diff --git a/doc/disk-40track_drive.md b/doc/disk-40track_drive.md index f5ce0ddb..203628f8 100644 --- a/doc/disk-40track_drive.md +++ b/doc/disk-40track_drive.md @@ -1,2 +1,15 @@ --d -\r +40track_drive +==== +## Adjust configuration for a 40-track drive + + +This is an extension profile; adding this to the command line will configure +FluxEngine to read from 40-track, 48tpi 5.25" drives. You have to tell it because there is +no way to detect this automatically. + +For example: + +``` +fluxengine read ibm --180 40track_drive +``` + diff --git a/doc/disk-acornadfs.md b/doc/disk-acornadfs.md index f5ce0ddb..2e63eff3 100644 --- a/doc/disk-acornadfs.md +++ b/doc/disk-acornadfs.md @@ -1,2 +1,39 @@ --d -\r +acornadfs +==== +## BBC Micro, Archimedes + + +Acorn ADFS disks are used by the 6502-based BBC Micro and ARM-based Archimedes +series of computers. They are yet another variation on MFM encoded IBM scheme +disks, although with different sector sizes and with the 0-based sector +identifiers rather than 1-based sector identifiers. The index hole is ignored +and sectors are written whereever, requiring FluxEngine to do two revolutions +to read a disk. + +There are various different kinds, which should all work out of the box. + +Be aware that Acorn logical block numbering goes all the way up side 0 and +then all the way up side 1. However, FluxEngine uses traditional disk images +with alternating sides, with the blocks from track 0 side 0 then track 0 side +1 then track 1 side 0 etc. Most Acorn emulators will use both formats, but +they might require nudging as the side order can't be reliably autodetected. + +## Options + + - Format variants: + - `160`: 160kB 3.5" or 5.25" 40-track SSDD; S format + - `320`: 320kB 3.5" or 5.25" 80-track SSDD; M format + - `640`: 640kB 3.5" or 5.25" 80-track DSDD; L format + - `800`: 800kB 3.5" 80-track DSDD; D and E formats + - `1600`: 1600kB 3.5" 80-track DSHD; F formats + +## Examples + +To read: + + - `fluxengine read acornadfs --160 -s drive:0 -o acornadfs.img` + - `fluxengine read acornadfs --320 -s drive:0 -o acornadfs.img` + - `fluxengine read acornadfs --640 -s drive:0 -o acornadfs.img` + - `fluxengine read acornadfs --800 -s drive:0 -o acornadfs.img` + - `fluxengine read acornadfs --1600 -s drive:0 -o acornadfs.img` + diff --git a/doc/disk-acorndfs.md b/doc/disk-acorndfs.md index f5ce0ddb..a36b0ddc 100644 --- a/doc/disk-acorndfs.md +++ b/doc/disk-acorndfs.md @@ -1,2 +1,38 @@ --d -\r +acorndfs +==== +## Acorn Atom, BBC Micro series + + +Acorn DFS disks are used by the Acorn Atom and BBC Micro series of computers. +They are pretty standard FM encoded IBM scheme disks, with 256-sectors and +0-based sector identifiers. There's nothing particularly special here. + +DFS disks are all single-sided, but allow the other side of the disk to be +used as another volume. + +They come in two varieties, 40 track and 80 track. These should both work. +Some rare disks are both at the same time. FluxEngine can read these but it +requires a bit of fiddling as they have the same tracks on twice. + +## Options + + - Format variants: + - `100`: 100kB 40-track SSSD + - `200`: 200kB 80-track SSSD + +## Examples + +To read: + + - `fluxengine read acorndfs --100 -s drive:0 -o acorndfs.img` + - `fluxengine read acorndfs --200 -s drive:0 -o acorndfs.img` + +To write: + + - `fluxengine write acorndfs --100 -d drive:0 -i acorndfs.img` + - `fluxengine write acorndfs --200 -d drive:0 -i acorndfs.img` + +## References + + - [The Acorn DFS disc format](https://beebwiki.mdfs.net/Acorn_DFS_disc_format) + diff --git a/doc/disk-aeslanier.md b/doc/disk-aeslanier.md index f5ce0ddb..73073fc3 100644 --- a/doc/disk-aeslanier.md +++ b/doc/disk-aeslanier.md @@ -1,2 +1,48 @@ --d -\r +aeslanier +==== +## 616kB 5.25" 77-track SSDD hard sectored + + +Back in 1980 Lanier released a series of very early integrated word processor +appliances, the No Problem. These were actually [rebranded AES Data Superplus +machines](http://vintagecomputers.site90.net/aes/). They were gigantic, +weighed 40kg, and one example I've found cost ꆲ5bb4 +of nearly ㇢1a6e + +8080 machines with 32kB of RAM, they ran their own proprietary word +processing software off twin 5.25" drive units, but apparently other software +was available. + +The disk format is exceptionally weird. They used 77 track, 32 sector, single-sided +_hard_ sectored disks, where there were multiple index holes, +indicating to the hardware where the sectors start. The encoding scheme +itself is [MMFM (aka +M2FM)](http://www.retrotechnology.com/herbs_stuff/m2fm.html), an early +attempt at double-density disk encoding which rapidly got obsoleted by the +simpler MFM --- and the bytes are stored on disk _backwards_. Even aside from +the encoding, the format on disk was strange; unified sector header/data +records, so that the sector header (containing the sector and track number) +is actually inside the user data. + +FluxEngine can read these, but I only have a single, fairly poor example of a +disk image, and I've had to make a lot of guesses as to the sector format +based on what looks right. If anyone knows _anything_ about these disks, +[please get in touch](https://github.com/davidgiven/fluxengine/issues/new). + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read aeslanier -s drive:0 -o aeslanier.img` + +## References + + * [SA800 Diskette Storage Drive - Theory Of + Operations](http://www.hartetechnologies.com/manuals/Shugart/50664-1_SA800_TheorOp_May78.pdf): + talks about MMFM a lot, but the Lanier machines didn't use this disk + format. + diff --git a/doc/disk-agat.md b/doc/disk-agat.md index f5ce0ddb..dc2b579d 100644 --- a/doc/disk-agat.md +++ b/doc/disk-agat.md @@ -1,2 +1,36 @@ --d -\r +agat +==== +## 840kB 5.25" 80-track DS + + +The Agat (Russian: ↊fd74 +1983. These were based around a 6502 and were nominally Apple II-compatible +although with enough differences to be problematic. + +They could use either standard Apple II 140kB disks, or a proprietary 840kb +MFM-based double-sided format. FluxEngine supports both of these; this profile +is for the proprietary format. for the Apple II format, use the `apple2` +profile. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read agat -s drive:0 -o agat.img` + +To write: + + - `fluxengine write agat -d drive:0 -i agat.img` + +## References + + - [Magazine article on the + Agat](https://sudonull.com/post/54185-Is-AGAT-a-bad-copy-of-Apple) + + - [Forum thread with (some) documentation on the + format](https://torlus.com/floppy/forum/viewtopic.php?t=1385) + diff --git a/doc/disk-amiga.md b/doc/disk-amiga.md index f5ce0ddb..fe40ed94 100644 --- a/doc/disk-amiga.md +++ b/doc/disk-amiga.md @@ -1,2 +1,41 @@ --d -\r +amiga +==== +## 880kB 3.5" DSDD + + +Amiga disks use MFM, but don't use IBM scheme. Instead, the entire track is +read and written as a unit, with each sector butting up against the previous +one. This saves a lot of space which allows the Amiga to not just store 880kB +on a DD disk, but _also_ allows an extra 16 bytes of metadata per sector. + +This metadata is mostly unused, so the default for FluxEngine is to ignore it +and just use the 512 bytes of main sector data. If you want it, specify a +528-byte sector size. The metadata will come after the user data. + +Bizarrely, the data in each sector is stored with all the odd bits first, and +then all the even bits. This is tied into the checksum algorithm, which is +distinctly subpar and not particularly good at detecting errors. + +## Options + + - Sector size: + - `without_metadata`: 512-byte sectors + - `with_metadata`: 528-byte sectors + +## Examples + +To read: + + - `fluxengine read amiga -s drive:0 -o amiga.adf` + +To write: + + - `fluxengine write amiga -d drive:0 -i amiga.adf` + +## References + + - [The Amiga Floppy Boot Process and Physical + Layout](https://wiki.amigaos.net/wiki/Amiga_Floppy_Boot_Process_and_Physical_Layout) + + - [The Amiga Disk File FAQ](http://lclevy.free.fr/adflib/adf_info.html) + diff --git a/doc/disk-ampro.md b/doc/disk-ampro.md index f5ce0ddb..f1b51f32 