24 Oct
2022
24 Oct
'22
5:41 p.m.
Wanted to pass along that PALM has been added as a target to the Macro
Assembler AS. This means writing some assembly stuff for the old IBM
5100/5110/5120 systems.
Assembler available here:
http://john.ccac.rwth-aachen.de:8000/as/
So no one has to re-type it, the "bounce" sample is located here:
https://github.com/voidstar78/5110VEMU/blob/main/bin/sample1_palm.asm
Norbert has the online IBM5110 emulator, but it doesn't have a way to
"script inputs". Not sure if he's in the mood to crack that open again,
but it would be a nice feature to add (a kind of "File Open" and just
script the content of the file like typing on its emulated keyboard)..
Alternative way to test some compiled assembly is to script it into the
5110VEMU located here:
https://github.com/voidstar78/5110VEMU/tree/main/bin
(Wintel builds, uses pdcurses which might not be too hard to port to other
systems)
Or if have a physical system willing to power up, you can use a serial port
as a proxy to the keyboard:
https://github.com/voidstar78/KBD5110/tree/main/CODE
Screen is port mapped at 0x0200, so drawing stuff is easy. I'll work on an
example to poll keyboard inputs - then we're in business for some
interactive software content :D (well, will need some RNG solution next)
Cheers!
25 Oct
25 Oct
1:01 a.m.
On Mon, 24 Oct 2022, Steve Lewis wrote:
Wanted to pass along that PALM has been added as a
target to the Macro
Assembler AS. This means writing some assembly stuff for the old IBM
5100/5110/5120 systems.
Assembler available here:
http://john.ccac.rwth-aachen.de:8000/as/
Funny, as this is the very assembler that I had added the PALM target
years ago for my simulator ;-)
It would have been nice if you had talked to me before as to have one
single implementation. But as I see, you took the original IBM mnemonics
whereas I have created more readable ones that make more sense :-)
Christian
2 a.m.
Christian! I tried contacting you a few months ago. Apologies, I suspect
some intermediate ISP may be blocking gmail? I've seen this happen with
other contacts.
Alfred Arnold made the changes to Macro AS himself, and he used both yours
and the IBM mnemonics since there were no conflicts.
And now, I think it's time for a "Bad Apple" style demoscene for the 5100
(as one was made for the 5150), before its 50th in 2025. Let's see what
some raw PALM can do.
Quick question, Christian were you aware of the microcode Appendix C in the
original IBM 5100 MIM? I'd assume Yes - but it's something I've always
been curious about, if that appendix had ever helped you at all, or if it
was too lacking in details? That Appendix C got removed for the 5110
(which has the same Processor/Controller). I'd assume you alway had
access to the 5100 manuals, not just the 5110's.
-Steve
On Tue, Oct 25, 2022 at 3:01 AM Christian Corti via cctalk <
cctalk(a)classiccmp.org> wrote:
On Mon, 24 Oct 2022, Steve Lewis wrote:
Wanted to pass along that PALM has been added as
a target to the Macro
Assembler AS. This means writing some assembly stuff for the old IBM
5100/5110/5120 systems.
Assembler available here:
http://john.ccac.rwth-aachen.de:8000/as/
Funny, as this is the very assembler that I had added the PALM target
years ago for my simulator ;-)
It would have been nice if you had talked to me before as to have one
single implementation. But as I see, you took the original IBM mnemonics
whereas I have created more readable ones that make more sense :-)
Christian
26 Oct
26 Oct
1:23 a.m.
On Tue, 25 Oct 2022, Steve Lewis wrote:
Christian! I tried contacting you a few months ago.
Apologies, I suspect
some intermediate ISP may be blocking gmail? I've seen this happen with
other contacts.
Oh my, yes I have some mail chaos on my system and I guess yours must have
been "lost" somewhere. I need to dig through my pile....
Alfred Arnold made the changes to Macro AS himself,
and he used both yours
and the IBM mnemonics since there were no conflicts.
Perfect! I didn't know that.
And now, I think it's time for a "Bad
Apple" style demoscene for the 5100
(as one was made for the 5150), before its 50th in 2025. Let's see what
some raw PALM can do.
