8 Dec
2024
8 Dec
'24
10:54 a.m.
The quick version : Does anyone know the exact physical and logical
disk format used by CP/M on the Philips P2000C portable computer?
The long version with explanations :
I am having reasonable success transfering files to/from disk images
for my Osborne 1A (using cpmtools) and IBM5155 (MS-DOS, of course
using editdisk). And I've even got the Greaseweazle to transfer those
images to/from real floppies. The Greaseweazle still annoys me in that
I know it's capable of a lot more if only I could work out how to do
it, but at least it does something useful
[The less said about floppy disks shedding oxide and/or suffering from
'sticky shed' the better. I'm spending far too much time dismantling
and cleaning drives....]
Any, I'd like to do the same for another of my machines, a Philips
P2000C cp/m 'portable'. My machine is the version with 2 internal 40
cylinder single head drives (about 160K each, MFM) but I can also plug
in an external 80 cylinder double head drive to handle this machine's
other native format (about 640K).
Unfortunately, this machine is not common, and neither cpmtools nor
the greaseweazle software has the formats predefined. I could add them
myself -- if I knew what they were. Things like #sectors/track, sector
size, #system tracks, skew, etc.
It's not obviously given in any of the manuals I have, so does anyone
know it before I try to work it out.
Alternatively there are rumours that the P2000C could read/write at
least one more common cp/m disk type. The hardware should be capable
of it, sure. Doe anyone know if software to do something like this
exists anywhere for the P2000C. I can't find it on any of the obvious
sites
-tony
8 Dec
8 Dec
1:42 p.m.
Tony,
Maybe this software catalogue will help you:
https://electrickery.nl/comp/p2000c/doc/PTC_Katalogus.pdf
-Rik
-----Oorspronkelijk bericht-----
Van: Tony Duell via cctalk <cctalk(a)classiccmp.org>
Verzonden: zondag 8 december 2024 19:55
Aan: General Discussion: On-Topic and Off-Topic Posts <cctalk(a)classiccmp.org>
CC: Tony Duell <ard.p850ug1(a)gmail.com>
Onderwerp: [cctalk] Philips P2000C disk format
The quick version : Does anyone know the exact physical and logical disk format used by
CP/M on the Philips P2000C portable computer?
The long version with explanations :
I am having reasonable success transfering files to/from disk images for my Osborne 1A
(using cpmtools) and IBM5155 (MS-DOS, of course using editdisk). And I've even got the
Greaseweazle to transfer those images to/from real floppies. The Greaseweazle still annoys
me in that I know it's capable of a lot more if only I could work out how to do it,
but at least it does something useful
[The less said about floppy disks shedding oxide and/or suffering from 'sticky
shed' the better. I'm spending far too much time dismantling and cleaning
drives....]
Any, I'd like to do the same for another of my machines, a Philips P2000C cp/m
'portable'. My machine is the version with 2 internal 40 cylinder single head
drives (about 160K each, MFM) but I can also plug in an external 80 cylinder double head
drive to handle this machine's other native format (about 640K).
Unfortunately, this machine is not common, and neither cpmtools nor the greaseweazle
software has the formats predefined. I could add them myself -- if I knew what they were.
Things like #sectors/track, sector size, #system tracks, skew, etc.
It's not obviously given in any of the manuals I have, so does anyone know it before I
try to work it out.
Alternatively there are rumours that the P2000C could read/write at least one more common
cp/m disk type. The hardware should be capable of it, sure. Doe anyone know if software to
do something like this exists anywhere for the P2000C. I can't find it on any of the
obvious sites
-tony
9 Dec
9 Dec
5:02 a.m.
This one seems to have information on the diskette layout starting on page
230:
https://electrickery.nl/comp/p2000c/doc/P2000C-SystemRefServiceManual.pdf
-Peter
On Sun, Dec 8, 2024 at 4:52 PM Rik Bos via cctalk <cctalk(a)classiccmp.org>
wrote:
Tony,
Maybe this software catalogue will help you:
https://electrickery.nl/comp/p2000c/doc/PTC_Katalogus.pdf
-Rik
-----Oorspronkelijk bericht-----
Van: Tony Duell via cctalk <cctalk(a)classiccmp.org>
Verzonden: zondag 8 december 2024 19:55
Aan: General Discussion: On-Topic and Off-Topic Posts <
cctalk(a)classiccmp.org>
CC: Tony Duell <ard.p850ug1(a)gmail.com>
Onderwerp: [cctalk] Philips P2000C disk format
The quick version : Does anyone know the exact physical and logical disk
format used by CP/M on the Philips P2000C portable computer?