100644 --- a/doc/disk-ampro.md +++ b/doc/disk-ampro.md @@ -1,2 +1,52 @@ --d -\r +ampro +==== +## CP/M + + +The Ampro Little Board was a very simple and cheap Z80-based computer from +1984, which ran CP/M. It was, in fact, a single PCB which you could mount +on the bottom of a 5.25" drive. + +[All about the Ampro Little Board](http://oldcomputers.net/ampro-little-board.html) + +It stored either 400kB on a double-sided 40-track drive or 800kB on a +double-sided 80 track drive. The disk format it used was a slightly quirky +variation of the standard MFM IBM scheme --- sector numbering starts at 17 +rather than 1 (or Acorn's 0). FluxEngine supports this. + +FluxEngine has direct filesystem support for these disks, or you can pass the +disk images into [cpmtools](http://www.moria.de/~michael/cpmtools/): + +``` +$ cpmls -f ampdsdd ampro.img +0: +-a60014.e +amprodsk.com +bitchk.doc +bitchk.mac +cpmmac.mac +dir.com +himem.doc +himem.mac +kaydiag.lbr +kayinfo.lbr +...etc... +``` + +## Options + + - Format variants: + - `400`: 400kB 40-track DSDD + - `800`: 800kB 80-track DSDD + +## Examples + +To read: + + - `fluxengine read ampro --400 -s drive:0 -o ampro.img` + - `fluxengine read ampro --800 -s drive:0 -o ampro.img` + +## References + + - [The Ampro Little Board](http://oldcomputers.net/ampro-little-board.html) + diff --git a/doc/disk-apple2.md b/doc/disk-apple2.md index f5ce0ddb..d866a497 100644 --- a/doc/disk-apple2.md +++ b/doc/disk-apple2.md @@ -1,2 +1,74 @@ --d -\r +apple2 +==== +## Prodos, Appledos, and CP/M + + +Apple II disks are nominally fairly sensible 40-track, single-sided, 256 +bytes-per-sector jobs. However, they come in two varieties: DOS 3.3/ProDOS and +above, and pre-DOS 3.3. They use different GCR encoding systems, dubbed +6-and-2 and 5-and-3, and are mutually incompatible (although in some rare +cases you can mix 6-and-2 and 5-and-3 sectors on the same disk). + +The difference is in the drive controller; the 6-and-2 controller is capable +of a more efficient encoding, and can fit 16 sectors on a track, storing +140kB on a disk. The 5-and-3 controller can only fit 13, with a mere 114kB. + +Both formats use GCR (in different varieties) in a nice, simple grid of +sectors, unlike the Macintosh. Like the Macintosh, there's a crazy encoding +scheme applied to the data before it goes down on disk to speed up +checksumming. + +In addition, a lot of the behaviour of the drive was handled in software. +This means that Apple II disks can do all kinds of weird things, including +having spiral tracks! Copy protection for the Apple II was even madder than +on other systems. + +FluxEngine can only read well-behaved 6-and-2 disks. It doesn't even try to +handle the weird stuff. + +Apple DOS also applies logical sector remapping on top of the physical sector +numbering on the disk, and this _varies_ depending on what the disk is for. +FluxEngine can remap the sectors from physical to logical using modifiers. If +you don't specify a remapping modifier, you get the sectors in the order they +appear on the disk. + +If you don't want an image in physical sector order, specify one of the +filesystem ordering options. These also select the appropriate file system; +FluxEngine has read-only support for all of these. + +In addition, some third-party systems use 80-track double sides drives, with +the same underlying disk format. The complication here is that the AppleDOS +filesystem only supports up to 50 tracks, so it needs tweaking to support +larger disks. It treats the second side of the disk as a completely different +volume. + +## Options + + - Format variants: + - `140`: 140kB 5.25" 35-track SS + - `640`: 640kB 5.25" 80-track DS + - Filesystem and sector skew: + - `nofs`: use physical CHS sector order and no file system + - `appledos`: use AppleDOS soft sector skew and file system + - `prodos`: use ProDOS soft sector skew and filesystem + - `cpm`: use CP/M soft sector skew and filesystem + - `side1`: for AppleDOS file system access, read the volume on side 1 of a disk + +## Examples + +To read: + + - `fluxengine read apple2 --140 -s drive:0 -o apple2.img` + - `fluxengine read apple2 --640 -s drive:0 -o apple2.img` + +To write: + + - `fluxengine write apple2 --140 -d drive:0 -i apple2.img` + - `fluxengine write apple2 --640 -d drive:0 -i apple2.img` + +## References + + - [Beneath Apple DOS](https://fabiensanglard.net/fd_proxy/prince_of_persia/Beneath%20Apple%20DOS.pdf) + + - [MAME's ap2_dsk.cpp file](https://github.com/mamedev/mame/blob/4263a71e64377db11392c458b580c5ae83556bc7/src/lib/formats/ap2_dsk.cpp) + diff --git a/doc/disk-apple2_drive.md b/doc/disk-apple2_drive.md index f5ce0ddb..e631d138 100644 --- a/doc/disk-apple2_drive.md +++ b/doc/disk-apple2_drive.md @@ -1,2 +1,16 @@ --d -\r +apple2_drive +==== +## Adjust configuration for a 40-track Apple II drive + + +This is an extension profile; adding this to the command line will configure +FluxEngine to adjust the pinout and track spacing to work with an Apple II +drive. This only works on Greaseweazle hardware and requires a custom +connector. + +For example: + +``` +fluxengine read apple2 --160 apple2_drive +``` + diff --git a/doc/disk-atarist.md b/doc/disk-atarist.md index f5ce0ddb..25f3e11e 100644 --- a/doc/disk-atarist.md +++ b/doc/disk-atarist.md @@ -1,2 +1,61 @@ --d -\r +atarist +==== +## Almost PC compatible + + +Atari ST disks are standard MFM encoded IBM scheme disks without an IAM header. +Disks are typically formatted 512 bytes per sector with between 9-10 (sometimes +11!) sectors per track and 80-82 tracks per side. + +For some reason, occasionally formatting software will put an extra IDAM record +with a sector number of 66 on a disk, which can horribly confuse things. The +Atari profiles below are configured to ignore these. + +Be aware that many PC drives (including mine) won't do the 82 track formats. + +## Options + + - Format variants: + - `360`: 360kB 3.5" 80-track 9-sector SSDD + - `370`: 370kB 3.5" 82-track 9-sector SSDD + - `400`: 400kB 3.5" 80-track 10-sector SSDD + - `410`: 410kB 3.5" 82-track 10-sector SSDD + - `720`: 720kB 3.5" 80-track 9-sector DSDD + - `740`: 740kB 3.5" 82-track 9-sector DSDD + - `800`: 800kB 3.5" 80-track 10-sector DSDD + - `820`: 820kB 3.5" 82-track 10-sector DSDD + +## Examples + +To read: + + - `fluxengine read atarist --360 -s drive:0 -o atarist.img` + - `fluxengine read atarist --370 -s drive:0 -o atarist.img` + - `fluxengine read atarist --400 -s drive:0 -o atarist.img` + - `fluxengine read atarist --410 -s drive:0 -o atarist.img` + - `fluxengine read atarist --720 -s drive:0 -o atarist.img` + - `fluxengine read atarist --740 -s drive:0 -o atarist.img` + - `fluxengine read atarist --800 -s drive:0 -o atarist.img` + - `fluxengine read atarist --820 -s drive:0 -o atarist.img` + +To write: + + - `fluxengine write atarist --360 -d drive:0 -i atarist.img` + - `fluxengine write atarist --370 -d drive:0 -i atarist.img` + - `fluxengine write atarist --400 -d drive:0 -i atarist.img` + - `fluxengine write atarist --410 -d drive:0 -i atarist.img` + - `fluxengine write atarist --720 -d drive:0 -i atarist.img` + - `fluxengine write atarist --740 -d drive:0 -i atarist.img` + - `fluxengine write atarist --800 -d drive:0 -i atarist.img` + - `fluxengine write atarist --820 -d drive:0 -i atarist.img` + +## References + + - [Atari ST Floppy Drive Hardware + Information](https://info-coach.fr/atari/hardware/FD-Hard.php) by Jean + Louis-Guerin + + - [Atari ST Floppy Drive Software + Information](https://info-coach.fr/atari/software/FD-Soft.php) by Jean + Louis-Guerin + diff --git a/doc/disk-bk.md b/doc/disk-bk.md index f5ce0ddb..fa18d96a 100644 --- a/doc/disk-bk.md +++ b/doc/disk-bk.md @@ -1,2 +1,30 @@ --d -\r +bk +==== +## 800kB 5.25"/3.5" 80-track 10-sector DSDD + + +The BK (an abbreviation for 1ba9 +is a Soviet era personal computer from Elektronika based on a PDP-11 +single-chip processor. It was the _only_ official, government approved home +computer in mass production at the time. + +It got a floppy interface in 1989 when the 128kB BK-0011 was released. This +used a relatively normal double-sided IBM scheme format with 80 sectors and ten +sectors per track, resulting in 800kB disks. The format is, in fact, identical +to the Atari ST 800kB format. Either 5.25" or 3.5" drives were used depending +on what was available at the time, with the same format on both. