> All I want to know is who decided to use \ instead of / for
> directory separators (warming up the tar and holding feathers).
I use Windows and Solaris. The driectory separator drives me batty.
One thing that DOS got right is to retain the file date when
a file is copied to another directory.
-------------------------------
Michael Holley
www.swtpc.com
-------------------------------
Well with Bruce's help (and a bit of actually paying attention to
what I was doing) I got Kermit to compile -- properly -- on RDOS.
Kermit for RDOS consists of about 20+ smaller source files,
separately compiled, and linked with the linking loader. THe
source is D.G. FORTRAN5, and the output of a seemingly-lost RATFOR
preprocessor.
Kermit was archived as one big file containing all of the sources,
with comments to indicate where to break them into files:
CCCCCCCCCCCCC STDIO.FT CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
C code here....
...
Bruce game me an awk script that does it, but since I don't have
any way to get files in/out to RDOS (except open-loop "text
download" through the console) I did it locally with an nspeed
command, looks like teco:
1WMGRKERMIT.FR$<NCCCCCCCCCCCCC ^T$T1BC1BS10KS $-1D1KBS0GCGW^B1$(^)!I;>GC$$
$ is escape key, ^T is Control-T, etc.
< ... > is a loop; the semicolon is a conditional exit (if the search fails).
N is "yank pages, copying to the output file, buffer-fulls of text
until 'CCCCCCCCCCCCC ^T' is found". ^t means any number of space
characters.
The transmission line noise is really copying the current line,
under the virtual cursor, to buffer 1, stripping the comments off
the embedded filename and building a command to create anoutput
file...
Hi Fred;
Glad you signed on. I have been waiting for someone older than myself to
speak up. Just didn't want to be the oldest.
Born 1930 and I will be 75 in April. Started playing around with an Atwater
Kent radio around 1939, went to a Radio Mech. course at a Vocational School
, 1944 to 45. Left school and joined the Army where I served 5+ years as a
brass pounder and comm. chief. Two years of which were at the Signal Corps
Evans Lab.
After the service went to work for the Navy, mostly on Shipboard radar and
taught in their apprentice program ( Math and Physics )
Moved to the FAA when they opened their Research Center at Pomona, N.J. in
1960. Worked mostly on Radars and Auto Landing Systems. All Analog stuff.
Was assigned to the ATC Labs in early 1970 and it didn't take long to
realize I was in a "Mushroom Farm" so I went back to school for a degree in
Information Science. That is when I got into computers. I remember the 1401,
7090, sigma 7 and the IBM 360's ( we called them 9020's ).
the ATC program was being developed in JOVIAL. My language however was
FORTRAM.
Retired in 1981.
Fred if you look hard enough you will find a few of us are still here.
Bob Yannetti, age 74 going on 75
Ironically, the DVD-R media that I've had the best luck with -- and my
experience with this brand has been superb, no failures, perfect burns, and
perfect subsequent readability -- is a bargain private-label off-brand.
It's "K-Hypermedia" brand, sold by OfficeMax (and, very rarely, by other
retailers). I've gone through over 10 25-packs over a period of 2 years, 1x
(which burn well at 2x on Pioneer burners) and 4X, and my experience with
these has been wonderful. And they are cheap, sometimes on sale for as
little as $5 for a 25-pack (more commonly $10 for a 25-pack).
On my Pioneer A04 burner, I could burn the 1X media at 2x by default. On my
A08, I had to get "hacked" firmware to allow me to select any burn speed
with any media. But in both cases, 2x works fine for the nominally 1X
media. I'm now also using 4X media with equally good results.
[They have a DVD+R product as well, I have used one spindle of it, it was
fine, but my experience suggests that when given a choice, "-" is the format
to prefer. Cheaper, more available and more likely to be readable, not
withstanding that the "propaganda" says just the opposite.]
Ashley Carder wrote:
> Anyone out there have a working VT-100 terminal
> and keyboard they'd be interested in selling? I
> need to get one to replace one that I'm giving
> away with a system.
Funny -- I was about to post the same "Wanted" message last week.
Let me jump in line here behind Ashley for anyone out there with a working
VT-100 terminal they feel like parting with (for $'s, or for trade).
Contact me off-list.
Eric F.
I have quite a bit of experience with these (and, by the way, I'm looking
for one or two ....).
The Persci drives for the most part are single full-size drives that accept
two different media in the same drive. They are not quite the same thing as
"dual drives" because there is only one head positioner for both drives.
However, it's a huge voice coil and it's very fast (1 millisecond per
track).
The drives have a horrendous number of very, very different configuration
achieved by about 4 dozen jumpers. There are several dozen "standard"
configurations, and many of them are quite non-interchangeable.
The interface is a standard 8" floppy interface (50-pin connector), and they
can be used with most controllers ***IF*** they are properly
jumpered/configured.