The long version with explanations :
I am having reasonable success transfering files to/from disk images for
my Osborne 1A (using cpmtools) and IBM5155 (MS-DOS, of course using
editdisk). And I've even got the Greaseweazle to transfer those images
to/from real floppies. The Greaseweazle still annoys me in that I know it's
capable of a lot more if only I could work out how to do it, but at least
it does something useful
[The less said about floppy disks shedding oxide and/or suffering from
'sticky shed' the better. I'm spending far too much time dismantling and
cleaning drives....]
Any, I'd like to do the same for another of my machines, a Philips P2000C
cp/m 'portable'. My machine is the version with 2 internal 40 cylinder
single head drives (about 160K each, MFM) but I can also plug in an
external 80 cylinder double head drive to handle this machine's other
native format (about 640K).
Unfortunately, this machine is not common, and neither cpmtools nor the
greaseweazle software has the formats predefined. I could add them myself
-- if I knew what they were. Things like #sectors/track, sector size,
#system tracks, skew, etc.
It's not obviously given in any of the manuals I have, so does anyone know
it before I try to work it out.
Alternatively there are rumours that the P2000C could read/write at least
one more common cp/m disk type. The hardware should be capable of it, sure.
Doe anyone know if software to do something like this exists anywhere for
the P2000C. I can't find it on any of the obvious sites
-tony
5:43 a.m.
On Mon, Dec 9, 2024 at 1:02 PM Peter Ekstrom via cctalk
<cctalk(a)classiccmp.org> wrote:
This one seems to have information on the diskette layout starting on page
230:
https://electrickery.nl/comp/p2000c/doc/P2000C-SystemRefServiceManual.pdf
Unofortunately that doesn't go far enough. It doesn't give the number
of sectors/track or their size (I think it's 16 sectors, each of 256
bytes)
It also doesn't give the 'skew. Under CP/M the sectors may not be used
in numerical order, maybe it uses every third one until all are used
then goes on to the next track. This is something that is very hard to
determine by lookng at the disk but is obviously essential to know to
make use of the diak image.
-tony
6:26 a.m.
Out of curiosity I downloaded the P2000C system disk images from Dave
Dunfield's archive and looked through them in Linux using the strings
command.
Turns out, in the 4th disk image (P2000_4.IMD) there appears to be the
assembler source for a format command that can handle multiple diskette
formats.
There are table entries for the P2000C's 160KB format as well. I haven't
had time to look through it all and figure out all the values but perhaps
that can be
a good starting point?
As an example, a couple of snippets of what I found:
;==============================================
;THIS FILE HAS THE DISK TABLES FOR UTILITY
;THE TABLE HAS FOLLOWING STRUCTURE:
;BYTES 00-01 LENGTH OF TH
B 0E5H ;FORMAT INFO
DEFB 10H ;
OF SECTORS ON SIDE-0
DEFB 1,02H,03H,04H,05H,06H,07H,08H ;SECTORS IN TRACK FORMAT
DEFB 9,0AH,0BH,0CH,0DH,0EH,0FH,10H
DEFB 10H ;
OF SECTORS ON SIDE-1
DEFB 1,02H,03H,04H,05H,06H,07H,08H ;SECTORS IN TRACK FORMAT
DEF
;DESCRIPTOR FOR 2000C 160K DISK
DT2: DEFB DT3-DT2 ;TABLE LENGTH
DEFT 'P2000C 160K - CP/M' ;NAME
FORMATTING INFO (E5, OR F6)
; 1 BYTE FOR THE
OF SECTORS ON SIDE-0
; BYTES FOR THE SECTOR NUMBERS ON SIDE-0
; 1 BYTE FOR THE
OF SECTORS ON SIDE-1
; BYTES FOR THE SECTOR NUMBERS ON SIDE-1
; SYSTEM IDENTIFIER 1=CP/M
; 3=MSDOS
; IF MSDOS: 1
DEFB 81H ;FLAG FOR SINGLE SIDED
DEFB 28H ;NUMBER OF TRACKS
DEFB 1 ;TR MULTIPLIER ACTIVE
SRL H
RET ;TRANSL SUBR FOR BIOS
DEFB 0 ;TO MAKE UP 4 BYTES
DEFB 1 ;1 SUBTABLE
DT21: DEFB DT22-DT21 ;LENGTH OF SUBTABLE
DEFT 'CP/M' ;SUBTABLE NAME
DEFB 1 ;SECT LGTH = 256 BYTES
DEFB 0E5H ;FORMAT INFO
DEFB 10H ;
OF SECTORS ON SIDE-0
DEFB 1,02H,03H,04H,05H,06H,07H,08H ;SECTORS IN TRACK FORMAT
DEFB 9,0AH,0BH,0CH,0DH,0EH,0FH,10H
DEFB 0 ;
OF SECTORS ON SIDE-1
DEFB 1 ;FOR CP/M
DEFB 20H,0,
7,08,11,12,15,16
DEFB 19,20,23,24,27,28,31,32
On Mon, Dec 9, 2024 at 8:44 AM Tony Duell <ard.p850ug1(a)gmail.com> wrote:
On Mon, Dec 9, 2024 at 1:02 PM Peter Ekstrom via
cctalk
<cctalk(a)classiccmp.org> wrote:
This one seems to have information on the diskette layout starting on
page
230:
https://electrickery.nl/comp/p2000c/doc/P2000C-SystemRefServiceManual.pdf
Unofortunately that doesn't go far enough. It doesn't give the number
of sectors/track or their size (I think it's 16 sectors, each of 256
bytes)
It also doesn't give the 'skew. Under CP/M the sectors may not be used
in numerical order, maybe it uses every third one until all are used
then goes on to the next track. This is something that is very hard to
determine by lookng at the disk but is obviously essential to know to
make use of the diak image.