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read bk -s drive:0 -o bk800.img` + +To write: + + - `fluxengine write bk -d drive:0 -i bk800.img` + diff --git a/doc/disk-brother.md b/doc/disk-brother.md index f5ce0ddb..6e12dced 100644 --- a/doc/disk-brother.md +++ b/doc/disk-brother.md @@ -1,2 +1,128 @@ --d -\r +brother +==== +## GCR family + + +Brother word processor disks are weird, using custom tooling and chipsets. +They are completely not PC compatible in every possible way other than the +size. + +Different word processors use different disk formats --- the only ones supported +by FluxEngine are the 120kB and 240kB 3.5" formats. Use the `--120` and `--240` +options to select which one. + +Apparently about 20% of Brother word processors have alignment issues which +means that the disks can't be read by FluxEngine (because the tracks on the disk +don't line up with the position of the head in a PC drive). The word processors +themselves solved this by microstepping until they found where the real track +is, but normal PC drives aren't capable of doing this. Particularly with the +120kB disks, you might want to fiddle with the head bias (e.g. +`--drive.head_bias=3`) to get a clean read. Keep an eye on the bad sector map +that's dumped at the end of a read. My word processor likes to put logical track +0 on physical track 3, which means that logical track 77 is on physical track +80, so I need that `head_bias` value of 3; luckily my PC drive can access track +80. + +Using FluxEngine to *write* disks isn't a problem, so the +simplest solution is to use FluxEngine to create a new disk, with the tracks +aligned properly, and then use a word processor to copy the files you want +onto it. The new disk can then be read and you can extract the files. +Obviously this sucks if you don't actually have a word processor, but I can't +do anything about that. + +If you find one of these misaligned disks then *please* [get in +touch](https://github.com/davidgiven/fluxengine/issues/new); I want to +investigate. + +## Options + + - Format variants: + - `120`: 120kB 3.5" 39-track SS GCR + - `240`: 240kB 3.5" 78-track SS GCR + +## Examples + +To read: + + - `fluxengine read brother --120 -s drive:0 -o brother.img` + - `fluxengine read brother --240 -s drive:0 -o brother.img` + +To write: + + - `fluxengine write brother --120 -d drive:0 -i brother.img` + - `fluxengine write brother --240 -d drive:0 -i brother.img` + +Dealing with misaligned disks +----------------------------- + +While FluxEngine can't read misaligned disks directly, Brother word processors +_can_. If you have access to a compatible word processor, there's a fairly +simple workaround to allow you to extract the data: + + 1. Format a disk using FluxEngine (by simply writing a blank filesystem image + to a disk). This will have the correct alignment to work on a PC drive. + + 2. Use a word processor to copy the misaligned disk to the newly formatted + disk. The machine will happily adjust itself to both sets of alignments. + + 3. Use FluxEngine to read the data off the correctly aligned disk. + +I realise this is rather unsatisfactory, as the Brother hardware is becoming +rarer and they cope rather badly with damaged disks, but this is a limitation +of the hardware of normal PC drives. (It _is_ possible to deliberately misalign +a drive to make it match up with a bad disk, but this is for experts only --- I +wouldn't dare.) + +Low level format +---------------- + +The drive is a single-sided 3.5" drive spinning at not 300 rpm (I don't know +the precise speed yet but FluxEngine doesn't care). The 240kB disks have 78 +tracks and the 120kB disks have 39. + +Each track has 12 256-byte sectors. The drive ignores the index hole so they're +lined up all anyhow. As FluxEngine can only read from index to index, it +actually reads two complete revolutions and reassembles the sectors from that. + +The underlying encoding is exceptionally weird; they use two different kinds of +GCR, one kind for the sector header records and a completely different one for +the data itself. It also has a completely bizarre CRC variant which a genius on +StackOverflow reverse engineered for me. However, odd though it may be, it does +seem pretty robust. + +See the source code for the GCR tables and CRC routine. + +Sectors are about 16.2ms apart on the disk (at 300 rpm). The header and +data records are 0.694ms apart. (All measured from the beginning of the +record.) The sector order is 05a3816b4927, which gives a sector skew of 5. + +High level format +----------------- + +Once decoded, you end up with a file system image. FluxEngine supports direct +filesystem access for both kinds of disks. + +### 120kB disks + +These disks use a proprietary and very simple file system. It's FAT-like +with an obvious directory and allocation table. It's supported by FluxEngine. + +Any files whose names begin with an asterisk (`*`) will be marked as hidden. If +the file is named `*boot`, then a boot sector will be created which will load +and run the file at 0x7000 if the machine is started with CODE+Q pressed. So +far this has only been confirmed to work on a WP-1. + +### 240kB disks + +Conversely, the 240kB disks turns out to be a completely normal Microsoft FAT +file system with a media type of 0x58 --- did you know that FAT supports 256 +byte sectors? I didn't --- of the MSX-DOS variety. There's a faint +possibility that the word processor is based on MSX-DOS, but I haven't +reverse engineered it to find out. + +Standard Linux mtools will access the filesystem image and allow you to move +files in and out. However, you'll need to change the media type bytes at +offsets 0x015 and 0x100 from 0x58 to 0xf0 before mtools will touch it. The +supplied `brother240tool` will do this. Additionally, FluxEngine's own FAT +file system supports this. + diff --git a/doc/disk-commodore.md b/doc/disk-commodore.md index f5ce0ddb..c2ad45f9 100644 --- a/doc/disk-commodore.md +++ b/doc/disk-commodore.md @@ -1,2 +1,74 @@ --d -\r +commodore +==== +## 1541, 1581, 8050 and variations + + +Commodore 8-bit computer disks come in two varieties: GCR, which are the +overwhelming majority; and MFM, only used on the 1571 and 1581. The latter were +(as far as I can tell) standard IBM PC format disks with a slightly odd sector +count. + +The GCR disks are much more interesting. They could store 170kB on a +single-sided disk (although later drives were double-sided), using a proprietary +encoding and record scheme; like [Apple Macintosh disks](macintosh.md) they +stored varying numbers of sectors per track to make the most of the physical +disk area, although unlike them they did it by changing the bitrate rather than +adjusting the motor speed. + +The drives were also intelligent and ran DOS on a CPU inside them. The +computer itself knew nothing about file systems. You could even upload +programs onto the drive and run them there, allowing all sorts of custom disk +formats, although this was mostly used to compensate for the [cripplingly +slow connection to the +computer](https://ilesj.wordpress.com/2014/05/14/1541-why-so-complicated/) of +300 bytes per second (!). (The drive itself could transfer data reasonably +quickly.) + + - a 1541 disk has 35 tracks of 17 to 21 sectors, each 256 bytes long + (sometimes 40 tracks), and uses GCR encoding. + + - a standard 1581 disk has 80 tracks and two sides, each with 10 sectors, 512 + bytes long, and uses normal IBM encoding. + + - an 8050 disk has 77 tracks and two sides, with four speed zones; the number + of sectors varies from 23 to 29, using GCR encoding. These will store + 1042kB. These drives are peculiar because they are 100tpi and therefore the + disks cannot be read in normal 96tpi