Customizing JiffyDOS/128D?
Peter T Karlsson
>the trouble is that I don't know the layout of the C128D EPROM chip (the one
>that contains the C64 KERNAL, BASIC and C128 KERNAL, BASIC (it's all in one
>chip)).
TP> As I recall, there are four sections to the EPROM for the 128D -
TP> 64 mode stock, 64 mode jiffy, 128 mode stock, and 128 mode jiffy.
The big problem is, that also the original ROM code is divided up, that
would be to three parts, I guess, if it's not done in RAM:
C64 mode Swedish
C128 mode Swedish
C128 mode English
I switch between Swedish and English mode (will change charset and keyboard
layout in C128 mode, and only charset in C64 mode) by using the ASCII/CC
key (relabeled "CAPS LOCK"). The CAPS LOCK key is hardwired to the charset
ROM, which contains the Swedish and the English charset, but I think that
the C128 mode transition between Swedish ROM (Swedish keyboard) and English
ROM (English keyboard) is done in software (i.e not hardwired).
That's the problem; I need to get the JiffyDOS ROM to 'feel' this change,
so that I can switch between Swedish and English mode with the ASCI/CC
key, removing the CAPS LOCK function, plus that I need to be able to switch
between stock and JiffyDOS mode...
Sounds easy, doesn't it?
I've done a dump of the original Swedish EPROM, and found it to be a 32KB
EPROM (at least the auto-detect got it as a such), but I haven't checked
the layout of it yet.
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
... "How to Upgrade to MS-DOS 6.0 from OS/2" -- MS-DOS 6.0 manual (1993)
Tony Postmayer
>I've done a dump of the original Swedish EPROM, and found it to be a 32KB
>EPROM (at least the auto-detect got it as a such), but I haven't checked
>the layout of it yet.
>
jiffydos replacement is 64k since it comprises both a copy of the
original as well as the jiffy code. With that in mind, I would expect
your 32k ROM to be "split" right down the middle - one half of it
being the 64 mode stuff, and the other half the 128 mode stuff. As
far as the keyboard layout, I can't help - my experience is only with
North American hardware.
What does your 128d look like? Does it come in a slimline metal case?
Or is it a somewhat bulky plastic case with a handle? Does it follow
the NTSC or the PAL video standard?
Tony -
Pontus Berg
>> the trouble is that I don't know the layout of the C128D EPROM chip (the
>> one that contains the C64 KERNAL, BASIC and C128 KERNAL, BASIC (it's all
>> in one chip)).
TP>> As I recall, there are four sections to the EPROM for the 128D - 64
TP>> mode stock, 64 mode jiffy, 128 mode stock, and 128 mode jiffy.
PTK> The big problem is, that also the original ROM code is divided up,
PTK> that would be to three parts, I guess, if it's not done in RAM:
PTK> C64 mode Swedish
PTK> C128 mode Swedish
PTK> C128 mode English
How big is the ROM in you machine? I have a 16K one for Basic+Kernal and then I
guess some sort of ROM for the 128 part (16K as well I assume). This means you
have to switch two different chips to switch the 64 as well as the 128 mode.
PTK> I switch between Swedish and English mode (will change charset and
PTK> keyboard layout in C128 mode, and only charset in C64 mode) by using
PTK> the ASCII/CC key (relabeled "CAPS LOCK"). The CAPS LOCK key is
PTK> hardwired to the charset ROM, which contains the Swedish and the
PTK> English charset, but I think that the C128 mode transition between
PTK> Swedish ROM (Swedish keyboard) and English ROM (English keyboard) is
PTK> done in software (i.e not hardwired).
Wait a minute here; Do you actually have different KEYMAPS for the two modes?
I'll be damned - they are different in 128 mode! But the switching takes a bit
to long to be only a flip of a switch if you ask me!
PTK> I've done a dump of the original Swedish EPROM, and found it to be a
PTK> 32KB EPROM (at least the auto-detect got it as a such), but I haven't
PTK> checked the layout of it yet.
Tron was with me a weekend and we coded a kernal patcher, that installed the
swedish keys into their proper places so if you have the default dump of a
kernal, you can easily make a swedish one out of it!
)==========================================================(
) o/ Bac...@FairLight.COM Fido: 2:201/411.71 (
) (# I'm a "geek tool" so ask about emulators elsewhere! (
)=/7=======================================================(
... The C64 is my only religion!
It's for real and can be proven and it's for sure divine!