-tony
8:23 p.m.
On Mon, Dec 9, 2024 at 7:52 PM Peter Ekstrom via cctalk
<cctalk(a)classiccmp.org> wrote:
Out of curiosity I downloaded the P2000C system disk images from Dave
Dunfield's archive and looked through them in Linux using the strings
command.
Turns out, in the 4th disk image (P2000_4.IMD) there appears to be the
assembler source for a format command that can handle multiple diskette
formats.
There are table entries for the P2000C's 160KB format as well. I haven't
had time to look through it all and figure out all the values but perhaps
that can be
a good starting point?
Interesting, I will take a look.
-tony
1:09 p.m.
Teaching granny to suck eggs:
STAT DSK: on a running CP/M machine gives some information, including the
number of reserve tracks, and the records per block, which is useful when
parsing the DIRectory entries. The "sectors per track" is actually 128
byte records, not the physical sectors. Although if it has an odd number
of physical sectors per track, that could help.
On Mon, 9 Dec 2024, Tony Duell via cctalk wrote:
Unofortunately that doesn't go far enough. It
doesn't give the number
of sectors/track or their size (I think it's 16 sectors, each of 256
bytes)
For bytes per sector, and sectors per track, write some trivial code.
Or, on the PC, you could use the 765 "read ID" command.
or, on a machine with WD controller, try a track read.
OR write some trivial code with Int1Eh and Int13h to read a 256 byte
sector. (NOT one of the ones in the reserve tracks; those sometimes have
a different format)
It also doesn't give the 'skew. Under CP/M the
sectors may not be used
in numerical order, maybe it uses every third one until all are used
then goes on to the next track. This is something that is very hard to
determine by lookng at the disk but is obviously essential to know to
make use of the diak image.
Statements found on the web can be misleading, or even wrong. I have seen
"skew of 3" used to mean every third sector, skip three sectors between
one and the next, or that the track is broken into three passes.
If you have the WD "track read", you can look at it, looking for text,
source code, or other recognizable content that extends past the end of a
sector, and look for which sector it continues in.
Otherwise, you need to write, or find, a sector editor that can be patched
for different sector sizes. For example, I patched Roxton Baaker's
"TRAKCESS" on the TRS80. And the first function that I
implemented/completed in the never completed XenoXap was displaying
sectors of various sizes.
I kept adding extravagant new functions, and never finished more than I
needed at the time of the functions (cf Randy Cook's "TRS-DOS" and
"VTOS")
Use sectors past the reserved tracks, as some formats have a different
format on the reservec tracks.
Then you can look for the "half a worm", to try to find what sequence the
sectors are.
Watch out for any "deleted files" whose content is still on the disk.
They could work, or they could tend to be misleading. ("Which of these two
almost identical sectors is the real next one, and which is residual
content from previous edits")
Disks that have been used in one format, and then re-formatted in
another format can be misleading. (a disk with 256 byte sectors on the
first side and 512 byte sectors on the second side, is probably a re-used
disk)
Text, even the messages in a program, and/or source code work well.
If you really know your machine language, you can look for the first
byte(s) of a multi-byte instruction. I never got skilled at that.
For double sided disks, ANOTHER issue that will require looking at some
sectors is how the disk handles using the second side. You will need a
disk that is over half full. Over half full, but not completely full is
by far the easiest.
Some formats fill the first side, and then start at track 0 of the second
side going up.
Some formats fill the first side, and then start after the reserve
tracks of the second side, and go up.
Some formats fill the first side, and then start at track 79, 39, or 34 of
the second side going down.
Some formats fill the first track of the first side, and then have a track
on the second side.
Some formats treat both track of each cylinder as if it were one large
track. Skew can bridge both sides of the cylinder.
Some formats will come up with even weirder approaches.
--
Grumpy Ol' Fred cisin(a)xenosoft.com
16
days inactive
18
days old
6 comments
4 participants
participants (4)
-
Fred Cisin -
hp-fix＠xs4all.nl -
Peter Ekstrom -
Tony Duell