drives. + + - a CMD FD2000 disk (a popular third-party Commodore disk drive) has 81 + tracks and two sides, each with 10 1024-byte sectors, for a massive 1620kB + of storage. This also uses IBM encoding. + +A CMD FD2000 disk (a popular third-party Commodore disk drive) + +## Options + + - Format variants: + - `171`: 171kB 1541, 35-track variant + - `192`: 192kB 1541, 40-track variant + - `800`: 800kB 3.5" 1581 + - `1042`: 1042kB 5.25" 8051 + - `1620`: 1620kB, CMD FD2000 + +## Examples + +To read: + + - `fluxengine read commodore --171 -s drive:0 -o commodore.d64` + - `fluxengine read commodore --192 -s drive:0 -o commodore.d64` + - `fluxengine read commodore --800 -s drive:0 -o commodore.d64` + - `fluxengine read commodore --1042 -s drive:0 -o commodore.d64` + - `fluxengine read commodore --1620 -s drive:0 -o commodore.d64` + +To write: + + - `fluxengine write commodore --171 -d drive:0 -i commodore.d64` + - `fluxengine write commodore --192 -d drive:0 -i commodore.d64` + - `fluxengine write commodore --800 -d drive:0 -i commodore.d64` + - `fluxengine write commodore --1620 -d drive:0 -i commodore.d64` + +## References + + - [Ruud's Commodore Site: 1541](http://www.baltissen.org/newhtm/1541c.htm): + documentation on the 1541 disk format. + diff --git a/doc/disk-eco1.md b/doc/disk-eco1.md index f5ce0ddb..f31ca3a1 100644 --- a/doc/disk-eco1.md +++ b/doc/disk-eco1.md @@ -1,2 +1,42 @@ --d -\r +eco1 +==== +## CP/M; 1210kB 77-track mixed format DSHD + + +The Eco1 is a Italian CP/M machine produced in 1982. It had 64kB of RAM, in +later models expandable up to 384kB, and _two_ Z80 processors. One of these was +used solely for managing the twin 8" drives, each storing 1.2MB, which was +quite impressive for a CP/M machine in those days. Visually it is best +described as 'very brown'. + +
+ A contemporary advert for the Eco1 +
+ +Its format is standard IBM scheme, but with an interesting wrinkle: there are +_three_ different formatting zones on the disk: + + - Track 0 side 0: 26 sectors, 128 bytes per sector (3296 bytes) + - Track 0 side 1: 26 sectors, 256 bytes per sector (6656 bytes) + - All others: 16 sectors, 512 bytes per sector (8192 bytes) + +The standard `read ibm` command will autodetect and read these disks, but due +to the format confusing the size autodetection the images need postprocessing +to be useful, so there's a custom profile for the Eco1 which produces sensible +images. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read eco1 -s drive:0 -o eco1.img` + +## References + + - [Apulio Retrocomputing's page on the + Eco1](https://www.apuliaretrocomputing.it/wordpress/?p=8976) + diff --git a/doc/disk-epsonpf10.md b/doc/disk-epsonpf10.md index f5ce0ddb..4f272f09 100644 --- a/doc/disk-epsonpf10.md +++ b/doc/disk-epsonpf10.md @@ -1,2 +1,19 @@ --d -\r +epsonpf10 +==== +## CP/M; 3.5" 40-track DSDD + + +The Epson PF10 is the disk unit for the Epson Z80 series of 'laptops', running +CP/M. It uses a single-sided 40-track 3.5" format, which is unusual, but the +format itself is yet another IBM scheme variant. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read epsonpf10 -s drive:0 -o epsonpf10.img` + diff --git a/doc/disk-f85.md b/doc/disk-f85.md index f5ce0ddb..3f99b343 100644 --- a/doc/disk-f85.md +++ b/doc/disk-f85.md @@ -1,2 +1,47 @@ --d -\r +f85 +==== +## 461kB 5.25" 77-track SS + + +The Durango F85 was an early office computer based around a 5MHz 8085 processor, +sold in 1977. It had an impressive 64kB of RAM, upgradable to 128kB, and ran +its own multitasking operating system call DX-85M, as well as CP/M. It had an +interesting electric-typewriter form factor, with a little monitor sitting on +the side of it --- in operation you were facing the 14" printer. + +It was touted as being portable. Which it was, if you were strong; the story +is that they had to do an extensive search to find someone capable of lifting +it for the following photo... + +
+A Durango F85, held precariously +
+ +...and even then, they had to airbrush out the tendons in her neck from the +effort! + +It used 5.25 soft-sectored disks storing an impressive-for-those-days +480kBish on a side, using a proprietary 4-in-5 GCR encoding. They used 77 +tracks, 12 sectors and 512 bytes per sector. Later models used double-sided +disks; I don't have access to an image of one so don't know how they work +(there's a suspicious looking spare byte in the sector header which could +store the side). As always, if you have one, please [get in +touch](https://github.com/davidgiven/fluxengine/issues/new). + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read f85 -s drive:0 -o f85.img` + +## References + +There's amazingly little information about these things. + + * [Chuck Guzis' F85 page](http://www.sydex.com/durango/durango.html) with + lots of pictures + diff --git a/doc/disk-fb100.md b/doc/disk-fb100.md index f5ce0ddb..2ba81422 100644 --- a/doc/disk-fb100.md +++ b/doc/disk-fb100.md @@ -1,2 +1,48 @@ --d -\r +fb100 +==== +## 100kB 3.5" 40-track SSSD + + +The Brother FB-100 is a serial-attached smart floppy drive used by a several +different machines for mass storage, including the Tandy Model 100 and +clones, the Husky Hunter 2, and (bizarrely) several knitting machines. It was +usually rebadged, sometimes with a cheap paper label stuck over the Brother +logo, but the most common variant appears to be the Tandy Portable Disk Drive +or TPDD: + +
+ A Tandy Portable Disk Drive +
+ +It's a bit of an oddball: the disk encoding is FM with a very custom record +scheme: 40-track single-sided 3.5" disks storing 100kB or so each. Each track +had only _two_ sectors, each 1280 bytes, but with an additional 12 bytes of +ID data used for filesystem management. + +There was also apparently a TPDD-2 which could store twice as much data, but +I don't have access to one of those disks. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read fb100 -s drive:0 -o fb100.img` + +## References + + - [Tandy Portable Disk Drive operations + manual](http://www.classiccmp.org/cini/pdf/Tandy/Portable%20Disk%20Drive%20Operation%20Manual.pdf) + + - [Tandy Portable Disk Drive service + manual](https://archive.org/details/TandyPortableDiskDriveSoftwareManual26-3808s) + + - [TPDD design notes (including a dump of the + ROM)](http://bitchin100.com/wiki/index.php?title=TPDD_Design_Notes) + + - [Knitting machine FB-100 + resources](http://www.k2g2.org/wiki:brother_fb-100) + diff --git a/doc/disk-hplif.md b/doc/disk-hplif.md index f5ce0ddb..f5cf22f6 100644 --- a/doc/disk-hplif.md +++ b/doc/disk-hplif.md @@ -1,2 +1,42 @@ --d -\r +hplif +==== +## a variety of disk formats used by HP + + +LIF, a.k.a. Logical Interchange Format, is a series of formats used by +Hewlett-Packard across their entire range of computers, from calculators to +modern servers. It also defines a simple non-hierarchical filesystem which is +bizarrely _still_ supported by HP-UX systems. + +Floppy-disk wise, they're yet more variations of the standard IBM floppy +encoding scheme. + +## Options + + - Format variants: + - `264`: 264kB 3.5" 66-track SSDD; HP9121 format + - `608`: 608kB 3.5" 76-track DSDD; HP9122 format + - `616`: 616kB 3.5" 77-track DSDD + - `770`: 770kB 3.5" 77-track DSDD + +## Examples + +To read: + + - `fluxengine read hplif --264 -s drive:0 -o hplif.img` + - `fluxengine read hplif --608 -s drive:0 -o hplif.img` + - `fluxengine read hplif --616 -s drive:0 -o hplif.img` + - `fluxengine read hplif --770 -s drive:0 -o hplif.img` + +To write: + + - `fluxengine write hplif --264 -d drive:0 -i hplif.img` + - `fluxengine write hplif --608 -d drive:0 -i hplif.img` + - `fluxengine write hplif --616 -d drive:0 -i hplif.img` + - `fluxengine write hplif --770 -d drive:0 -i hplif.img` + +## References + + * [A summary of the Hewlett Packard floppy disk + formats](http://www.bitsavers.org/pdf/hp/disc/912x/HP_Flexible_Disk_Formats.pdf) + diff --git a/doc/disk-ibm.md b/doc/disk-ibm.md index f5ce0ddb..742be528 100644 --- a/doc/disk-ibm.md +++ b/doc/disk-ibm.md @@ -1,2 +1,122 @@ --d -\r +ibm +==== +## Generic PC 3.5"/5.25" disks + + +IBM scheme disks are _the_ most common disk format, ever. They're used by a +huge variety of different systems, and they come in a huge variety of different +forms, but they're all fundamentally the same: either FM or MFM, either single- +or double-sided, with distinct sector header and data records and no sector +metadata. Systems which use IBM scheme disks include but are not limited to: + + - IBM PCs (naturally) + - Atari ST + - late era Apple machines + - Acorn machines + - the TRS-80 + - late era Commodore machines (the 1571 and so on) + - most CP/M machines + - NEC PC-88 series + - NEC PC-98 series + - Sharp X68000 + - Fujitsu FM Towns + - VAX & PDP-11 + - etc + +FluxEngine supports reading these. However, some variants are more peculiar +than others, and as a result there are specific decoders which set the defaults +correctly for certain formats (for example: on PC disks the sector numbers +start from 1, but on Acorn disks they start from 0). The IBM decoder described +here is the generic one, and is suited for 'conventional' PC disks. While you +can read all the variant formats with it if you use the right set of arguments, +it's easier to use the specific decoder. + +There is a generic decoder which should adjust automatically to whichever disk +format you are using, but it's unreliable and not recommended. This format +should also be used if you are writing images such as DIM which specify the +image format. FluxEngine will use these parameters. + +## Options + + - Format variants: + - `auto`: try to autodetect the format (unreliable) + - `160`: 160kB 5.25" 40-track 8-sector SSDD + - `180`: 180kB 5.25" 40-track 9-sector SSDD + - `320`: 320kB 5.25" 40-track 8-sector DSDD + - `360`: 360kB 5.25" 40-track 9-sector DSDD + - `720_96`: 720kB 5.25" 80-track 9-sector DSDD + - `720_135`: 720kB 3.5" 80-track 9-sector DSDD + - `1200`: 1200kB 5.25" 80-track 15-sector DSHD + - `1232`: 1232kB 5.25" 77-track 8-sector DSHD + - `1440`: 1440kB 3.5" 80-track 18-sector DSHD + - `1680`: 1680kB 3.5" 80-track 21-sector DSHD; DMF + +## Examples + +To read: + + - `fluxengine read ibm --auto -s drive:0 -o ibm.img` + - `fluxengine read ibm --160 -s drive:0 -o ibm.img` + - `fluxengine read ibm --180 -s drive:0 -o ibm.img` + - `fluxengine read ibm --320 -s drive:0 -o ibm.img` + - `fluxengine read ibm --360 -s drive:0 -o ibm.img` + - `fluxengine read ibm --720_96 -s drive:0 -o ibm.img` + - `fluxengine read ibm --720_135 -s drive:0 -o ibm.img` + - `fluxengine read ibm --1200 -s drive:0 -o ibm.img` + - `fluxengine read ibm --1232 -s drive:0 -o ibm.img` + - `fluxengine read ibm --1440 -s drive:0 -o ibm.img` + - `fluxengine read ibm --1680 -s drive:0 -o ibm.img` + +To write: + + - `fluxengine write ibm --160 -d drive:0 -i ibm.img` + - `fluxengine write ibm --180 -d drive:0 -i ibm.img` + - `fluxengine write ibm --320 -d drive:0 -i ibm.img` + - `fluxengine write ibm --360 -d drive:0 -i ibm.img` + - `fluxengine write ibm --720_96 -d drive:0 -i ibm.img` + - `fluxengine write ibm --720_135 -d drive:0 -i ibm.img` + - `fluxengine write ibm --1200 -d drive:0 -i ibm.img` + - `fluxengine write ibm --1232 -d drive:0 -i ibm.img` + - `fluxengine write ibm --1440 -d drive:0 -i ibm.img` + - `fluxengine write ibm --1680 -d drive:0 -i ibm.img` + +Mixed-format disks +------------------ + +Some disks, such as those belonging to early CP/M machines, or N88-Basic disks +(for PC-88 and PC-98), have more than one format on the disk at once. Typically, +the first few tracks will be low-density FM encoded and will be read by the +machine's ROM; those tracks contain new floppy drive handling code capable of +coping with MFM data, and so the rest of the disk will use that, allowing them +to store more data. + +FluxEngine can read these fine, but it tends to get a bit confused when it sees +tracks with differing numbers of sectors --- if track 0 has 32 sectors but +track 1 has 16, it will assume that sectors 16..31 are missing on track 1 and +size the image file accordingly. This can be worked around by specifying the +size of each track; see the `eco1` read profile for an example. + +N88-Basic format floppies can be written by either specifying the `n88basic` +format, or by using D88 or NFD format images which include explicit sector +layout information. + +Writing other formats can be made to work too, by creating a custom format +specifier, using the `n88basic` format as an example. +Please [get in touch](https://github.com/davidgiven/fluxengine/issues/new) if +you have specific requirements. + +360rpm 3.5" disks +----------------- + +Japanese PCs (NEC PC-98, Sharp X68000, Fujitsu FM Towns) spin their floppy +drives at 360rpm rather than the more typical 300rpm. This was done in order +to be fully backwards compatible with 5.25" disks, while using the exact +same floppy controller. Later models of the PC-9821, as well as most USB floppy +drives, feature "tri-mode" support which in addition to normal 300rpm modes, +can change their speed to read and write 360rpm DD and HD disks. + +Neither the FluxEngine or Greaseweazle hardware can currently command a +tri-mode drive to spin at 360rpm. However on both devices the FluxEngine +software is capable of both reading and writing 300rpm disks at 360rpm and vice +versa, so it shouldn't matter. + diff --git a/doc/disk-icl30.md b/doc/disk-icl30.md index f5ce0ddb..a5333a8d 100644 --- a/doc/disk-icl30.md +++ b/doc/disk-icl30.md @@ -1,2 +1,19 @@ --d -\r +icl30 +==== +## CP/M; 263kB 35-track DSSD + + +The ICL Model 30 is a reasonably standard CP/M machine using 35-track single +density disks and the traditional CP/M 128-byte secotrs --- 30 of them per +track! Other than that it's another IBM scheme variation. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read icl30 -s drive:0 -o icl30.img` + diff --git a/doc/disk-mac.md b/doc/disk-mac.md index f5ce0ddb..655c5175 100644 --- a/doc/disk-mac.md +++ b/doc/disk-mac.md @@ -1,2 +1,73 @@ --d -\r +mac +==== +## 400kB/800kB 3.5" GCR + + +Macintosh disks come in two varieties: the newer 1440kB ones, which are +perfectly ordinary PC disks you should use the `ibm` profile to read them, and +the older 800kB disks (and 400kB for the single sides ones). They have 80 +tracks and up to 12 sectors per track. + +They are also completely insane. + +It's not just the weird, custom GCR encoding. It's not just the utterly +bizarre additional encoding/checksum built on top of that where [every byte +is mutated according to the previous bytes in the +sector](https://www.bigmessowires.com/2011/10/02/crazy-disk-encoding-schemes/). +It's not just the odd way in which disks think they have four sides, two on one +side and two on the other, so that the track byte stores only the bottom 6 bits +of the track number. It's not just the way that Macintosh sectors are 524 bytes +long. No, it's the way the Macintosh drive changes speed depending on which +track it's looking at, so that each track contains a different amount of data. + +The reason for this is actually quite sensible: the tracks towards the centre +of the disk are obviously moving more slowly, so you can't pack the bits in +quite as closely (due to limitations in the magnetic media). You can use a +higher bitrate at the edge of the disk than in the middle. Many platforms, for +example the Commodore 64 1541 drive, changed bitrate this way. + +But Macintosh disks used a constant bitrate and changed the speed that the disk +spun instead to achieve the same effect... + +_Anyway_: FluxEngine will read them fine on conventional drives. Because it's +clever. + +Macintosh computers never really used the twelve bytes of metadata and the +standard for disk images is to omit it. If you want them, specify that you want +524-byte sectors. The metadata will follow the 512 bytes of user data. + +## Options + + - Format variants: + - `400`: 400kB 80-track SSDD + - `800`: 800kB 80-track DSDD + - `metadata`: read/write 524 byte sectors + +## Examples + +To read: + + - `fluxengine read mac --400 -s drive:0 -o mac.dsk` + - `fluxengine read mac --800 -s drive:0 -o mac.dsk` + +To