sam-astro / astro8-computer Goto Github PK

In memory layout range of "Variable and General Purpose Memory" is from 16528 to 61439, and range of "Video Memory" form 61439 to 65535
End of VGPM overlaps with begin of VM and (if I calculated it correctly) VGPM end must be 61438.

BTW, I'm noticed this bug on creating arrays for Armstrong, I haven't finished yet but maybe you would be interested ;).

Is your feature request related to a problem? Please describe.
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
Bitwise math is like the core of binary

Describe the solution you'd like
A clear and concise description of what you want to happen.
or xor xnor and nand left-shift right-shift etc

Additional context
Add any other context or screenshots about the feature request here.

Is your feature request related to a problem? Please describe.
The current assembly is very spaced out and takes up much more space than it needs to by only using actual CPU instructions.

Describe the solution you'd like
Substituting some commands that take up multiple line with single-line ones.

which is faster; Astro8, 8086, or MC68000?

Is your feature request related to a problem? Please describe.
To be honest earlier version of documentation was better, felt less bloated. Everything was on one page, no need to jump through couple pages to find that part you were interested.

Describe the solution you'd like
Revert or rework documentation. New documentation feels over the top.

Is your feature request related to a problem? Please describe.
The current system automatically throttles the video to only render at/around 60fps, which keeps the video circuitry from taking up most of the CPU's execution time. Although this doesn't scale the CPU cycle rate, so calculations per second will vary drastically between different computers,

Describe the solution you'd like
Better speed scaling so all systems execute at a similar speed.

Describe the bug
jmp instruction is limited between 0-2047, so any programs longer than that fail

A new jmpl instruction should be added for jumping to large addresses (larger than 2047). This will use the following word as an address, rather than the 11-bit instruction area.

Describe the bug
The logisim version of the CPU does not function the same as the emulator, it possibly has memory accessing problems or something that causes programs to execute improperly.

Documentation related to how the CPU works and processes data, and also how the assembly and Armstrong work and their syntaxes.

Describe the solution you'd like
Maybe 8 or so 6x6 characters which can be redefined in memory

two indexs are equal. And other two indexs are not sequentially

In https://sam-astro.github.io/Astro8-Computer/docs/Architecture/Memory%20Layout.html the addresses are decimal, when in Armstrong it says that hexadecimal should be used for addresses

Describe the bug
Its not possible to set or see (idk for now) last pixel on emulator on Linux (maybe on windows too but I dont has windows so idk) from version 0.2.0 upwards (0.1.1 works)

To Reproduce
Steps to reproduce the behavior:

1. Open rainbow_gradient.armstrong
2. See error (last pixel is black)

Expected behavior
Whole screen is "rainbow"

Additional context
Screenshot of bug

Is your feature request related to a problem? Please describe.
It is currently not possible to easily access fixed areas of memory without direct addressing, and that way also doesn't allow for dynamic indexing and isn't allocated automatically in variable memory.

Describe the solution you'd like
Add arrays, where values are accessed by array[<index>], and new arrays are created array $name, <size>

Describe the bug
When you run the emulator, and dont give an input (just press enter) the program will segfault.

To Reproduce
Steps to reproduce the behavior:

1. run program
2. press enter
3. question your life choices
4. See error

Expected behavior
not entirely sure, but definitely not a segfault. Maybe a help message?

Additional context
Add any other context about the problem here.

The current implementation of the sound synthesizer has audio artifacts and pitch inaccuracies, mainly because each channel is loaded from an audio file and its speed changed, instead of being generated dynamically.

The system needs to be remade to create 4 channels of audio (1 and 2 are square waves, 3 is a triangle wave, and 4 is noise).

Unfortunately, I don't have much experience with creating an audio synthesizer, so any help would be appreciated.
An application you can use to test the sound is the music tracker at https://github.com/sam-astro/MusicalYabal/blob/main/musical.asm
(press underscore/minus key to play/pause)

Is your feature request related to a problem? Please describe.
The current resolution is somewhat low and unrealistic to use legitimately.

Describe the solution you'd like
Increase the resolution to 240x180 pixels (40x30 characters). This will require a rework of the memory and an additional 43.2kb is needed for pixels, along with 1.2kb for characters.

The Linux version almost works, but there are a few errors preventing it from running correctly.

Is your feature request related to a problem? Please describe.
I'd love to make a raytracing demo, but without floats it seems beyond impractical, and more just impossible.

Describe the solution you'd like
Signed fractional numbers with add, sub, mult and div operators. Even very limited precision will do the job, since screen resolution is so small.

Describe the bug
Passing path to command prompt for emulator results in error.

The pong game isn't very fun because the EX port can only register one key at a time.

I'm not sure how to fix this, but you mentioned in the video that the EX port is 16 bit and "available for any device to use", so something could be done more with that, and utilize more of the port.

It is pretty disorganized right now, so separating these parts would make it better.

Is your feature request related to a problem? Please describe.
Some complex programs run much slower than they should, even though the emulator stays at ~60fps.

Describe the solution you'd like
CPU frequency to be controlled instead of the FPS, which could cause lower FPS ~20 or 30, but will make sure all emulators run at the same speed across computers.