However, some of their quirks are very, very unique. One is the single
positioner, moving the head of the "A:" drive also moves the head of the
"B:" drive, and vice-versa. This requires a different BIOS for Persci
drives vs. a more conventional configuration of two separate drives.
Another thing is seek timing. The seek speed is very, very fast, and the
servo system is capable of dynamic acceleration and deceleration. To use
this, which is all but mandatory, instead of sending step pulses at, say, 6
nSec. per pulse (track), the step pulses are sent at a VERY high rate, over
100,000 pulses per second. The drive buffers them, and dynamically
accelerates the head to max speed for the seek, and dynamically decelerates
it at the end of the seek. There is a new interface signal "seek complete"
to tell you when the head has arrived at the desired track. This is totally
different than a normal stepper motor drive, and requires both some minor
hardware changes (which are supported by almost all LSI-based S-100
controllers, e.g. Western Digital) and some software changes. It's not
difficult and it works well, BUT it means that a BIOS for a conventional
stepper motor drive and a Persci drive are two separate BIOS'.
These drives are a wonder to behold, the main board is huge (something like
200 square inches) and beautiful. But they are almost unserviceable, since
parts are not available and they used some very unique components, and also
they are quite fragile (there is a special insert required for shipment, and
they are unlikely to survive shipment without that insert).
I personally scanned all of the Persci service data on the 270/277 drives
and they are up on Howard's site for download, so at least documentation is
available. If you have any questions, I'll try to help, but if they are
"dead" or grossly non-functional, the chances of resurrecting them are slim.
>From: "Eric Smith" <eric at brouhaha.com>
>
>river wrote:
>> Yes, the 8080/8085/8086 series chips did parity checking. The PSW had a
>> flag for odd/even parity. Naturally, the Z80, being a superset of the 8080,
>> also had this ability.
>
>In that regard, the Z80 is not a proper superset of the 8080. The
>parity flag is usurped to act as an overflow flag after arithmetic
>operations. It is still partity after moves and logical operations.
>
>Eric
>
>
Hi
Giving it some thought, the only compute application
for parity that I can think of would be reverse bit
swapping. This is generally done on an address though
and might not be all that useful on a byte value.
( FFT's often need this function ).
Dwight
Been kinda busy (especially at home - which is where all the fun stuff is
supposed to happen...)
However, on Sunday, I did get enough of a chance to dead-bug connect 2
24-pin wirewrap sockets together, and wirewrap the legs - this made a
hackish but working adapter for the HHC Roms. I read several different Roms
into the PC, verified 'em, looked at .bin files and I can say that
everything that looked like it should be text looked like text.
The Basic ROM just came in today - and despite the lack of anti-stat looks
to be in fair shape. It also looks like the seller put 2 dabs of epoxy on
the chip so it couldn't be removed from the carrier and copied. (I say that
because I could still smell the epoxy - Everything that I've ever epoxied
stopped smelling after 2-3 weeks; a month at most.) Of course, my adapter
doesn't give a whit about that... ;-)
I'm looking to snipe an EPROM eraser on ePay tonite about 10pm EST... If I
don't get that one tonite, I'm just gonna buy one so I'll be able to burn
new EPROMs very soon.
If anyone just wants the code and can make their own EPROMs, I'll have that
available soon; prolly tonite. ;-)
I'll keep in touch,
"Merch"
--
Roger "Merch" Merchberger | A new truth in advertising slogan
SysAdmin, Iceberg Computers | for MicroSoft: "We're not the oxy...
zmerch at 30below.com | ...in oxymoron!"
Hi,
>Rumor has it that William Donzelli may have mentioned these words:
>>Parity checking is the job of the memory controller, not the processor. In
>>fact, I am having a hard time thinking of a processor that did its own
>>parity checking in software (yes, I know any processor could do it, but
>>did any really do it?).
>8085? There's a P bit in the condition code register (in 6809 speak) -- I
>think it's called the PSW -- Program Status Word? I'm just beginning
>learning assembly on my Tandy 10x/200 machines... Anyway, there are several
>arithmetic operations that automatically set the Parity bit and there are
>branch and return instructions that utilize the status of the P bit.
Yes, the 8080/8085/8086 series chips did parity checking. The PSW had a flag
for odd/even parity. Naturally, the Z80, being a superset of the 8080, also had this
ability. I assume that the later generations of x86 series chips, for backward compatability,
also had the ability to check for parity.
The TMS9900 series also did parity checking.
>From a cursory glance at the status registers of the 6809 and 68000 processors, it
appears these devices did not have inherent parity checking.
river
Long story short, I got it working. Want some pictures?
http://www.gjcp.net/~gordonjcp/pdp-11.html
Want to play? Get me in #classiccmp on irc.freenode.net sometime
tonight. I plan on going to bed about midnight, and I will be turning
the noisy bloody contraption off. Failing that, get me tomorrow.
Gordon.