write: + + - `fluxengine write mac --400 -d drive:0 -i mac.dsk` + - `fluxengine write mac --800 -d drive:0 -i mac.dsk` + +## References + + - [MAME's ap_dsk35.cpp file](https://github.com/mamedev/mame/blob/4263a71e64377db11392c458b580c5ae83556bc7/src/lib/formats/ap_dsk35.cpp), + without which I'd never have managed to do this + + - [Crazy Disk Encoding + Schemes](https://www.bigmessowires.com/2011/10/02/crazy-disk-encoding-schemes/), which made + me realise just how nuts the format is + + - [Les Disquettes et le drive Disk II](http://www.hackzapple.com/DISKII/DISKIITECH.HTM), an + epicly detailed writeup of the Apple II disk format (which is closely related) + + - [The DiskCopy 4.2 + format](https://www.discferret.com/wiki/Apple_DiskCopy_4.2), described on + the DiskFerret website. + diff --git a/doc/disk-micropolis.md b/doc/disk-micropolis.md index f5ce0ddb..a5c2576f 100644 --- a/doc/disk-micropolis.md +++ b/doc/disk-micropolis.md @@ -1,2 +1,87 @@ --d -\r +micropolis +==== +## 100tpi MetaFloppy disks + + +Micropolis MetaFloppy disks use MFM and hard sectors. Mod I was 48 TPI and +stored 143k per side. Mod II was 100 TPI and stored 315k per side. Each of the +16 sectors contains 266 bytes of "user data," allowing 10 bytes of metadata for +use by the operating system. Micropolis DOS (MDOS) used the metadata bytes, but +CP/M did not. + +Some later systems were Micropolis-compatible and so were also 100 TPI, like +the Vector Graphic Dual-Mode Disk Controller which was paired with a Tandon +drive. + +**Important note:** You _cannot_ read these disks with a normal PC drive, as +these drives are 96tpi. The track spacing is determined by the physical geometry +of the drive and can't be changed in software. You'll need to get hold of a +100tpi Micropolis drive. Luckily these seem to use the same connector and +pinout as a 96tpi PC 5.25" drive. In use they should be identical. + +While most operating systems use the standard Micropolis checksum, Vector +Graphic MZOS uses a unique checksum. The decoder will automatically detect +the checksum type in use; however, a specific checksum type may be forced +using the `--decoder.micropolis.checksum_type=TYPE` where TYPE is one of: + +| Checksum | Description | +|------------|-----------------------------------------| +| AUTO | Automatically detect | +| MICROPOLIS | Standard Micropolis (MDOS, CP/M, OASIS) | +| MZOS | Vector Graphic MZOS | + +Later versions of the Micropolis format supported ECC, especially in +controllers with HDD support. The ECC can detect and correct errors. However, +it is unclear what ECC algorithm was used by each vendor. ECC is disabled by +default, but available for checking and correcting using +`--decoder.micropolis.ecc_type=TYPE` and for writing from IMG files using +`--encoder.micropolis.ecc_type=TYPE`, where TYPE is one of: + +| ECC | Description | +|--------|------------------------------------------| +| NONE | No ECC processing enabled | +| VECTOR | Vector Graphic Dual-Mode Disk Controller | + +The [CP/M BIOS](https://www.seasip.info/Cpm/bios.html) defined SELDSK, SETTRK, +and SETSEC, but no function to select the head/side. Double-sided floppies +could be represented as having either twice the number of sectors, for CHS, or +twice the number of tracks, HCS; the second side's tracks in opposite order +logically followed the first side (e.g., tracks 77-153). Micropolis disks +tended to be the latter. FluxEngine always emits CHS format disks, so you may +need to apply extra options to change the format if desired. + +## Options + + - : + - `143`: 143kB 5.25" SSDD hard-sectored; Micropolis MetaFloppy Mod I + - `287`: 287kB 5.25" DSDD hard-sectored; Micropolis MetaFloppy Mod I + - `315`: 315kB 5.25" SSDD hard-sectored; Micropolis MetaFloppy Mod II + - `630`: 630kB 5.25" DSDD hard-sectored; Micropolis MetaFloppy Mod II + - `vgi`: Read/write VGI format images with 275 bytes per sector + +## Examples + +To read: + + - `fluxengine read micropolis -s drive:0 -o micropolis.img` + +To write: + + - `fluxengine write micropolis -d drive:0 -i micropolis.img` + +## References + + - [Micropolis 1040/1050 S-100 Floppy Disk Subsystems User's Manual][micropolis1040/1050]. + Section 6, pages 261-266. Documents pre-ECC sector format. Note that the + entire record, starting at the sync byte, is controlled by software + + - [Vector Graphic Dual-Mode Disk Controller Board Engineering Documentation][vectordualmode]. + Section 1.6.2. Documents ECC sector format + + - [AltairZ80 Simulator Usage Manual][altairz80]. Section 10.6. Documents ECC + sector format and VGI file format + +[micropolis1040/1050]: http://www.bitsavers.org/pdf/micropolis/metafloppy/1084-01_1040_1050_Users_Manual_Apr79.pdf +[vectordualmode]: http://bitsavers.org/pdf/vectorGraphic/hardware/7200-1200-02-1_Dual-Mode_Disk_Controller_Board_Engineering_Documentation_Feb81.pdf +[altairz80]: http://www.bitsavers.org/simh.trailing-edge.com_201206/pdf/altairz80_doc.pdf + diff --git a/doc/disk-ms2000.md b/doc/disk-ms2000.md index f5ce0ddb..71f144e9 100644 --- a/doc/disk-ms2000.md +++ b/doc/disk-ms2000.md @@ -1,2 +1,33 @@ --d -\r +ms2000 +==== +## MS2000 Microdisk Development System + + +The RCA MicroDisk Development System MS2000 is a highly obscure (i.e. I gather +that single digit numbers of original machines exist) development system for the +RCA1802 series of CPUs, as made famous by the Cosmac ELF. It was a fairly +straightforward big bag o'RAM system with a 2kB boot ROM, 62kB of RAM, twin +floppy drives and a serial terminal --- CP/M users will find it very familiar. + +Read and writing disks is currently not supported by FluxEngine, but there is +basic support for the MicroDisk operating system's file system. This should +allow files to be read from MS2000 disk images. + +The disks are normal DD 3.5" disks, using a 70-track, single sided variation of +the venerable IBM floppy disk scheme, so allowing 315kB of storage per disk. + +If you have access to flux files for MS2000 disks, please [get in +touch](https://github.com/davidgiven/cpm65/issues/new) --- I would like to add +better support for these. + +## Options + +(no options) + +## Examples + +## References + + - [The EMMA-02 emulator](https://www.emma02.hobby-site.com/ms2000.html), which + supports the MS2000 and provides information on it. + diff --git a/doc/disk-mx.md b/doc/disk-mx.md index f5ce0ddb..0e3102e9 100644 --- a/doc/disk-mx.md +++ b/doc/disk-mx.md @@ -1,2 +1,69 @@ --d -\r +mx +==== +## Soviet-era PDP-11 clone + + +The DVK (in Russian, 沾7d65 +Computing Complex) was a late 1970s Soviet personal computer, a cut-down +version of the professional SM EVM (ⵁ241c +--- literally System of Mini Computers), which _itself_ was an unlicensed +clone of the PDP-11. The MX board was an early floppy drive controller board +for it. + +
+A DVK computer +
+ +The MX format is interesting in that it has to be read a track at a time. The +format contains the usual ID prologue at the beginning of the track, then +eleven data blocks and checksums, then the epilogue, then it stops. The +actual encoding is normal FM. There were four different disk variants, in all +combinations of single- and double-sided and 40- and 80-tracked; but every +track contained eleven 256-byte sectors. + +The format varies subtly depending on whether you're using the 'new' driver +or the 'old' driver. FluxEngine should read both. + +A track is: + + * 8 x 0x0000 words (FM encoded as 01010101...) + * 1 x 0x00F3 --- start of track + * 1 x 0xnnnn --- track number + * 11 of: + * 128 words (256 bytes) of data + * 16 bit checksum + * **if 'new' format:** + * 3 x 0x83nn --- `n = (track_number<<1) + side_number` + * **if 'old' format:** + * 3 x 0x8301 + +The checksum is just the unsigned integer sum of all the words in the sector. +Words are all stored little-endian. + +## Options + + - Format variants: + - `110`: 110kB 5.25" 40-track SSSD + - `220ds`: 220kB 5.25" 40-track DSSD + - `220ss`: 220kB 5.25" 80-track SSSD + - `440`: 440kB 5.25" 80-track DSSD + +## Examples + +To read: + + - `fluxengine read mx --110 -s drive:0 -o mx.img` + - `fluxengine read mx --220ds -s drive:0 -o mx.img` + - `fluxengine read mx --220ss -s drive:0 -o mx.img` + - `fluxengine read mx --440 -s drive:0 -o mx.img` + +## References + + - [The Soviet Digital Electronics + Museum](http://www.leningrad.su/museum/main.php) (source of the image + above) + + - [a random post on the HxC2001 support + forum](http://torlus.com/floppy/forum/viewtopic.php?t=1384) with lots of + information on the format + diff --git a/doc/disk-n88basic.md b/doc/disk-n88basic.md index f5ce0ddb..749b256c 100644 --- a/doc/disk-n88basic.md +++ b/doc/disk-n88basic.md @@ -1,2 +1,26 @@ --d -\r +n88basic +==== +## PC8800/PC98 5.25" 77-track 26-sector DSHD + + +The N88-BASIC disk format is the one used by the operating system of the same +name for the Japanese PC8800 and PC98 computers. It is another IBM scheme +variant, and is very similar to some mixed-format CP/M disk formats, where +track 0 side 0 uses 128-byte single density sectors and the rest of the disk +uses 512-byte double density sectors. (The reason for this is that the PC8800 +boot ROM could only read single density data.) + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read n88basic -s drive:0 -o n88basic.img` + +To write: + + - `fluxengine write n88basic -d drive:0 -i n88basic.img` + diff --git a/doc/disk-northstar.md b/doc/disk-northstar.md index f5ce0ddb..26386574 100644 --- a/doc/disk-northstar.md +++ b/doc/disk-northstar.md @@ -1,2 +1,50 @@ --d -\r +northstar +==== +## 5.25" hard sectored + + +Northstar Floppy disks use 10-sector hard sectored disks with either FM or MFM +encoding. They may be single- or double-sided. Each of the 10 sectors contains +256 (FM) or 512 (MFM) bytes of data. The disk has 35 cylinders, with tracks 0- +34 on side 0, and tracks 35-69 on side 1. Tracks on side 1 are numbered "back- +wards" in that track 35 corresponds to cylinder 34, side 1, and track 69 +corresponds to cylinder 0, side 1. + +The Northstar sector format does not include any head positioning information. +As such, reads from Northstar floppies need to by synchronized with the index +pulse, in order to properly identify the sector being read. This is handled +automatically by FluxEngine. + +Due to the nature of the track ordering on side 1, an .nsi image reader and +writer are provided for double-sided disks. The .nsi image writer supports +both single- and double-sided disks; however single-sided .nsi images are +equivalent to .img images. + +## Options + + - Format variants: + - `87`: 87.5kB 5.25" 35-track SSSD hard-sectored + - `175`: 175kB 5.25" 40-track SSDD hard-sectored + - `350`: 350kB 5.25" 40-track DSDD hard-sectored + +## Examples + +To read: + + - `fluxengine read northstar --87 -s drive:0 -o northstar.nsi` + - `fluxengine read northstar --175 -s drive:0 -o northstar.nsi` + - `fluxengine read northstar --350 -s drive:0 -o northstar.nsi` + +To write: + + - `fluxengine write northstar --87 -d drive:0 -i northstar.nsi` + - `fluxengine write northstar --175 -d drive:0 -i northstar.nsi` + - `fluxengine write northstar --350 -d drive:0 -i northstar.nsi` + +## References + + - [MICRO-DISK SYSTEM MDS-A-D DOUBLE DENSITY Manual][northstar_mds]. + Page 33 documents sector format for single- and double-density. + +[northstar_mds]: http://bitsavers.org/pdf/northstar/boards/Northstar_MDS-A-D_1978.pdf + diff --git a/doc/disk-psos.md b/doc/disk-psos.md index f5ce0ddb..27df0224 100644 --- a/doc/disk-psos.md +++ b/doc/disk-psos.md @@ -1,2 +1,32 @@ --d -\r +psos +==== +## 800kB DSDD with PHILE + + +pSOS was an influential real-time operating system from the 1980s, used mainly +on 68000-based machines, lasting up until about 2000 when it was bought (and +cancelled) by Wind River. It had its own floppy disk format and file system, +both of which are partially supported here. + +The PHILE file system is almost completely undocumented and so many of the data +structures have had to be reverse engineered and are not well known. Please +[get in touch](https://github.com/davidgiven/fluxengine/issues/new) if you know +anything about it. + +The floppy disk format itself is an IBM scheme variation with 1024-byte sectors +and, oddly, swapped sides. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read psos -s drive:0 -o pme.img` + +To write: + + - `fluxengine write psos -d drive:0 -i pme.img` + diff --git a/doc/disk-rolandd20.md b/doc/disk-rolandd20.md index f5ce0ddb..92eaa1dc 100644 --- a/doc/disk-rolandd20.md +++ b/doc/disk-rolandd20.md @@ -1,2 +1,48 @@ --d -\r +rolandd20 +==== +## 3.5" electronic synthesiser disks + + +The Roland D20 is a classic electronic synthesiser with a built-in floppy +drive, used for saving MIDI sequences and samples. + +Weirdly, it seems to use precisely the same format as the Brother word +processors: a thoroughly non-IBM-compatible custom GCR system. + +FluxEngine supports both reading and writing D20 disks, as well as basic support +for the filesystem, allowing files to be read from and written to D20 disks. +Note that the D20 was never intended to support arbitrary files on its disks and +is very likely to crash if you put unexpected files on a disk. In addition, +while the file format itself is currently unknown, there is a header at the top +of the file containing what appears to be the name shown in the D20 file +browser, so the name by which you see it is not necessarily the filename. + +A word of warning --- just like the Brother word processors, the D20 floppy +drive isn't very well aligned. The drive itself uses quarter-stepping to +automatically adapt to whatever alignment the disk was formatted with. This +means that trying to read such a disk on a PC drive, which does _not_ have +adjustable alignment, may not work very well. In these situations it is possible +to adjust the alignment of most modern drives, but this is a somewhat risky +process and may result in permanently wrecking the drive alignment. + +Please [get in touch](https://github.com/davidgiven/fluxengine/issues/new) if +you know anything about it. + +Many thanks to trondl [on the VCF +forums](https://forum.vcfed.org/index.php?threads/roland-d-20-decoding-the-mysterious-floppy-format.1243226/) +for assistance with this! + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read rolandd20 -s drive:0 -o rolandd20.img` + +To write: + + - `fluxengine write rolandd20 -d drive:0 -i rolandd20.img` + diff --git a/doc/disk-rx50.md b/doc/disk-rx50.md index f5ce0ddb..d39a5ca2 100644 --- a/doc/disk-rx50.md +++ b/doc/disk-rx50.md @@ -1,2 +1,23 @@ --d -\r +rx50 +==== +## 400kB 5.25" 80-track 10-sector SSDD + + +The Digital RX50 is one of the external floppy drive units used by Digital's +range of computers, especially the DEC Rainbow microcomputer. It is a fairly +vanilla single-sided IBM scheme variation. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read rx50 -s drive:0 -o rx50.img` + +To write: + + - `fluxengine write rx50 -d drive:0 -i rx50.img` + diff --git a/doc/disk-shugart_drive.md b/doc/disk-shugart_drive.md index f5ce0ddb..086e1d39 100644 --- a/doc/disk-shugart_drive.md +++ b/doc/disk-shugart_drive.md @@ -1,2 +1,15 @@ --d -\r +shugart_drive +==== +## Adjust configuration for a Shugart drive + + +This is an extension profile; adding this to the command line will configure +FluxEngine to adjust the pinout to work with a Shugart drive. This only works +on Greaseweazle hardware. + +For example: + +``` +fluxengine read ibm --720 shugart_drive +``` + diff --git a/doc/disk-smaky6.md b/doc/disk-smaky6.md index f5ce0ddb..03dec9f8 100644 --- a/doc/disk-smaky6.md +++ b/doc/disk-smaky6.md @@ -1,2 +1,34 @@ --d -\r +smaky6 +==== +## 308kB 5.25" 77-track 16-sector SSDD, hard sectored + + +The Smaky 6 is a Swiss computer from 1978 produced by Epsitec. It's based +around a Z80 processor and has one or two Micropolis 5.25" drives which use +16-sector hard sectored disks. The disk format is single-sided with 77 tracks +and 256-byte sectors, resulting in 308kB disks. It uses MFM with a custom +sector record scheme. It was later superceded by a 68000-based Smaky which used +different disks. + +FluxEngine supports these, although because the Micropolis drives use a 100tpi +track pitch, you can't read Smaky 6 disks with a normal PC 96tpi drive. You +will have to find a 100tpi drive from somewhere (they're rare). + +There is experimental read-only support for the Smaky 6 filesystem, allowing +the directory to be listed and files read from disks. It's not known whether +this is completely correct, so don't trust it! + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read smaky6 -s drive:0 -o smaky6.img` + +## References + + - [Smaky Info, 1978-2002 (in French)](https://www.smaky.ch/theme.php?id=sminfo) + diff --git a/doc/disk-tartu.md b/doc/disk-tartu.md index f5ce0ddb..4831d4db 100644 --- a/doc/disk-tartu.md +++ b/doc/disk-tartu.md @@ -1,2 +1,48 @@ --d -\r +tartu +==== +## The Palivere and variations + + +The Tartu Palivere is a 1988 Z80-based computer from Estonia. It is a CP/M +machine with 64kB of RAM, running off a 2MHz ꃣ0e30 +clone; it operated off punched tape, cassette, external hard drive or floppy, and was notable as being the first ever computer with an Estonian keyboard. + +