To Reproduce
Steps to reproduce the behavior:
1.git clone https://github.com/sam-astro/Astro8-Computer
2. cd Astro8-Computer/Astro8-Emulator/linux-build
4. cmake ..
5. make

[ 50%] Linking CXX executable astro8
/usr/bin/ld: CMakeFiles/astro8.dir/main.cpp.o: in function `destroy(SDL_Renderer*, SDL_Window*)':
main.cpp:(.text+0x14a3): undefined reference to `Mix_FreeChunk'
/usr/bin/ld: CMakeFiles/astro8.dir/main.cpp.o: in function `Update()':
main.cpp:(.text+0x20e5): undefined reference to `Mix_Playing'
/usr/bin/ld: main.cpp:(.text+0x2108): undefined reference to `Mix_Playing'
/usr/bin/ld: main.cpp:(.text+0x2145): undefined reference to `Mix_Playing'
/usr/bin/ld: main.cpp:(.text+0x2422): undefined reference to `Mix_Playing'
/usr/bin/ld: main.cpp:(.text+0x2452): undefined reference to `Mix_HaltChannel'
/usr/bin/ld: main.cpp:(.text+0x2489): undefined reference to `Mix_PlayChannel'
/usr/bin/ld: CMakeFiles/astro8.dir/main.cpp.o: in function `InitGraphics(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, int, int, int)':
main.cpp:(.text+0x3f3e): undefined reference to `Mix_OpenAudio'
/usr/bin/ld: main.cpp:(.text+0x3f68): undefined reference to `Mix_QuerySpec'
/usr/bin/ld: main.cpp:(.text+0x3f7a): undefined reference to `Mix_AllocateChannels'
/usr/bin/ld: main.cpp:(.text+0x400a): undefined reference to `Mix_LoadWAV'
/usr/bin/ld: main.cpp:(.text+0x40c1): undefined reference to `Mix_LoadWAV'
/usr/bin/ld: main.cpp:(.text+0x4180): undefined reference to `Mix_LoadWAV'
/usr/bin/ld: main.cpp:(.text+0x423f): undefined reference to `Mix_LoadWAV'
/usr/bin/ld: CMakeFiles/astro8.dir/main.cpp.o: in function `destroy(SDL_Renderer*, SDL_Window*)':
main.cpp:(.text+0x14e0): undefined reference to `Mix_Quit'
/usr/bin/ld: CMakeFiles/astro8.dir/main.cpp.o: in function `setupPlaybackSpeedEffect(Mix_Chunk const*, float const&, int, bool, bool)':
main.cpp:(.text+0x15d7): undefined reference to `Mix_RegisterEffect'
collect2: error: ld returned 1 exit status
make[2]: *** [CMakeFiles/astro8.dir/build.make:99: astro8] Error 1
make[1]: *** [CMakeFiles/Makefile2:83: CMakeFiles/astro8.dir/all] Error 2
make: *** [Makefile:91: all] Error 2

Expected behavior
able to compile and use the Emulator from github build

Additional context
Yes i have SDL installed

extra/sdl2 2.26.2-1 [0B 4.43MiB] [Installed]
extra/sdl2_image 2.6.2-4 [0B 199.44KiB] [Installed]
extra/sdl2_mixer 2.6.2-2 [0B 393.83KiB] [Installed]
multilib/lib32-sdl2 2.26.2-1 [0B 2.21MiB] [Installed]

i do not have extra/sdl2_gfx or extra/sdl2_net or extra/sdl2_ttf installed
these errors look like sound errors though (which should be handled by mixer

Is your feature request related to a problem? Please describe.
This can much improve jumps in the code, for example we can write programCounter to @A, jump to other part of code and after few instructions jump back to address written in @A

Describe the solution you'd like
I think best name for that register is @PC

Describe the bug
Pointers used directly in an arithmetic operation fail to compile properly.

Doesn't work:

mult *$memoryLocationA,64 -> $tmp

Works:

mult $variable,64 -> $tmp

This will either be a pull request or a fork.

Using Qt, the interface system will be reworked to hopefully make it windowed-only. This will help with MacOS .app bundling and making the application simple to use for people not familiar with the console. This will also expand upon debugging, for example, pausing, playing, frame-by-frame, and memory viewing. If you still wish to use the console command, then you still can.

And no, I am not doing this just for the purpose of Mac support, it's actually a feature that would be helpful for all the other platforms as well.

If I end up not being able to do this fully in C++, then I may make an emulator in Python or Java and see what I can do with that. If anybody want to help, then that would be nice, since it would give me motivation.

Thanks, everyone, and thanks @sam-astro for making a project that I actually have the passion to help for.

I'll try building this through MacOS. If it works, I'll upload it here. If not, I'll see what I can do to make it work.

To build it yourself:

Make sure you have:

XCode Command Line Tools (May be installed already on your computer)
SDL2
CMake
Make
A C++ compiler

Now make sure you have Homebrew

Install the requirements on Terminal with:

#xcode-select --install
brew install sdl2
brew install cmake

The rest of the building is the same as the normal source building.

If this brings you an error than let me know and I'll try and find the cause.

This is probably a stupid question, but where is the windows exe? Do I have to compile it myself?