(Pontus Berg)
Jouko Valta
> In a message of 19 Mar 96 Peter T Karlsson wrote to All:
>>> the trouble is that I don't know the layout of the C128D EPROM
>>> chip (the one that contains the C64 KERNAL, BASIC and C128
>>> KERNAL, BASIC (it's all in one chip)).
I don't know the order (I have the roms as 4 chips) but the sizes are:
C64 basic+kernal (same as on C64C) 16K (8+8)
C128 editor, BIOS and kernal 16K (4+4+8)
Basic Low 16K (16)
Basic High and Monitor 16K (10+4)
Pontus> Wait a minute here; Do you actually have different KEYMAPS
Pontus> for the two modes? I'll be damned - they are different in
Pontus> 128 mode! But the switching takes a bit to long to be only
Pontus> a flip of a switch if you ask me!
It copies the 4 pointers to keyboard decoding tables and then all the
character set onto the 80 column display memory.
--j;
Peter T Karlsson
TP> 32k is normal. The american 128d comes with a 32k ROM,
Okay, then I guess that the keyboard switching and stuff must be in
software, and that the pointers ot the keyboard maps are changed.
That should be 'fairly easy' to implement in the JiffyDOS chipset.
TP> What does your 128d look like? Does it come in a slimline metal case
TP> ? Or is it a somewhat bulky plastic case with a handle? Does it
TP> follow the NTSC or the PAL video standard?
I have a PAL C128D (I think all Swedish C128D's are PAL ones, since PAL is
used in Sweden and Finland), in a metal case. It only has one kernel chip.
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
... Liberalism is futile. We are EIB. You will be assimilated
Peter T Karlsson
PB> How big is the ROM in you machine?
I read it here at school, and the EPROM reader found it to be 32K. I have a
dump of it.
There is only one chip containing everything, and I only switched one chip
when changing to JiffyDOS.
PB> Wait a minute here; Do you actually have different KEYMAPS for the
PB> two modes? I'll be damned - they are different in 128 mode!
Yes, they are. It's only unfortunate that this is not available in C64
mode, where I would need it (the thinnest underscore character is missing
on the Swedish keyboard -- this was one of the things I was thinking about
correcting).
PB> But the switching takes a bit to long to be only a flip of a switch if
PB> you ask me!
It does? Hmm. Maybe it's done in software then. I'm not sure. The
character generator is hardwired at least, there's an extra wire from the
keyboard connector to it on the motherboard :-)
(In 80 column mode, the charset is re-read by software).
PB> Tron was with me a weekend and we coded a kernal patcher, that
PB> installed the swedish keys into their proper places so if you have
PB> the default dump of a kernal, you can easily make a swedish one out
PB> of it!
I don't have a dump of the JiffyDOS chip, only of my Swedish KERNAL.
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
... I am Reagan of Borg. Prepare to be...uh, I don't recall
paul....@bbs.wline.se
About changhing the 128 KERNAL,
the problem I ran into (I changed my jiffy to swedish keyboard map in 64 mode)
was that whenever you change a few bytes here and there in the 128 ROM the 128
will no longer boot at all. I assume this is because of the rutine that checks
a memory location for a CHECKSUM, which I never figured out how to RE-caculate.
the checksum is missing from the 64 ROM so one can change it to you hearts
content,
I figured out how to get around the thing in 128 mode (not sure about the
keyboard mapping, but i think it switched too) start the machine with the caps
lock (ASCII/CC) key down then switch in jiffy and enable it) just to read in
the character set you can do somthing like wait 1:sys 65378 and push the ascii
capslock key while in the wait.
(hope you get som ideas from this I've got a terrible memory and could be
totaly wrong about the details but the general idea is correct)
__________________________________________________
/ Paul Fisher aka Natas | paul....@bbs.wline.se \
\_________________/Malmo Sweden\___________________/
/ C128D 7Mb RamLink 1.60Mb floppy 105Mb SCSI HD + \
/ 14.4 USR modem and a bunch of other stuff!!!!!!!! \
\ keeping up with the big guys /
\__________________________________________________
David Gahris
> Pontus Berg <pontu...@p71.anet.canit.se> writes:
> > In a message of 19 Mar 96 Peter T Karlsson wrote to All:
> > > > chip (the one that contains the C64 KERNAL, BASIC and C128
Here is the layout of all 64K of OS ROM in the 128(D) and 128DCR:
Use Start End Size 128DCR K 128 upgr K
--- ----- --- ---- ------ - -------- -
BASlo $4000 $7fff 16K 318022-02 0-16 318018-0x 0-16
BAShi $8000 $bfff 16K 318022-02 16-32 318019-0x 0-16
BAS2 $a000 $bfff 8K 318023-02 0-8 251913-01 0-8
K64 $e000 $ffff 8K 318023-02 8-16 251913-01 8-16
MLmon $c000 $cfff 4K 318023-02 16-20 318020-0y 0-4
ZBIOS $0000 $0fff 4K 318023-02 20-24 318020-0y 4-8
K128 $e000 $ffff 8K 318023-02 24-32 318020-0y 8-16
You asked about the one that appears in the bottom five lines of this table.
