One-stop Visual Studio Code Extension to compile, debug and profile Amiga C/C++ programs compiled by the bundled gcc 12.2 with the bundled WinUAE/FS-UAE.
License: GNU General Public License v3.0
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot]")
3 seconds to stop
I had been wondering how I was going to integrate a mod player, but it turns out your sample does this. However the samples generated from the Marketplace install don't include the mod player. Not sure why that is but it seemed worth mentioning.
My codebase uses dynamic memory management - I can't just register bitmap / copperlist / etc. once and be cool with it - I need to be able to unregister them at some point and register next ones a bit later.
One new issue I've run into with the 1.1.0 preview is that if I change the makefile so it'll compile with -O0 for easier debugging then it'll give me compile error in the support code around memcopy and memset. I can fix it by commenting out the special optimization options those functions have but that didn't use to be necessary.
When compiling with -Wextra I get tons of messages near os fn calls:
warning: 'register' is not at beginning of declaration [-Wold-style-declaration]
Of course adding -Wno-old-style-declaration to compile flags fixes problem, but I think this shouldn't be required.
This is of course a very low priority, but a nice fix to have in future.
Hello
I am unable to open console.device despite me following the steps in the reference manuals. I'm thinking it may be an issue with
vscode-amiga-debug since I simply cannot see any issues with the code.
The window, port and IORequest all open up with no errors. It only fails when OpenDevice() is called on line 58.
Feel free to tell me my code is wrong but this is what is in the reference manual for KS 1.3.
My repo is at https://github.com/sacredbanana/DF0Engine
Thanks so much, I'm pulling my hair out trying to get console.device to open :(
Btw I also had to comment out the inclusion of the <rexx/storage.h> header file since that appears to be missing in your extension.
First of all thank you! This is incredible and has really given me the easy path I needed to revisit the Amiga.
I'm not sure if this is the appropriate place to ask questions but here goes.
When debugging C I'm sometimes able to single step through the assembly code, not sure if it works %100 but it certain is possible.
However if I write some inline asm code then I'm unable to single step through that code, it'll keep bouncing back to the C source.
Is this something would I can fix? Or is there something in the setup that would need to be improved?
This was super easy to set up but it would be good to have integration with ThePlayer or another ProTracker replay routine in the example project as it is rather hard to get that one working with C (at least for me).
So, I am guessing from your configuration that we are supposed to add a "config" JSON property that matches configurations[x].config for one of the configuration block objects in the launch.json file. However, the only way I can get it to do anything other than choose the first configuration is to make something else the first configuration manually. I can only specify a newer kickstart. Nothing I do can get it to use the A1200 or A4000 configs.
Also, the A4000 configuration is not setup in the version of your extension pulled from the Visual Studio Code extension search. Am I doing something wrong? I'd like to develop some apps, not games, using this extension and they are aimed at "newer" Amigas running 3.1.4 and so on.
Is there a way to tweak the WinUAE settings applied to each config?
Hi!
Perhaps not the right place, or the right forum even.
I am new to C on Amiga (emulated), and would like to test it. I found your(s) excelent work here, it looks nice, uses visual studio code which is a nice editor...
I did some tests, opened a few windows and wrote Hello World as one should, after some changes and added includes (like missing rexx if you include alib_proto.h) it worked.
However, for me I am stuck and mostly that depends on missing documantation or the inability to find the right docs.
I also tried to do an Hello World with MUI and this is why I am adressing this topic, I got completely stucked, MUI apparantly wish to call a routine called DoMethod, I scanned the whole MUI NDK and could not find it, I scanned the whole internet and with only sparse information I could figure out that the routine where located in alib.lib or amiga.lib, I think the latter.
So I got an error when linking could not find where it was defined when linking (more or less) I added in different combinations lamiga lalib llibamiga and whatnot, still no luck, error was could not find whatever .lib I added.
So I explored the [user]*abyss**. thing, could not find any alib.lib or amiga.lib... so I "stole" a version from vbcc and put in various places, no luck.
