Search results

I don't think they cloned everything that IBM made, just the most useful models. There probably were't enough installs to warrant cloning the 5100, unlike, say, Wang 2200.

6580 emulation is still broken, needs more reverse engineering (or a copy of schematics, if these ever were published by IBM). It's possible that Displaywriter System Manual S544-2023-0 (?) -- mentioned in US patents 4648071 and 5675827 -- has them, but bitsavers does not have a copy.

Interrupt list says this call was added for AT/EGA, and Poisk was neither of those. There's not much point in using RK or another tool like it -- Poisk has cyrillic font in ROM and BIOS does keyboard layout switching.

This game was inspired by the same book (Roadside Picnic) as the S.T.A.L.K.E.R. series. Runs under RT-11. Gameplay is appropriately deadly and user-unfriendly (only 6 items in the inventory, no identify, all scrolls, potions and wands are random, and so on).

I am working on emulation of GRiD Compass in MAME (see [2]); these machines used bubble memory controlled by Intel D7220 chipset. Since MAME also emulates CoCo, I thought to add emulation of Dennis Kitz's bubble memory expansion board (which used same chipset). July'1983 CoCo Magazine [1]...

Keyboard can be made to talk USB via controller replacement -- http://downloads.cornall.co/ibm-capsense-usb-web/ibm-capsense-usb.html My reverse-engineering/emulation effort is on hold -- the emulator (MAME) can boot some disks (diags, CP/M) but not the actual word processor. Of course, anyone...

Textpack still does not work in the emulator -- the crash message I've posted earlier means "insufficient memory", even when 320KB is configured. Memory size is actually probed correctly, but is then reset to 128KB -- in the "new" BIOS dumped by Nigel, this happens at FD45Eh. This code...

"Asynchronous Communication Facility" described in http://www.computerhistory.org/collections/catalog/102678942? It's a terminal emulator, apparently. There was also a "Displaywriter/Personal Computer Attach Convenience Kit":

Kermit under CP/M could work -- but "current release of CP/M-86 for the Displaywriter does not support these [serial communications] interfaces" (from http://www.nostalgia8.nl/cpm/ibm/cpm6dwrm.pdf) :(

What's the part number on it? (also, just noticed the cat :))

My code was accepted into next MAME release (0.182 -- should happen later this month). Teledisk and Imagedisk floppy images (single-density ones) are usable with MAME; of those on Bitsavers, C.E. diags and CP/M boot but aren't stable. POST (aka BAT) errors can be skipped with Memory Record...

Getting there... bit 6 in floppy status port (8150) apparently reflects 'idle' status of MCU and upd765 both. Diagnostics disks are now usable and CCP/M boots. Textpack does not yet.

... but exposure to this much IBM over-engineering may leave permanent scarring :) Seriously though, at this point a logic analyzer could help a lot to improve high-level emulation of floppy board.

upd765 chip is actually available on system bus (ports 8158, 815A) and with a few hacks I forced firmware to IPL; there are still a lot of unknowns and nothing completely boots yet. This is Textpack A crash message:

Thanks to Al Kossow (who dumped the remaining ROMs) and nigwil (hands-on tests -- I don't have the hardware), here's some progress (key tests are still failing -- this screenshot required a debugger hack to ignore these failures). EDIT: this is how semi-graphics mode looks like (looks close...

Some diagnostics are passing (and others I've simply skipped via debugger hack :) ) and... (It's saying that keyboard test is failing.) Character generator ROM is not dumped (and it's in EBCDIC), I've used a generic PC font + translation table. Actual font looks like this:

IBM's patents refer to a "Displaywriter System Manual" which should have everything we need :) Firmware accesses these port ranges. I've identified some of them: AM_RANGE(0x0000, 0x0007) AM_DEVREADWRITE8("pic8259", pic8259_device, read, write, 0x00ff) AM_RANGE(0x0008, 0x000f)...

pdf/ibm/6580_Displaywriter/S241-6248-3_Displaywriter_Product_Support_Manual_Feb83.pdf on bitsavers is quite useful -- Chapter 6 describes the POST sequence (called BAT, Basic Assurance Test) and interrupt level assignments. Floppy cable is simply a bus extension, so the floppy controller sits...

Maybe there's a simpler way -- does maintenance manual explain how the devices are mapped in memory? + there's some info in CP/M-86 reference manual for DW (http://www.nostalgia8.nl/cpm/ibm/cpm6dwrm.pdf) -- f.e. that Control-End causes a power-on-reset, brief description of video memory layout...

I am collecting info on this system, to improve its emulation in MAME from current 'skeleton' level and found retroComputingTasmania's info page -- thanks for that. Code in images from bitsavers (8493023.BIN and 8493822.BIN) doesn't look sensible, though -- f.e. here's the startup sequence...