"Start" and "End" are addresses within the CPU address space. "BAS2" and "K64" are visible only in 64 mode and "ZBIOS" is for CP/M mode.
The 123456-12 numbers are the Commodore part numbers of the ROM chips for the USA. In North America, "x" is either 2, 3, or 4, and y = x + 1.
The "K" columns measure, in kilobytes, the chip internal locations of each part of the OS. These are the locations that an EPROM reader will give you.
There are a bunch of ROM dumps and diffs at: ftp://ftp.funet.fi/pub/cbm/firmware/c128
I think this is how character set switching works: The 128 Programmer's Reference Guide schematic shows a jumper leading to the high-order address bit of the character ROM. This jumper can connect either to the 128/64 mode signal or to one of the locking keys on the keyboard. On NTSC systems this jumper connects to the 128/64 signal and the locking key bears the label CAPS LOCK. So over here the 128-mode character set is different from the 64's character images.
In Europe the jumper connects to the locking key, which on some keyboards is ASC/CC and on others is ASCII/DIN. So in Europe the user can switch at will between two character sets in both 128 mode and 64 mode. This switching may work for 80 columns only at reset, when the Kernal copies the character bitmaps to VDC RAM. Otherwise, the editor would have to monitor the key continuously.
-David Gahris
dga...@eece.maine.edu
Please reply by E-mail.
David Gahris
Peter Karlsson
discussions:
> whenever you change a few bytes here and
> there in the 128 ROM the 128 will no longer boot at all. I assume
> this is because of the rutine that checks a memory location for a
> CHECKSUM, which I never figured out how to RE-caculate.
Ahh. Not good.
> start the machine with the caps lock (ASCII/CC) key down then
> switch in jiffy and enable it) just to read in the character set you can do
> somthing like wait 1:sys 65378 and push the ascii capslock key
> while in the wait.
Hmm, isn't that the SYS to get the 80 column screen to re-load the character
set? The one that is issued every time the ASCII/CC key changes position? Or am
I totally wrong here? If not, what does it do?
BTW, are you still reading R20_8BITAR, the Swedish eight bit computer echo on
Fidonet? Haven't seen you there lately.
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
- Fido 2:204/145.42
... "Everything's gotta have rules, rules, rules!" -- Calvin
--- FleetStreet 1.15+
Peter Karlsson
> Out Keypatcher actually fixes this! You gat access to the
> underscore, the checkmark
Ah, nice. I'd better use this program sometime, then.
> Well the actual flipping of the chars is still as fast, but the
> cursor is gone for ~half a second...
Hmm, never thought of that. That might be reloading of keytables, then. Or,
rather, changing of pointers (the pointers are in RAM if I remember correctly).
PTK>> I don't have a dump of the JiffyDOS chip, only of my Swedish
PTK>> KERNAL.
> You could read out the JiffyKernal and patch ...
I wonder if that will work, with the switch and everyting.
But I do have a couple of 'empty' EPROMS that I could use, one of them has the
same numbers as the original ROM, and some others seem to be the double amount.
Maybe I could use one of them, and install my own 'switch'?
> With a pure patched Swedish kernal, you'd need no switching back to a fully
> english one!
Probably. If I install the original Swedish kernel in the place of the original
English in the JiffyDOS, I would have access to both the original English and
the original Swedish. The problem is, that I would have to modify the JiffyDOS
English part, and since I'm not very good at that stuff, I don't know if I
could do it.
But then again, I have access to real hackers in via this newsgroup :-)
If I were to read out the JiffyDOS EPROM, and send that to you together with
the Swedish EPROM of the 128D, could you then 'hack' it so it works in Swedish?