I started Process Monitor (sysinternals) and did see that ld where trying to do something with /libamiga.(a|so) files, i read the docs for ld, and indeed it stated that it puts lib infront of the libraries I wish to use, I tried to change the name of the lib just to test. Did not work...
Anyway despite this long and for you I guess, kind of boring description of my problem. I wonder if you would like to include some help documentation how to solve various easy problems for an experienced coder, but perhaps a "stopper" for any unexperienced ones with your toolchain solution, or perhaps links and references to where one could solve the problem at hand.
//Thank you for your time, and my appology for the rather unusual "request" of mine. And thanks for a nice tool to test out and perhaps learn to use
//HEA
Would it be possible to add the ability to create a bootable ADF containing the compiled app along with its required assets as long as it fits onto a floppy?
Hi there,
Due to recent developments of your extension, I'm trying to support both yours and Bebbo's GCC in my code.
To this point I've used stdio.h, string.h and other std headers excensively. I see that your implementation lacks lots of fns. Perhaps you could include https://github.com/bebbo/libnix (even the ks1.3+ variant) in your toolchain?
More regarding libnix: https://github.com/bebbo/amiga-gcc/wiki/Libraries
Although not released yet, I checked the sound playback capability available on master branch. Everything works fine with A500 configuration (around 20-25% CPU), but switching to A1200 config with the same binary gives big lags and CPU on the level of 130-150%, with clicky sound and slowed down execution.
In order to separate the problem, I reassembled P61 player example from original sources available on Aminet and it sounds exactly the same (I can provide the exe if needed). Surprisingly, running the same binary on vanilla A1200 WinUAE setup (although 4.3.0, which IMHO makes no difference) runs perfectly smooth, so I guess there's something in the WinUAE configuration that makes it laggy. I tried to adjust the settings to match the not lagging one using the WinUAE GUI, but cannot find the culprit.
I'm just wondering if the problem also appears on your setup. I'm running Windows 10, WinUAE 4.4.0 as attached to current version of the extension.
PS: great thanks for bringing back my memories and allow me to learn how machine of my childhood/adolescence works and can be programmed. You made easier to fulfil my nostalgic dream of programming an Amiga computer!
I'm pretty sure i would work round this issue but I wanted to give you feedback on a clean machine running the project for the first time.
Perhaps some hard coded path in there somwhere that needs updated?
> Executing task: c:\Users\stephen\.vscode\extensions\bartman-abyss.amiga-debug-0.7.0\bin\gnumake.exe -f Makefile8mingw -j4 <
Makefile8mingw:41: target 'obj8mingw/main.o' given more than once in the same rule
Compiling bin/elf2hunk.c
C:/github/vscode-amiga-debug/bin/elf2hunk.c:10:10: fatal error: stdio.h: No such file or directory
#include <stdio.h>
^~~~~~~~~
compilation terminated.
gnumake: *** [Makefile8mingw:44: obj8mingw/elf2hunk.o] Error 1
The terminal process terminated with exit code: 1
I wonder if this is only triggered due to the modification date on a clean checkout?
In order to contribute to the project, it would be nice to know how to set up a dev environment so we can best provide you with PRs that you'll likely approve and merge.
I presume it starts with
I saw some steps on requirements on the README.md for porting, but this is different. What do we have to have present in order to contribute. Help me and I'll add a wiki page to the repo for others to follow.
The supplied scripts in the package.json file are as follows:
{
"scripts": {
"vscode:prepublish": "concurrently -r \"npm:compile:ext\" \"webpack --mode=production\"",
"compile": "concurrently -r \"npm:compile:*\"",
"compile:ext": "tsc -p ./",
"compile:client": "webpack --mode=development",
"watch": "concurrently -r \"npm:watch:*\"",
"watch:ext": "tsc -watch -p ./",
"watch:client": "webpack --watch --mode=development --info-verbosity verbose",
"serve": "webpack-dev-server --mode=development",
"pretest": "npm run compile:ext",
"test": "node ./out/test/runTest.js"
}
}None of these create a vsix file for testing.
Hello,
is it possible to run the cool extension under OS? You could exchange WinUAE with FS-UAE.