+The Tartu computer's developer Leo Humal working with one. +
+ +From a floppy disk perspective, it is interesting because the floppy drive +interface is almost entirely handled in software --- necessary at the time as +the usual floppy disk interface chip at the time, the ⎲fba5 +of the WD1793), was hard to find. Instead, the floppy controller board was +implemented entirely using TTL logic. Despite this, the encoding is fairly high +density, using MFM and with up to 780kB on a double-sided 80 track disk. + +
+The Tartu FDC with Soviet TTL logic chips. +
+ +FluxEngine supports reading and writing Tartu disks with CP/M filesystem access. + +## Options + + - Format variants: + - `390`: 390kB 5.25" 40-track DSDD + - `780`: 780kB 5.25" 80-track DSDD + +## Examples + +To read: + + - `fluxengine read tartu --390 -s drive:0 -o tartu.img` + - `fluxengine read tartu --780 -s drive:0 -o tartu.img` + +To write: + + - `fluxengine write tartu --390 -d drive:0 -i tartu.img` + - `fluxengine write tartu --780 -d drive:0 -i tartu.img` + +## References + + - [The Estonia Museum of Electronics](https://www.elektroonikamuuseum.ee/tartu_arvuti_lugu.html) + diff --git a/doc/disk-tids990.md b/doc/disk-tids990.md index f5ce0ddb..e25dea48 100644 --- a/doc/disk-tids990.md +++ b/doc/disk-tids990.md @@ -1,2 +1,39 @@ --d -\r +tids990 +==== +## 1126kB 8" DSSD + + +The Texas Instruments DS990 was a multiuser modular computing system from 1998, +based around the TMS-9900 processor (as used by the TI-99). It had an 8" floppy +drive module, the FD1000, which was a 77-track, 288-byte sector FM/MFM system +with 26 sectors per track. The encoding scheme was very similar to a simplified +version of the IBM scheme, but of course not compatible. A double-sided disk +would store a very satisfactory 1126kB of data; here's one at old-computers.com: + +
+ +A DS990 at old-computers.com +
+ +FluxEngine will read and write these (but only the DSDD MFM variant). + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read tids990 -s drive:0 -o tids990.img` + +To write: + + - `fluxengine write tids990 -d drive:0 -i tids990.img` + +## References + + - [The FD1000 Depot Maintenance + Manual](http://www.bitsavers.org/pdf/ti/990/disk/2261885-9701_FD1000depotVo1_Jan81.pdf) + diff --git a/doc/disk-tiki.md b/doc/disk-tiki.md index f5ce0ddb..bcf4cf89 100644 --- a/doc/disk-tiki.md +++ b/doc/disk-tiki.md @@ -1,2 +1,27 @@ --d -\r +tiki +==== +## CP/M + + +The Tiki 100 is a Z80-based Norwegian microcomputer from the mid 1980s intended +for eductional use. It mostly ran an unbranded CP/M clone, and uses fairly +normal CP/M disks --- IBM scheme and from 128 to 512 bytes per sector depending +on the precise format. + +## Options + + - Format variants: + - `90`: 90kB 40-track 18-sector SSSD + - `200`: 200kB 40-track 10-sector SSDD + - `400`: 400kB 40-track 10-sector DSDD + - `800`: 800kB 80-track 10-sector DSDD + +## Examples + +To read: + + - `fluxengine read tiki --90 -s drive:0 -o tiki.img` + - `fluxengine read tiki --200 -s drive:0 -o tiki.img` + - `fluxengine read tiki --400 -s drive:0 -o tiki.img` + - `fluxengine read tiki --800 -s drive:0 -o tiki.img` + diff --git a/doc/disk-victor9k.md b/doc/disk-victor9k.md index f5ce0ddb..82e69814 100644 --- a/doc/disk-victor9k.md +++ b/doc/disk-victor9k.md @@ -1,2 +1,62 @@ --d -\r +victor9k +==== +## 1224kB 5.25" DSDD GCR + + +The Victor 9000 / Sirius One was a rather strange old 8086-based machine +which used a disk format very reminiscent of the Commodore format; not a +coincidence, as Chuck Peddle designed them both. They're 80-track, 512-byte +sector GCR disks, with a variable-speed drive and a varying number of sectors +per track --- from 19 to 12. Disks can be double-sided, meaning that they can +store 1224kB per disk, which was almost unheard of back then. Because the way +that the tracks on head 1 are offset from head 0 (this happens with all disks), +the speed zone allocation on head 1 differs from head 0... + +| Zone | Head 0 tracks | Head 1 tracks | Sectors | Original period (ms) | +|:----:|:-------------:|:-------------:|:-------:|:--------------------:| +| 0 | 0-3 | | 19 | 237.9 | +| 1 | 4-15 | 0-7 | 18 | 224.5 | +| 2 | 16-26 | 8-18 | 17 | 212.2 | +| 3 | 27-37 | 19-29 | 16 | 199.9 | +| 4 | 38-47\* | 30-39\* | 15 | 187.6 | +| 5 | 48-59 | 40-51 | 14 | 175.3 | +| 6 | 60-70 | 52-62 | 13 | 163.0 | +| 7 | 71-79 | 63-74 | 12 | 149.6 | +| 8 | | 75-79 | 11 | 144.0 | + +(The Original Period column is the original rotation rate. When used in +FluxEngine, the disk always spins at 360 rpm, which corresponds to a rotational +period of 166 ms.) + +\*The Victor 9000 Hardware Reference Manual has a bug in the documentation +and lists Zone 4 as ending with track 48 on head 0 and track 40 on head 1. +The above table matches observed data on various disks and the assembly +code in the boot loader, which ends Zone 4 with track 47 on head 0 +and track 39 on Head 1. + +FluxEngine can read and write both the single-sided and double-sided variants. + +## Options + + - Format variants: + - `612`: 612kB 80-track DSHD GCR + - `1224`: 1224kB 80-track DSHD GCR + +## Examples + +To read: + + - `fluxengine read victor9k --612 -s drive:0 -o victor9k.img` + - `fluxengine read victor9k --1224 -s drive:0 -o victor9k.img` + +To write: + + - `fluxengine write victor9k --612 -d drive:0 -i victor9k.img` + - `fluxengine write victor9k --1224 -d drive:0 -i victor9k.img` + +## References + + - [The Victor 9000 technical reference manual](http://bitsavers.org/pdf/victor/victor9000/Victor9000TechRef_Jun82.pdf) + + - [DiskFerret's Victor 9000 format guide](https://discferret.com/wiki/Victor_9000_format) + diff --git a/doc/disk-zilogmcz.md b/doc/disk-zilogmcz.md index f5ce0ddb..7d2cb218 100644 --- a/doc/disk-zilogmcz.md +++ b/doc/disk-zilogmcz.md @@ -1,2 +1,42 @@ --d -\r +zilogmcz +==== +## 320kB 8" 77-track SSSD hard-sectored + + +The Zilog MCZ is an extremely early Z80 development system, produced by +Zilog, which came out in 1976. It used twin 8-inch hard sectored floppy +drives; here's one at the Centre +for Computing History: + +
+ +A Zilog MCZ at the Centre For Computing History +
+ +The MCZ ran Zilog's own operating system, Z80-RIO, and used 77 track +single-sided disks, with 32 sectors (each marked by an index hole), with 132 +bytes per sector --- 128 bytes of user payload plus two two-byte metadata +words used to construct linked lists of sectors for storing files. These +stored 320kB each. + +FluxEngine has read support for these, including support for RIO's ZDOS file +system. + +## Options + +(no options) + +## Examples + +To read: + + - `fluxengine read zilogmcz -s drive:0 -o zilogmcz.img` + +## References + + * [About the Zilog MCZ](http://www.retrotechnology.com/restore/zilog.html), + containing lots of useful links + + * [The hardware user's manual](https://amaus.org/static/S100/zilog/ZDS/Zilog%20ZDS%201-25%20Hardware%20Users%20Manual.pdf) +