> ! Origin: BoArDgasm, Farsta, Sweden
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
- Fido 2:204/145.42
... Brought to you by TAG.FRX v1.32 (c) 1995 Peter Karlsson
--- FleetStreet 1.15+
Peter Karlsson
On 96-03-25, one could hear David say this in CBM computer discussions:
> Here is the layout of all 64K of OS ROM in the 128(D) and 128DCR:
> Use Start End Size 128DCR K 128 upgr K
> -!- ----- --- ---- ------ - -------- -
> BASlo $4000 $7fff 16K 318022-02 0-16 318018-0x 0-16
> BAShi $8000 $bfff 16K 318022-02 16-32 318019-0x 0-16
> BAS2 $a000 $bfff 8K 318023-02 0-8 251913-01 0-8
> K64 $e000 $ffff 8K 318023-02 8-16 251913-01 8-16
> MLmon $c000 $cfff 4K 318023-02 16-20 318020-0y 0-4
> ZBIOS $0000 $0fff 4K 318023-02 20-24 318020-0y 4-8
> K128 $e000 $ffff 8K 318023-02 24-32 318020-0y 8-16
Hmm, this would make it 64 KBytes, but when I read the KERNEL chip, I only got
32 KBytes?? There *is* only one ROM/EPROM chip in the computer (if you don't
count the C1571D ROM chip that is situated right under the drive bay).
> The 123456-12 numbers are the Commodore part numbers of the ROM
> chips for the USA. In North America, "x" is either 2, 3, or 4,
> and y = x + 1.
My chip has this label:
318034-01 DCR
C128 Ker.Sw/Fi
> I think this is how character set switching works: The 128
> Programmer's Reference Guide schematic shows a jumper leading to
> the high-order address bit of the character ROM. This jumper can
> connect either to the 128/64 mode signal or to one of the locking
> keys on the keyboard.
Ahh, interesting.
> On NTSC systems this jumper connects to the
> 128/64 signal and the locking key bears the label CAPS LOCK. So
> over here the 128-mode character set is different from the 64's
> character images.
What's the difference, then?
> In Europe the jumper connects to the locking key, which on some
> keyboards is ASC/CC and on others is ASCII/DIN.
ASCII/CC is on the Swedish/Finnish model that I have. ASCII/DIN I believe is
the German keyboard. I don't know if there are any more "nationalized" versions.
> This switching may work for 80 columns only at
> reset, when the Kernal copies the character bitmaps to VDC RAM.
> Otherwise, the editor would have to monitor the key continuously.
The editor does this, yes. Whenever I change the ASCII/CC key, the characters
on the screen are re-painted to the Swedish character set. This can be
"cheated" by pressing it rapidly, and thus having all four character sets on
the screen at the same time.
> -David Gahris
> dga...@eece.maine.edu
> Please reply by E-mail.
> ! Origin: University of Maine System
Thanks so much for your comments! I'll forward this to e-mail, but please reply
in the newsgroup if possible.
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
- Fido 2:204/145.42
... If at first you do succeed, try not to look *too* astonished!
--- FleetStreet 1.15+
Peter Karlsson
On 96-03-25, one could hear David say this in CBM computer discussions:
> Oops, why don't I put in carriage returns so that you can read this post.
Personally I prefer the ones without, they look better in my reader :-)
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
- Fido 2:204/145.42
--- FleetStreet 1.15+
David Gahris
> On 96-03-25, one could hear David say this in CBM computer discussions:
> > Here is the layout of all 64K of OS ROM in the 128(D) and 128DCR:
> when I read the KERNEL chip, I only got 32 KBytes?? There *is* only one ROM/
> EPROM chip in the computer (if you don't count the C1571D ROM chip that is
> situated right under the drive bay).
There is a total of four ROMs in mine:
(1) 32k x 8 BASIC
(1) 32k x 8 OS
(1) 8k x 8 character
(1) 32k x 8 DOS
Look again - how can you run BASIC programs if your computer has no BASIC ROM?
> My chip has this label:
> 318034-01 DCR
> C128 Ker.Sw/Fi
You might want to send a dump to ftp.funet.fi, directory /pub/cbm/incoming/c128/
The Swedish/Finnish kernals already on FUNet have no part numbers listed, so
they could be different from the kernal that you have. In any case,
associating a part number with each ROM image avoids confusion between
versions.
> > On NTSC systems the 128-mode character set is different from the 64's
> > character images.
> What's the difference?
The 64-mode set is the same as on a real 64 for compatibility reasons. In
128 mode, the lowercase letters are more rounded, and the characters with
descenders (like lowercase y and g) use the bottom row of the 8 x 8 character
cell. The lowercase letters seem to be the only characters that are
different between the two sets. The 128-mode characters in the US may be the
same as the Commodore 16 character set on FUNet:
http://ftp.funet.fi/pub/cbm/firmware/characters/characters-c16
----------------------------------------------------------------
David Gahris Commodore Users Unite! E-mail address:
Sophomore EE student N1TCL dga...@eece.maine.edu
If this is a news posting, please reply by E-mail.