Greetings
I've been trying to use mulsw to make sure the compiler does a 16x16=32 multiplication. But as things are the compiler seems to assume mulsw results can be truncated to 16bits. I'm seeing generated code that does this. I think its because the inline asm used 'a' for the result and 'a' is a short. So it optimizes out the additional bits you might hope for in the returning int.
Well, anyway I think this alternative works:
inline int mulsw(short a, short b) {
int r = a;
asm("mulsw %1,%0":"+d"(r): "mid"(b): "cc");
return r;
}
But you will probably want to check that yourself and do the same for muluw.
My current project has hundreds of fns - looking for specific one is very tedious. ;) Adding filtering out names when typing would significantly increase usability.
Alternatively, perhaps adding a "jump to disassembly" context menu on profiling blocks would do the trick.
Right now it is very easy and convenient to create a C/C++ project: simply press Ctrl+Shift+P, select Amiga: Init Project, and you're all set. This is awesome!
But unless I am missing something (in which case: sorry!), there is no such easy way to create an Assembler project.
Since changes so far seem to be stable, perhaps it's time to update extension in vscode marketplace? This is mainly for #7 because it's needed to set up CMake workflow correctly, which my library extensively use, and I have some ppl asking me when my project will be compatible with your extension. ;)
Hi,
I've started developing a project along with a friend. Everything works nice, except that we have kickstart ROMs in totally different filesystem paths. I know that we can just unify that on our side, but I think that it would be more beneficial to have it predefined globally for whole extension. This way the launch.json won't change and could be commited on repository and used by anyone without any changes.
There are two ways in which it could be solved:
"kickstart": "${amiga.kickstartPath}/kick13.rom",. It would still require filenames to be exactly the same, but it's still better than current approach.amiga.kickstarts object which could be assigned with "name": "filePath", like:"amiga.kickstarts": {
"kick13": "/path/to/my/kick13.rom",
"kick30": "/path/to/my/kick30dumpedFromMyAmiga.bin"
}Dunno which approach is better. This is a low priority issue for me. ;)
Hello, thanks for this awesome project. I've just tried it out with a fresh install from marketplace (v1.0.0) and wrote a simple helloworld program. It works as expected but winuae-gdb.exe crashes after a while, sometimes after 15 seconds and sometimes a little later, a little random when it occurs but seem to always happen. It is highly reproducible. This is when running on A500 KS 1.3 (kick34005.A500) rom with or without debugging. If I close the winuae window by clicking [x] it freezes and hangs. Trying the same on A1200 KS 3.1 (kick40068.A1200) or KS 2.0 (kick37175.A500) works fine without any crashes and doesn't hang. I'm using a win 8.1 machine and latest VSCode 1.47.3 64-bit. The crash seems to be caused by different dll calls e.g. to ntdll.dll, details below, any clue what is causing this? Tried on KS 1.2 too and same problem there so problem seems to be tied to when using older Kickstarts.
