior to NEWDOS80. That one really DID exist. RS, in TRS-DOS 2.3,
I can confirm that one exists, I've seen it myself.
changed "Randy Cook" in the easter egg to
"Tandy Corp". It was
invoked by running a boot file as if it were a program, with one
of the master paswords, and pressing 'J'? and 'N'? while it was loading.
I thought it was type BOOT/SYS.WHO, hold down 2 4 and 6 and press enter.
For the non-TRS-80 people here :
In TRS-DOS, disk space was allocated by 'granules', these being half a
track on single-density disks (which is all 2.x supported). All space
allocated on the disk has a directory entry, this includes the directory
itself (DIR/SYS) and the boot block (BOOT/SYS)
The boot block was the first sector of the disk, but of course the rest
of that granule was also allocated to BOOT/SYS. The ROM bootstrap just
copied the first sector into memory (note : sector, not granule) and
jumped to the start of that area of memory.
But executable files in general were more complex. In particular you
could haev several logical sections that loaded at non-contiguous
addresses, you could have sections that were taken as comments (and not
loadrd anywhere), and so on.
The other thing you need to know is how the keyboard was scanned. It was
done in software. The low 8 address lines, suitably buffered, were fed to
the ekyboard column lines. The 8 row lines, again suitable buffered, were
fed to the data lines, these buffers being enabled in a 256-byte section
of the memory map between the ROMs and the RAM. By rerading 8 selected
address in that region you could determine which keys were pressed.
Now for the easter egg. The boot sector was carefully designed so that if
BOOT/SYS was loaded as a program from the TRS-DOS prompt, it would be
taken as a comment. And thys executing BOOT/SYS would ignore the real
boot sector, and would execute the stuff in the remaining 4 sectors of
that granule.
That contained a little program to kick the display into 32*16 mode, then
read in 512 bytes from those sectors, decrypt them, which in part
involved XORing them with the keyboard row data (I think all the column
lines were turned on) -- this means the message is not obvious if you
search the disk, and write the result to video memory.
On TRS-DOS 2.3, the result is a Radio Shack copyright screen. I am told
that on some earlier versions it was a Randy Cook copyright screen, and
this was used to prove that Radio Shack were still using Randy Cook's
code, even though they claimed they weren't (and weren't paying him
royalties).
That last part I am not sure about. The existance of the easter egg I am.
-tony
6 Mar
6 Mar
6:55 p.m.
> about Randy Cook's easter egg in TRS-DOS,
On Sun, 6 Mar 2005, Tony Duell wrote:
I can confirm that one exists, I've seen it
myself.
I thought it was type BOOT/SYS.WHO, hold down 2 4 and 6 and press enter.
WHO, in that case, is a password. Since TRS-DOS hashed the passwords, any
password that produces one of the correct hashes will do. IIRC, 9642
is the hash for system master password, and anything that will
hash to that, such as NV36 will work on any file.
IIRC, I was able to use a combination of two letter keys. There are
probably many that would work.
On TRS-DOS 2.3, the result is a Radio Shack copyright
screen. I am told
ALMOST. It is actually a "TANDY CORP" copyright screen. RS wanted to
make as few changes from "RANDY COOK" as they could.
code, even though they claimed they weren't (and
weren't paying him
royalties).
When RS cut a deal to distribute LDOS, Randy finally started getting his
royalties.
--
Grumpy Ol' Fred cisin at xenosoft.com
4 Mar
4 Mar
11:37 p.m.
On Fri, 4 Mar 2005, John Allain wrote:
Ahh... that does reveal a bit.
I, and I'm sure many, translated programs from 8080/z80 to 8086
with a text editor 'search and replace'. Clearly this was Intel's
intent and it worked fine; the first thing I translated was my
EPROM debugger/boot code. (It obviously required hand coding too
to make it work but the bulk of hard-won code was ported.)
So if you disassembled 8080 binary, mechanically translated the
resulting source to 8086, that would preserve the easter egg.
(Man would that be one ugly program.)
It seems entirely likely that Kildall's hack (algorithmically
obscured text within the code somewhere) would only be preserved
by some sort of mechanical translation. It's doubtful an honest
re-write even with listings side by side would result in it being
preserved.
There are few enough CP/M binaries that the sources of which could
contain such things. CCP did not become COMMAND.COM, that is
clear, it's totally different.
MSDOS/CPM kept the same "BDOS::BIOS" relationship, jumptables and
all, but MSDOS (and 86DOS) have a totally different filesystem, so
the directory and block-management stuff is unlikely to contain
the egg.
I bet some puzzling would work out a list of the few places where
text would be brought to console out (the implication of the
accounts above).
If this were the case, I bet I (and many others) could identify
mechanically translated code by looking at it but (1) I'm rusty
and (2) I don't really give a shit, I'll wait to enjoy someone
else's work.
Gotta be 86DOS or MSDOS 1.25. The first 86DOS I wrote a BIOS for
was version (oh so clever) 0.86. It was basically 8086 CPM 1.4.
Then there was a 1.10 I think.
What CP/M
easter egg? That ensuing discussion lacks any useful detail.
For people who haven't read it, here's what it said:
"IIRC, Killdall stood up in court and entered a keystrokes at a PC
running MSDOS and brought up an easter egg he had programmed
into CP/M years earlier, proving they had used his code."
What we're trying to figure out is the keystrokes and the MSDOS version.
Here's another account: http://dfarq.homeip.net/article.php?story=935
"Back in the mid-1990s, PC Magazine columnist John C. Dvorak wrote
something curious about this operating system. He said he knew of an
easter egg present in CP/M in the late 1970s that caused Kildall's name
and a copyright notice to be printed. Very early versions (presumably
before the 1.0 release) of DOS had this same easter egg." 
7 Mar
7 Mar
11:32 a.m.
Tom Jennings wrote:
So if you disassembled 8080 binary, mechanically translated the
resulting source to 8086, that would preserve the easter egg.
(Man would that be one ugly program.)
It seems entirely likely that Kildall's hack (algorithmically
obscured text within the code somewhere) would only be preserved
by some sort of mechanical translation. It's doubtful an honest
re-write even with listings side by side would result in it being
preserved.
There are few enough CP/M binaries that the sources of which could
contain such things. CCP did not become COMMAND.COM, that is
clear, it's totally different.
I've always heard that it was in PIP, using some lame type of
encoding (an xor?).
5 Mar
5 Mar
6:15 a.m.
John Allain wrote:
For people who haven't read it, here's what it
said:
"IIRC, Killdall stood up in court and entered a keystrokes at a PC
running MSDOS and brought up an easter egg he had programmed
into CP/M years earlier, proving they had used his code."
"Back in the mid-1990s, PC Magazine columnist John C. Dvorak wrote
something curious about this operating system. He said he knew of an
easter egg present in CP/M in the late 1970s that caused Kildall's name
and a copyright notice to be printed. Very early versions (presumably
before the 1.0 release) of DOS had this same easter egg."
I call bullshit/urban legend. John C. Dvorak has been fooled many, many times
before (my favorite is a 286 he reviewed that he claimed blew the doors off of
everything else; it was found later that the benchmarks he ran on it had been
fudged by the manufacturer by deliberatly altering the internal clock to run
slow) so I automatically invalidate his story.
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Want to help an ambitious games project? http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
7073
days inactive
7075
days old
11 comments
8 participants
participants (8)
-
allain＠panix.com -
ard＠p850ug1.demon.co.uk -
cctech＠randy482.com -
cisin＠xenosoft.com -
eric＠brouhaha.com -
kth＠srv.net -
tomj＠wps.com -
trixter＠oldskool.org