Hehe, this sounds like a great idea!
Did I send you my Kermit implementation or my Infocom V2 interpreter?
Quick question, Christian were you aware of the
microcode Appendix C in the
original IBM 5100 MIM? I'd assume Yes - but it's something I've always
been curious about, if that appendix had ever helped you at all, or if it
was too lacking in details? That Appendix C got removed for the 5110
(which has the same Processor/Controller). I'd assume you alway had
access to the 5100 manuals, not just the 5110's.
Yes I'm aware of that. But I noticed that there are significant
differences between the 5100 and the 5110 concerning the different ROS and
the I/O functions like switching between ROS/RWS or controlling the
display adapter.
At the time when I reverse-engineered the 5110 I didn't have access to the
5100 docs, though, only to some copies of IBM internal papers (on my
5110 site).
Christian
10:55 a.m.
Hi!
On Mon, 2022-10-24 19:41:56 -0500, Steve Lewis via cctalk <cctalk(a)classiccmp.org>
wrote:
Wanted to pass along that PALM has been added as a
target to the Macro
Assembler AS. This means writing some assembly stuff for the old IBM
5100/5110/5120 systems.
Assembler available here:
http://john.ccac.rwth-aachen.de:8000/as/
I'm doing automated builds for some software (as a QA check) and
thought about pulling this Assembler in as well. Are there only the
release tarballs? Or is there some GIT repo somewhere as well? If
there isn't, I'd probably feed the tarballs into a repo by myself, but
I'd prefer to just pull from a publically hostet repo of course.
MfG, JBG
--
11:13 p.m.
Hehe, this sounds like a great idea!
Did I send you my Kermit implementation or my Infocom V2 interpreter?
Nope - would love to check them out, any PALM examples. Your annotated
disassemble of the Executive ROS is helping a lot, of examples of how to do
certain things in native PALM examples. Some idea for implementing a RNG
in PALM is probably going to be needed (for any gaming).
Does your Kermit use the rear external IO? I don't have any of the asynch
IO or comm. cards. But I've imagined it would be possible to attach a
WiModem232 to the rear external IO and (using 3 wires) to implement a
terminal in PALM. But being completely "CPU" based, unsure what baud rate
it could achieve. This is similar to the original Tandy Color Computer -
if you use its built in "Serial IO" connector, you can only get 1200 baud.
But if you use the "RS232 Pak" with the proper buffer, it can do 9600
baud. So, you can get a CoCo1 on the internet (or even a PET) - and I
think it actually would be possible to get a 5100 on the internet in a
similar way!
I recently got in contact with Hal Prewitt who made CoreNet for the IBM
5110. He confirmed that the hardware no longer exist (as far as he is
aware), but he does still have a copy of PC51 (essentially a 5110 BASIC
emulator that runs on the IBM PC -- so those fancy BASIC programs that use
the FORMS feature of the IBM 5110 can continue to run on the IBM PC;
unclear if he still has the source code for it).
differences between the 5100 and the 5110 concerning
the different ROS and
the I/O functions like switching between ROS/RWS or controlling the
display adapter.
Indeed, I ran into the same issue when running the extracted 5100 Executive
ROS in your PALM emulator. It can run the BUP, and I think I managed to
send the scancode early to activate its DCP (if you do so too early, the
BUP just shows you the scan code - but if you wait to a certain time just
before BUP finishes, some CMD-ATTN {I forget the exact scan code} like
combination activates the DCP). So the 5100 ROS binary is extracted
correctly, but I had issues as well in the ROS/RWS transition. Solving
that before 2025 would be a nice 50th gift as well, since otherwise there
is still no 5100 emulator -- where it helps to dev/test things in an
emulator, before powering up a physical 5100.
Aside from github, I'm not sure of a convenient repository to archive/share
any PALM assembly examples.
-Steve
On Wed, Oct 26, 2022 at 3:23 AM Christian Corti via cctalk <
cctalk(a)classiccmp.org> wrote:
> On Tue, 25 Oct 2022, Steve Lewis wrote:
> > Christian! I tried contacting you a few months ago. Apologies, I
> suspect
> > some intermediate ISP may be blocking gmail? I've seen this happen with
> > other contacts.
