z88dk forums

Interesting!
It is again a relevant quantity of functions. It's probably easier to first try replicating the exact functions by using the demonstration programs as test cases.
I haven't looked at the code yet but it should be feasible.

are you focusing on the spectrum? the Amstrad cpc is another interesting challenge

https://github.com/z88dk/z88dk/blob/a9b ... ns.asm#L29

I may have misunderstood your case, but isn't it the same also with the zx microdrive, opus discovery and others?

The FP calculator is so slow. Perhaps it was used in extended BASIC or compiler implementations, BASIC extensions, like extra commands, assembly written USR functions to fit into DEF FN declarations, etc. I think decoding the RST28 was often present in zx spectrum disassemblers, I remember "zoo...

> Page 102 of this document

wow, and I even knew it ! it's too long I left the TI calculator behind.

very interesting, thank you for pointing it out, I'll surely try to compare them!

cborn, what I wanted to say is that youwould hardly cover all the tricks the programmers used in their code. The Microsoft Basic, in example, uses an RST (where possible, on other targets it is a CALL) to do the syntax checking, by putting the required symbol just after the call. If the symbol doesn...

There's yet another answer I can give (after searching it online) : https://retrocomputing.stackexchange.com/questions/24899/why-does-cp-m-use-call-0005h-for-its-syscalls probably the CP/M BDOS entry was place at address 0005 to avoid conflicts with an EXISTING RST mechanism on computers already exi...

well, let me try to clarify a little bit. First of all, you should know that the RST calls (close relatives of the INT entries on the Intel x86 CPUs) can be considered simply as a trick to save memory, using an RST entry in place of the more used CALLs in your program. A Z80 often will have a ROM on...

On September 2020 Martin ..someone implemented the deflate algorithm for the z80 basing on a Hitech C source for the Amiga. This allowed an incredible improvement of the famous unzip tool for CP/M originally written in 1991: now it works with the modern zip archives too. https://github.com/agn453/UN...

we have finally news on this topic! EDIT: appmake is now creating TI83 Plus applications in the "right" format, a.k.a. the format expected by the recent tools and emulators, like Wabbitemu. I kept an option on appmake, "-Cz--oldfmt", to restore the ancient, probably wrong, way to...

Yes, I chose to stick the prefix to deal with the user number on top of the original BDS C parser to leave the rest of the code untouched.
I was so surprised that it was fitting with almost no changes that I was treating it like a magic item.

It probably depends on the asm implementation of parsefcb (in parsefcb.c).

Please try specifying a filename including the user number on top, it should be something like:

0/a:file.txt

Apologies for the mess, it looked like a brillant idea, long time ago..

On the ZXROM side, fmax() and fmin() are now fixed. I also added less common functions like cbrt(), hypoth(), exp2()...
The overall support is still poor but slightly better.

I updated the test program above.
Proceeding with the tests I discovered that we have an issue on asin() and acos() in genmath !
This time the crappy zxmath is doing well.

While interfacing the LAMBDA libraries to the the Sinclair BASIC, few bugs are emerging. So far I had to fix fmod() and log10(), I'm taking the opportunity to test it furtherly. I noticed that the game dallas.c is crashing, it could be my next test case. This very dumb piece of code is helping me. S...

Yes. Depending on the target system it could be not so straightforward, the problem is whether the position for ORG is flexible on that target or require a specific value (to be found in lib/targets/*/*crt0*. Run appmake +target -h and check if the "-zorg" option is available. Usually addi...

The library version depends on the z80asm version, which in turn is being massively improved. It shouldn't be a problem at all with the z88dk libraries, the way I'm rebuilding them on Windows with mingw: - move into the libsrc folder - run make - mv *.lib ../lib/clibs If you're willing to keep the o...

Could gcc 13.2 suffice?

https://github.com/BartmanAbyss/vscode-amiga-debug
https://github.com/BartmanAbyss/vscode- ... arto.patch

Back to the original "incipit" of this thread, I think it's quite common for most of the C / C++ environment to give a generous headache to whom is willing to set it up for the first time, no matter on which platform you are. A simpler compiler makes life easier but it won't be powerful, i...

The new modes: -subtype=mono is the one for the less common v1 ROM, which was closer to the ZX81 one asked lacked colour instructions. The optional mlambdav1.lib and mlambdav1_tiny.lib maths optimisations are the valid replacements equivalenti to m81 or mlambda. -subtype=cac3. Is for a less common v...

Few years ago the zx81.de team shared interesting details related to the Lambda clones. The well known color models had an earlier ROM version, probably too similar to the ZX81 one and missing the INK/BORDER/PAPER statements. Moreover another variant used in the CAC-3 and the NF300 included support ...

You're awakening my interest
Really I could never be of any help with the Coleco target, but those efforts with a Pascal cross development are something which happened also elsewhere (the earlier 16 bit Apple systems, in example).
Please keep on digging into the code

I described the -DAMALLOC1, -DAMALLOC2 and -DAMALLOC3 options here: https://www.z88dk.org/forum/viewtopic.php?t=5859 They're three possible presets based on the assumptions that you have some free memory between the user area and the stack, it's something an IT purist (@Alvin ? :D) possibly wouldn't...

- C is suited to this task much more than BASIC , thanks
..which, if you're using a Sinclair computer, you can still mix with BASIC