z88dk forums

+ Pickles & Trout CP/M.
So far we have can create disk images in 2 different CP/M formats for the TRS-80 Model II.
Pickles & Trout supported a Real Time Clock, but our current time library does not interface to it.

Omikron CP/M on Model I.

Double checking the skew table, this time I first had to guess it since I wasn't able to kick in DPB.COM.

I noticed that all the special disk images I could test are perfectly readable also by replacing the special boot track with a regular one. No format fingerprints or detection tricks seem to depend on it, so it is pointless to add complexity to appmake. If it will be necessary, it is probably more u...

Yet another interesting trick to forcefully run a CP/M executable in a TRS80 emulator in CP/M mode. Most of the emulators allow direct load of CMD files, thus we can use z88dk_appmake twice. In a first run we build the CP/M command file, then we add the TRS80 CMD header: zcc +cpm -create-app dpb.c z...

The resulting code is not a wonder but still acceptable. A couple of examples of those crazy formats: Often the first sector is in a basic format, to allow boot. The TRS80 puts the boot code in a single track and requires a lot of sectors to fit ! IMageDisk Utility 1.18 / Mar 07 2012 IMD TD 1.4 8&qu...

I noticed that, especially in 8" disk formats a track0 with different sectors is not so rare. I was thinking to extend appmake and force the IMD format to behave differently, extending the current structure in this way: uint8_t trk0_sectors_per_track; uint16_t trk0_sector_size; uint8_t trk0_gap...

Congratulations! I think we all hoped to see the whole IOT on that CPU when it was launched

Same technique, used multidsk on Ampro (on MAME) to tune and check the disk image configuration, with the support of z88dk/examples/cpm/dpb.c, and the Dunfield's TD02IMD and IMDU. Final check on trs80gp running Montezuma CP/M. The graphics library won't work, it surely requires page swapping. Sound ...

diskdefs in the cpmtools provides a lot of useful informations but most of those definitions were verified only as "logically valid", in example there's no difference in terms of file block allocation if we work on a double sided disk with 80 tracks per side or on a single sided disk with ...

I finished the options avaliable on the Zorba and discovered the Ampro ! This one included compatibility tools to configure a custom floppy disk format (always mapped to e:), format some of them, etc. I already pulled in a good number of new formats. Among the others, the VT180 "Robin", th...

Osborne 1 formats improved by addind the single density format and the correct disk speed and interleave settings.
Now the disk images work correctly also in IMD mode.

VT-180 disk mode pulled.
Having a good emulation level helps in creating accurate disk images, with transfer speed and skew settings closer to the real ones, raising the probability to get working suff on the real hardware.

that's not impossible, I'm confident that the emulation work better on the v1050 than on the mbc200
I simply couldn't figure out how to make it work

In this way I could verify that the disk format for the Televideo TS-803 and TPC1 is the same of the TS-802.
I'm renaming it in appmake and I'm adding a generic "televideo" subtype.

The Telcon Zorba portable CP/M computer had a lot of disk options, I'm using it to check the formats we've already implemented and possibly find the extra ones which could be nice to have. Well, I'm already disappointed, I can't understand how the IMD format could keep the 1..20 sector numbering wor...

It shouldn't happen, I wondered whether I forgot to update something in the public z88dk tree, but they're 2 files only and they are ok. zcc +cpm -subtype=v1050 -create-app tic.c mame v1050 -flop1 1050-sys.td0 -flop2 a.imd I could run it on a recent MAME build (self-built from the current dev tree, ...

I'm going to pull in the full VT-ANSI engine, which is quite perverse, because it encapsulates the ANSI ESC sequences in an ANSI ESC interpreter ! v1050ansi.png Alternatively programmers can totally skip the ansi VT engine and use the native mode which will accept almost all the ANSI VT sequences, w...

In short, you're expecting sccz80 to force h to zero, right?

https://user-images.githubusercontent.com/10224992/240363429-c733bd81-a20e-4a07-8d8b-01003b0f0fb6.png Support of generic console with 8x8 font and udg is now ready (the code is a nightmare of ESC sequences!). I still couldn't find a way to catch the talkback from the video section, it probably requ...

I'm now looking for the way to get data from the video section (6502 controlled).

A handy way to read the V1050 sources is on this link:
http://v1050.classiccmp.org/src/index.html

Sometimes the transmission breaks. I could observe it also with the original GSX driver, so I'm not sure it does not depend on the emulation (I'm using an old version of MESS because I miss some of the ROM images). Passing the parameters as ASCII-decimal values works much better, but sometimes the p...

I finally had to mix BIOS calls and direct out to allow a stable communication.
A first minimal set of graphics is now working, I think I'll open a PR and proceed by steps.

This is another interesting hw, its concept is very close to the Sanyo MBC-200, but the terminal control and graphics were delegated to a 6502 CPU. There's a nice sw set (TOSEC) in archive.org which include the rather rare "DR Graph" and useful original system oriented source code. https:/...

I think it is in a consistent status, now. Considering that mbc_send() has its own advantages in terms of speed and control over BIOS or BDOS, I decided not to change it. It is rather convenient to replace it with a function calling the BIOS on the other Sanyo systems. I can't wait to move on and tr...

It's very comfortable indeed. I must have messed with the underlined text attribute, bit 7 in CAPS, which is tempting me to provide both the implementations (there's very little code in this case) . I should also change the calls to mbc_send in the terminal code (OUTing to the 2nd CPU directly) and ...