Home Page Forums MAME MAME Requirements?

We can't post ROMs here, but that particular ROM has been available from fine illegal ROM sites for the last 3 months. Google is your friend.

Wow, that brings back some memories. I used to use these at work in the mid-80's.

The 2013 version of my Rainbow disassembly (HTML + IDA 5 freeware source included).

https://dl.dropbox.com/u/37819653/Jan-2013_DISASSEMBLY.zip

Documented & located a bunch of internal routines, for example

$1e15 - ROM_ConsoleLev2In
$1e01 - ROM_ConsoleLev2Stat
$1e24 - ROM_ConsoleLev1In
$001e - ROM_DisableCursor
$0021 - ROM_EnableCursor
$1e49 - ROM_InitVectors
$1dfa - ROM_ReturnClock
$1ebf - ROM_SetLeds (keyboard LED positions documented)
$1ec3 - ROM_ClearLeds
$1ecc - ROM_FastVideo (with parameters!)
$1f09 - ROM_InitToNVM
$1f0c - ROM_RawKbdData
$1f23 - ROM_RomVersion
$1f66 - ROM_ChangeVectorMap
$1fc2 - ROM_RingBell
$1fc8 - ROM_GetSet7or8

(just search for :1DBA in the HTML output)

Rainbow-100 B low rom disassembly updated
Here ist the June 2013 edition of my low ROM disassembly (documented IDA 5 Freeware source and HTML).

https://dl.dropboxusercontent.com/u/37819653/June-2013_DISASSEMBLY.zip

Managed to find out more on how DC011 and DC012 display controllers are initialized (search for DC011, DC012 and INIT_VIDEO_HW). NVM handling is better documented (-> keyword NVM in HTML).

Also, it is now clear how option boards (the RAM board for 128 - 896K, as well as the communication board, which has a DMA and a non-DMA mode) are detected: via interrupts (-> keywords: OPTION or ENABLE_RAM_BOARD_DETECTION)

It's a pity that high intensity and underline, blink, reverse, as well as 80 / 132 column mode (with different dot pitches) aren't emulated yet. VT100 emulation seems a clever decision, but i can't imagine how the more exotic modes of the real hardware could be emulated (BIOS unsupported 48 line modes like http://rainbow-100.com/archives/154/ OR : SQUINT from Latrobe http://www.classiccmp.org/rainbow/files/goodnight/LATROBE/s/squint.arc - will provide a photo later)

BTW, the pc98gdc.c / PC98 MESS driver seems to implement the NEC 7220 display controller used in the color graphics extension (required for the early Windows 1 DEC released).

-----------------------------------------------------------
I was asked to document the 4 test jumpers present on the 100-B motherboard.

W13, W14 and W15 manufacturing test jumpers are represented by bits 1,2,3 of Diagnostic Read Register $0a

Jumper W18 pulls DSR line LOW when set (bit 7 of Status Read Register $11). It is possibly the 'unknown' jumper near the Z80 (requires verification).

Example from the low ROM disassembly:

seg000:0DEF test al, 80h ; W18 jumper / DSR (bit 7)
seg000:0DF1 jnz short loc_DE1


I wish someone could add jumpers and accurate NEC V20 (80186+) emulation - the latter was one of the more affordable upgrades on the 100-B. Note that the original V20 patch published on Jeff's 'DRIVE W' introduces serious timing issues.

My real hardware quickly got a worn out EPROM socket when i tried these - and other - hacks

Speaking of Rainbow, I dumped the two PROMS and two PALS from one of the Rainbow's video boards (specifically, board "5015687 01" with the upd7220D-1 on it) rfka01 lent me.

These will be added to the driver shortly.

P.S. if you wore out the eprom socket, grab a soldering iron, a soldapult and some flux, and replace them; I suggest using dual wipe sockets from digikey, and to prevent this happening in the future, stack an extra socket on top of each one in the board, and use that extra socket when inserting/removing chips. (hence, if a socket fails, you can pull it out of the other socket, no desoldering needed)

LN

Hi there,

i am currently experimenting with semi graphics on the Rainbow-100. A modified character ROM and a simple QBASIC program (to read BMPs) are sufficient to achieve:

264 x 72 (264 x 96) resolution in 132 x 24 mode
140 x 144 (140 x 192) resolution in 70 x 48 mode (custom mode of DC011/DC012)

Resolutions in brackets use a 2x4 character set instead of 2x3.

As you can see in the image, you can even mix text and graphics - normally impossible on the Rainbow

https://dl.dropboxusercontent.com/u/37819653/Rainbow-100%20Semi%20graphics.jpg

I know it is an old trick, but the technique (and the source) could be adapted to other systems with more horsepower and screen RAM.

---- Back to our main topic: the current VT100 (or Rainbow?) driver does not allow reverse character attributes ($ef000 attribute RAM). Is this a bug?
Characters above $80 display in reverse, though.

To be more precise, even the error messages (shown when deliberately booting the 'non working' driver) have active reverse, bold, and blink attributes - if i remember correctly

The character space needed could be reduced easily if not only line, but character attributes also worked.

Any hints, suggestions or ideas?

It's quite possible attribute RAM isn't fully supported; L_N would know more about that since it's borrowed from the VT100.

Yeah the video stuff in vt100 is missing a number of graphics features (evident on the setup screens as well); I didn't write that code though, I think that was Micko or Judge.

LN

I looked into vtvideo.c, but don't want to interfere with the VT100 code (no experience and no way to test on a live machine).

That said, i recently added 3 of the 5 character attributes necessary (reverse, blink, underline) - only for RAINBOW_VIDEO.

Bold (intensity) and soft scrolling are not as easy to implement.

Additional to the 48 line mode (now also implemented & not present on the VT-100) there are possibly some undocumented DC012 pokes - see below.

Any info would be appreciated.

https://dl.dropboxusercontent.com/u/37819653/SQUINT_Disassembly__POKES.jpg
(excerpt from fun program, compresses image to a quarter of the normal screen size on a real DEC-100 B)

---------
Attached, you see a screenshot of a VR201 (orange-black) monitor connected to the Rainbow-100.

The combination of different attributes (bold + reverse characters on reverse base video) is a bit tricky to emulate correctly.

https://dl.dropboxusercontent.com/u/37819653/rainbow-1024x768.jpg

Question: what's the proper way to configure the palette to emulate different VR201 phosphor types (orange, white, ...)?

This is the doc for the WD2793 Floppy disk controller used in the Siemens PC-D

https://dl.dropboxusercontent.com/u/55419307/WD2793%20datasheet.pdf

It would be great if the PC-D could benefit from the recent changes to the 80186 driver and be broken out from PC compatibles into a seperate driver (it uses no standard PC bus and can not boot from generic MS-DOS floppies).

Robert