Problem signature:
Problem Event Name: APPCRASH
Application Name: winuae-gdb.exe
Application Version: 4.4.0.0
Application Timestamp: 5ed1714a
Fault Module Name: ntdll.dll
Fault Module Version: 6.3.9600.19678
Fault Module Timestamp: 5e82c88a
Exception Code: c0000005
Exception Offset: 00000000000358fa
OS Version: 6.3.9600.2.0.0.256.48
Locale ID: 1053
Additional Information 1: cee4
Additional Information 2: cee49baf695955cc06ce73bf396842b0
Additional Information 3: ac31
Additional Information 4: ac31d5743764a6b1b96c6fe1859d6916
Problem signature:
Problem Event Name: APPCRASH
Application Name: winuae-gdb.exe
Application Version: 4.4.0.0
Application Timestamp: 5ed1714a
Fault Module Name: StackHash_2269
Fault Module Version: 6.3.9600.19678
Fault Module Timestamp: 5e82c88a
Exception Code: c0000374
Exception Offset: PCH_C3_FROM_ntdll+0x0000000000090CBA
OS Version: 6.3.9600.2.0.0.256.48
Locale ID: 1053
Additional Information 1: 2269
Additional Information 2: 2269d2f9a7df7b894866693deedbe695
Additional Information 3: 2387
Additional Information 4: 238760144906e0a39667ea3de5604a11
Problem signature:
Problem Event Name: APPCRASH
Application Name: winuae-gdb.exe
Application Version: 4.4.0.0
Application Timestamp: 5ed1714a
Fault Module Name: GDI32.dll
Fault Module Version: 6.3.9600.19781
Fault Module Timestamp: 5f0d29d8
Exception Code: c0000005
Exception Offset: 000000000000972c
OS Version: 6.3.9600.2.0.0.256.48
Locale ID: 1053
Additional Information 1: e6da
Additional Information 2: e6dad71e24846c6e936596e26d8739b2
Additional Information 3: 23c4
Additional Information 4: 23c498ae9be98440f541d454c625566d
Hi Bartman
Is there a way to code a debug break point in C or inline asm? I thought maybe calling the exec function Debug(0) would do it. That seems to be an old way to hook into debuggers. How does the framework trigger break points? Maybe I can just emit the correct magic code?
When compiling inline asm I seem to always get a bunch of "unknown register name" warnings.
The sample project does this for the unpacking and player inline asm.
It hasn't been a problem until recently. I now have a few different blocks of inline asm that need to specify the registers to use for some variables because they'll be used with a "movem". I do that the usual way: "register int var asm("d0");" for example. However the compiler is not always enforcing the register specification, I suspect the optimizer is doing something because it works okay with compiling with -Og.
It seems plausible that the problem stems from those compiler warnings. Do you know why those warnings are happening and if they can be fixed correctly?
Fantastic project. I love the one stop shop idea.
I'm a big VSCode user. However I cannot get the extension install. I dont even know where to start.
What deps and steps are needed to get the vscode extension module built. Could we get a binary install package or even better integrate into the marget place?
Hi,
Having gcc_support files inside a project is little troublesome. When using newer versions of plugin I had problems with building code containing old support files, so now I'm trying to update them along the way. It raises some problems, though. Having different versions of those files in git history of my project is problematic - If I go back in history, I have to apply support files which are compatible with current version of extension, and I have to do it each time when I'm bisecting commits to find regression bug. The problem escalates to all of my projects because I've made a poor decision of adding those support files to my game library to have plug-and-play experience across repos, but it prevents me from freely selecting the commit from which the library is built.
I think it could be solved by bundling support libs directly with compiler:
#include <extamiga.h> and linking with -lextamiga (-lamiga name is already taken for amiga library present in most of other compilers)The following lines gets added to .vscode/settings.json:
"C_Cpp.default.compilerPath": "path\\to\\m68k-amiga-elf-gcc.exe",Which is not fine because it's also appended in non-amiga projects, including my work projects. I know I can disable extension for given workspace, but i don't need this line 95% of the time and I have 100+ projects on my disk. I keep removing it from non-amiga projects but that line comes back.
Perhaps this line could be added once when project gets created from template?
Now since my library and games are building correctly, I still need one thing to be completely happy - ability to load files. Right now I see that there is a dh0 directory in extension's directory, which I assume is used to run the executable.
Would it be possible for extension to dynamically bind executable's original directory as UAE's main DH, so that when launching executable via debug task it'd be able to find files via relative paths?
Hi,
I'm helping out Virgill with his Amigaklang exe generator which uses your cool gcc environment.
I updated his project from gcc8 to the gcc10 version and it stopped working (corrupt sounds). I have traced it to a commit on the 18th may where the mulsw and other macros were improved. If I leave the gcc10 in place but use the old mulsw macros it works OK again.
I had a look a the macros but the gcc/asm syntax is new to me so I couldn't spot any errors.
Virgill calls mulsw in two places as follows, maybe the error will be obvious to you:
short osc_sine (BYTE instance, short freq, UBYTE gain){ int temp; short res;static short counter[16]; short buf = counter[instance]+=freq; buf -= 16384; temp= mulsw(buf, 32767-abs(buf)); res =temp>>16; res<<=3; return vol(res,gain); }
inline short vol(short val, UBYTE gain){ return mulsw(val, gain) >> 7; }
You can message me here or on EAB if you need me to check/change something.