>
> Oh my, yes I have some mail chaos on my system and I guess yours must have
> been "lost" somewhere. I need to dig through my pile....
>
> > Alfred Arnold made the changes to Macro AS himself, and he used both
> yours
> > and the IBM mnemonics since there were no conflicts.
>
> Perfect! I didn't know that.
>
> > And now, I think it's time for a "Bad Apple" style demoscene for
the 5100
> > (as one was made for the 5150), before its 50th in 2025. Let's see what
> > some raw PALM can do.
>
Hehe, this sounds like a great idea!
Did I send you my Kermit implementation or my Infocom V2 interpreter?
>
> > Quick question, Christian were you aware of the microcode Appendix C in
> the
> > original IBM 5100 MIM? I'd assume Yes - but it's something I've
always
> > been curious about, if that appendix had ever helped you at all, or if it
> > was too lacking in details? That Appendix C got removed for the 5110
> > (which has the same Processor/Controller). I'd assume you alway had
> > access to the 5100 manuals, not just the 5110's.
>
> Yes I'm aware of that. But I noticed that there are significant
differences between the 5100 and the 5110 concerning
the different ROS and
the I/O functions like switching between ROS/RWS or controlling the
display adapter.
> At the time when I reverse-engineered the 5110 I didn't have access to
the
> 5100 docs, though, only to some copies of IBM internal papers (on my
> 5110 site).
>
> Christian
>
27 Oct
27 Oct
4:25 a.m.
On Thu, 27 Oct 2022, Steve Lewis wrote:
Ok, can you remind me next week perhaps?
Hehe, this
sounds like a great idea!
Did I send you my Kermit implementation or my Infocom V2 interpreter?
Nope - would love to check them out, any PALM examples. Your annotated disassemble of the Executive ROS is helping a lot, of
examples of how to do
certain things in native PALM examples. Some idea for implementing a RNG
in PALM is probably going to be needed (for any gaming).
I have implemented practically everything that is needed for the
interpreter, including the kind of virtual memory/swapping mechanism. And
of course, the ASCII-EBCDIC translation and screen output for the 64
chars/line.
Does your Kermit use the rear external IO? I
don't have any of the asynch
IO or comm. cards. But I've imagined it would be possible to attach a
WiModem232 to the rear external IO and (using 3 wires) to implement a
terminal in PALM. But being completely "CPU" based, unsure what baud rate
it could achieve. This is similar to the original Tandy Color Computer -
if you use its built in "Serial IO" connector, you can only get 1200 baud.
But if you use the "RS232 Pak" with the proper buffer, it can do 9600
baud. So, you can get a CoCo1 on the internet (or even a PET) - and I
think it actually would be possible to get a 5100 on the internet in a
similar way!
No, I have the Async/Serial I/O feature card. Originally, it only supports
up to 1200(?) baud and only half duplex, but since I have drawn the
schematics (it's only standard TTL) I know how to program it.
It has two modes, a bit transfer mode, and a byte transfer mode. The first
one is needed to find the start bit, then you can switch to byte mode and
have an interrupt after one transferred character.
The trick that I use for full duplex terminal emulation is to only use the
bit mode. I get an interrupt twice as many as the baud rate and examine
the input line to detect the start bit etc. In the same time, I output a
pending bit to be sent. With this, I ended with a speed of up to 4800
baud, already four times of the IBM supported byte mode rate.
For file transfers, I use the standard byte mode. The Kermit protocol is
made for half duplex transmission line, so there is no special protocol
handling needed. There, I can go up to 38400 baud.
Note, the I/O card is absolutely dumb. It is mainly a card with clock
dividers, shift registers and drivers/receivers. Is has no buffer,
nothing.
I have always wanted to build a bus interface card. It is a simple
open-collector I/O bus with an easy to implement bus protocol. I've dreamt
of building a bridge to an 8 bit ISA backplane in order to use simple
graphics cards and controllers like serial, parallel or even disk
interfaces. I have tons of these laying around here...
Christian
557
days inactive
559
days old
6 comments
3 participants
participants (3)
-
Christian Corti -
Jan-Benedict Glaw -
Steve Lewis