z88dk forums

I'll be editing this list to try tracking down the known CP/M formats for the Amstrad machines. The resulting configurations allow the file extraction from the RAW dumps of the disk images but are not enough to recreate the correct skew factor and sector numbering. Vortex CP/M . IMDU shows non-linea...

My "discovery" was well documented on the older Pro-DOS version. I think there are good possibilities to adapt the graphics library relying on the z88dk sections, it's probably way better than the trick I used on the Spectrum +3. @Dom, you can probably arrange it way more easily than me. T...

IIRC EightyOne runs on WINE

I *think* you get the final optimized code already. Sometimes those frequent PUSH / POP sequences give the feeling of absent optimization, you can try comparing the outputs with the normal build and a second one with "-O0", in example. Adding "-v" you will see the optimiser passe...

Dom, most of the Floating Point should be working since.. quite a few. The major problem was on the index registers but it is now partially addressed by intercepting and replacing the video interrupt handler also in text mode.

Also the ROM based FP implementation is now sightly improved/fixed.

The ProDOS, is just a good CP/M implementation for the SAM Coupe. After enough experiments to understand how the FD format should have been I discovered that it is 100% compatible with the Amstrad PCW 80T disks, its system disk has even the same heading on track#0 ! I'm inserting a basic configurati...

There are still a two or three CP/M formats for the Spectrum which I can't verify. They must wait. By the way one more target is in the supported formats lists. I noticed that the QUORUM emulator reads two disk formats, the already discovered Profi's KLUG CP/M but also also the HOBBIT CP/M disks ! A...

On the ZX81 you will miss everything explicitly as part of the so-called "newlib". Most of the remaining stuff *should* (could) work, I encourage you to experiment trying also examples in the subfolders (including zx81, graphics, console, clisp, but also something elsewhere, like in the Sp...

If you can limit your code to blind computation, then we have "ticks", it is part of the z88dk kit. It is a modified CP/M emulator benchmarking the code for you. Optimizing the code for speed usually involves assembly. I suppose that most of the optimization attempts will be something like...

And at last a bit of luck !! The "Cobra" disk images available here: http://cobrasov.com/CoBra%20Project/cpm_comments-ro.html#C1 Are perfectly read from the FUSE emulated HC-2000 environment. I'd assume the two 3,5 inch disk formats are compatible. I noticed that the boot sectors *in some ...

Pull request opened for the HC-91 and HC-2000.
The trap was in the OS disks! On the HC-2000 the 720K format is supported by drive A only, the 5,25 inch format by drive B only.
Moreover support for drive B works properly only on the "Z80CCP" version of the HC-2000 CP/M system (CPMZ80.DSK).

The HC-91 and HC-2000 disk definitions for CPMTOOLS (finally!) On the HC-91 there's surely a waste with the byte-sized sector allocation... perhaps the CP/M BIOS for the HC-91 was initially designed for 40 tracks disks. # ZX HC-91 diskdef hc2000 seclen 512 cylinders 57 sectrk 9 heads 2 blocksize 204...

More experiments on the Romanian clones. I finally discovered that there probably were two different disk formats. The HC-91 one had 40 tracks, and byte sized extents while the HC-2000 had 80 tracks, word sized extents. The "Z80CCP" version of the HC-2000 CP/M system seems to be able to re...

I havent worked on this one (yet), but I'm adding the link for tracking purposes. COBRA http://cobrasov.com/CoBra%20Project/cpm80.html HC-2000 https://sites.google.com/site/georgechirtoaca/projects http://www.forofpga.es/download/file.php?id=620 https://www.z80-romania.ro/Dischete_CPM_pentru_HC_91_s...

I woudn't exclude an sccz80 silent crash.
Can you successfully compile different programs ?
I'd suggest to try cutting away code sections and spot the odd condition, I'm sure Dom is interested in this.

I opened a last minute pull request for the Cortex disk format. This one was relative to the Visual 1050, I couldn't test it directly but the disk format was in ALIENC.COM As expected it is very similar to the Visual 1050 one, simply it was adjusted for 80 tracks, while the Cortex had 40 tracks disk...

I just put yet another CP/M disk format in appmake for an obscure Italian system: the "SAGA FOX". The main reason for putting a couple of curious formats in appmake is to show how flexible z88dk is. Moreover the SAGA computers require the same data inversion of the Sharp MZ-80 systems now ...

I think it is worth to mention, to make life easier in the eastern ZX clones world we could add a new disk container to "appmake +cpmdisk": the FDI format (same file extension of the Amiga specific files but Sinclair related). It is not immediate to implement though. A sector map must be p...

Well, I finally found a little bit of regression. The current fix is not very elegant, especially on the DSK format, but should work. For what I could understand, the disk dump utilities and the related formats allowed a bit of flexibility for permitting a last resort data gathering in case of troub...

Apparently I didn't break it :) I found that MAME unrolls the MZ80A/MZ80B IMD formats properly only if an initial skew table is provided. Unluckily enriching the file attributes indiscriminately would break other target images, probably some of the emulations are expecting the sectors shuffled OR th...

I forgot to mention two more disk tools very useful to convert and analyse the disk images.

- libdsk
- HxCFloppyEmulator

Well it works ! This crappy tool was of great help. On the MZ80B both the sector data and the side order must be inverted. Currenlty I got it working on the DSK format, I'm now checking this same configuration on all the other disk formats. SAMDISK should help when the format is not supported on the...

More about the MZ80B format. Clearly also the sideness is inverted ! The tool I used to analyse the disk format is probably showing me the data already "XORed", which helps the readability. A non-system disk shows the following sector/track sequence: mz80b_data_disk.png The three arrows sh...

Some of the lost CP/M boot images for the earlier Sharp MZ models were recently found. It allowed me to look at them looking for a way to deal with those odd disk formats. Few relevant tools are helping me in the intermediate steps: - The ALIENC.COM tool for the MicroBee claims to read the "Sha...

I was looking for totally different things and I noticed this one: https://archive.org/details/wristapp-timex-datalink-ironman-usb/ The Timex watches had an original way to download a software app, initially they had to be put in front of a computer monitor. A special program, tuned for the CRT disp...