Pontus Berg
PTK>>> I don't have a dump of the JiffyDOS chip, only of my Swedish KERNAL.
>> You could read out the JiffyKernal and patch ...
PK> I wonder if that will work, with the switch and everyting.
I bought myself a second hand US Jiffy and pathcing it with out program was no
problem! Installing the normal kernal in the other end is no problem!
PK> But I do have a couple of 'empty' EPROMS that I could use, one of them
PK> has the same numbers as the original ROM, and some others seem to be the
PK> double amount. Maybe I could use one of them, and install my own
PK> 'switch'?
Yes for sure. They are pretty easy to deciper 27(size i kbits) ... a 2764 is 8K
(normal cart $8000-$a000). 27256 is the 32K ones for f.ex Action Replay...
>> With a pure patched Swedish kernal, you'd need no switching back to a
>> fully english one!
PK> Probably. If I install the original Swedish kernel in the place of the
PK> original English in the JiffyDOS, I would have access to both the
PK> original English and the original Swedish. The problem is, that I would
PK> have to modify the JiffyDOS English part, and since I'm not very good at
PK> that stuff, I don't know if I could do it.
Mail me and we can take that between us... Have the number for calling?
)==========================================================(
) o/ Bac...@FairLight.COM Fido: 2:201/411.71 (
) (# I'm a "geek tool" so ask about emulators elsewhere! (
)=/7=======================================================(
... Coming soon: "PONTUS V1.3 /FLT"
(Jason of Cosine)
Jouko Valta
>> cursor is gone for ~half a second...
Peter> Hmm, never thought of that. That might be reloading of
Peter> keytables, then.
Not the keyboard tables but the character generator.
Peter Karlsson
> There *is* only one ROM/ EPROM chip in the computer (if you don't
> count the C1571D ROM chip that is situated right under the drive bay).
DG> There is a total of four ROMs in mine:
DG> (1) 32k x 8 BASIC
DG> (1) 32k x 8 OS
DG> (1) 8k x 8 character
DG> (1) 32k x 8 DOS
DG> Look again - how can you run BASIC programs if your computer has no
DG> BASIC ROM?
Yeah, I know it sounds strange, but there IS only one EPROM chip in my C128D.
At least what I can find. I was asked before, and was quoted to U-numbers that
would have the chip, but I could only find one of them.
It is a 32 KB chip. It seems to contain both the C64 BASIC and the C128 BASIC.
DG> You might want to send a dump to ftp.funet.fi, directory
DG> /pub/cbm/incoming/c128/ The Swedish/Finnish kernals already on
DG> FUNet have no part numbers listed, so they could be different from
DG> the kernal that you have. In any case, associating a part number
DG> with each ROM image avoids confusion between versions.
I'll do that if I remember to :-) I'll forward your message to myself (I'm
currently offline).
Peter Karlsson
PK> I wonder if that will work, with the switch and everyting.
PB> I bought myself a second hand US Jiffy and pathcing it with out
PB> program was no problem! Installing the normal kernal in the other
PB> end is no problem!
Okay, but how do one read out the chip with the switch installed? Should I
remove the chip from the "holder"?
PB> Yes for sure. They are pretty easy to deciper 27(size i kbits) ...
PB> a 2764 is 8K (normal cart $8000-$a000). 27256 is the 32K ones for
PB> f.ex Action Replay...
Let's see. Two are labeled M27C512 - 15XFI B88BD 9348C and one is labeled
M27C256B 15XFI BHHAB 9344B. Since the KERNEL chip is 32KB (a ..256), then the
JiffyDOS EPROM would use the ..512 chip? Correct?
PB> Mail me and we can take that between us...
Netmail coming your way.
Peter Karlsson
DG> (1) 32k x 8 BASIC
DG> (1) 32k x 8 OS
DG> (1) 8k x 8 character
DG> (1) 32k x 8 DOS
DG> Look again - how can you run BASIC programs if your computer has no
DG> BASIC ROM?
PK> It is a 32 KB chip. It seems to contain both the C64 BASIC and the
PK> C128 BASIC.
Plus one with the character ROM. But when I "upgraded" to JiffyDOS, I only
switched *one* chip.
Peter Karlsson
JV> Peter> Hmm, never thought of that. That might be reloading of
JV> Peter> keytables, then.
JV> Not the keyboard tables but the character generator.
So the keytables are changed by just altering a pointer?
\\//
Peter - dat9...@idt.mdh.se - http://www.mds.mdh.se/~dat95pkn
- Fido 2:204/145.42
... I seen strange before, but this breaks new ground.