Cheers
Antiriad
Both preview31 and preview32 do not compile and run the sample application generated by Amiga Init Project.
So I am trying to pare down the minimal required code for an application to work well with the system. The supplied main.c is impressive but it's also got a lot going on. If I wanted to be a good Amiga citizen and follow the rules, and print hello world in the console, I was thinking it would look something like this.
#include "support/gcc8_c_support.h"
#include <exec/exec.h>
#include <dos/dos.h>
#include <dos/stdio.h>
#include <proto/dos.h>
#include <proto/exec.h>
volatile struct Custom *custom;
struct ExecBase *SysBase;
struct DosLibrary *DOSBase;
int main(int argc, char **argv) {
SysBase = *((struct ExecBase**)4UL);
custom = (struct Custom*)0xdff000;
FPrintf(Output(), "Hello World\n");
return 0;
}From what I can tell we need to have the gcc8_c_support stuff, but what else is required? What is the struct Custom *custom bit? I haven't seen that before. This program, by the way, crashes when run in the emulator. Do you know why?
Hi Bartman,
First thing first, nice job on this Amiga tool :)
I started using it lately and noticed an error while trying to use command gnumake clean
Using plugin 1.0.0 with VSCode 1.45.1 on a Windows 10 64 bits machine
Let me know if you need more informations or help
Cheers
Kyle/MenTaSm
See this example:
00004a30 00000024 00004650 FDE cie=00004650 pc=00006ba4..00006c1c
LOC CFA r2 r10 r13 ra
00006ba4 r15+4 u u u c-4
00006ba8 r13+8 u u c-8 c-4
00006bac r13+8 c-4036 c-4032 c-8 c-4
Hi
I was wondering if it would be possible to add a feature where every directory and file inside a special folder in the project automatically gets deployed onto the virtual amiga hard disk rather than having to manually put them in there yourself.
This would make life so much easier! Thanks!
Hi Bartman
Loving this system and the profiler and all the updates! But wanted to ask if you are aware of a way to force alignment for static variables. Using __attribute__((aligned(n)) works fine for structs but in my case I'm just trying to align a statically allocated array.
I'm unable to debug for some regions of my code for some reason. The breakpoints completely get ignored and for earlier breakpoints that work, it is also unable to step into functions where my broken breakpoints reside. The code executes fine but debugging is broken for some reason.
The a500 and a1200 configs works out of the box with added-rom files,
cannot get the a4000 config to work though, got a HALT3 error and a black screen only.
Also kind of expected to be able to fire up a "1260", poking around in amigaDebug.js using
config['quickstart'] = 'a1200,4'; instead of config['quickstart'] = 'a1200,0';
and adding:
config['cpuboard_rom_file']= <Blizzard_1260.rom>
config['cpuboard_rom_file_id']='F88AE0F1,Blizzard 1240/1260'
Did not work either, got a black screen and a HALT3.
[edit] Using "winuae64.exe -portable" instead of "winuae-gdb -portable" with the generated default.uae works though :-)
I highly appreciate your effort making this!
I hope you don't have enough of me and my issues. ;)
The WinUAE 4.4.0 was released recently. It comes with printf-like logging, which works almost the same as your vscode-bound logger. Quoting the WinUAE release topic from EAB:
New easy to use and transparent printf()-like debug logging method for developers: simply write parameters to address $bfff00 (byte, word and long accepted) one by one, then format string to $bfff04 (must be long) and formatted log message will appear in log window. Address may change, currently only active if 128k UAE boot rom mode is enabled and accepts only %d, %u, %x, %p, %s and %b (BSTR). Quickly made for quick and easy UAESND AHI driver debug logging..
I think it'd be cool to be consistent and use same interface to pass the debug messages to vscode. This way, I can use same code to send messages to vscode and later give the exact same build to betatesters with uae logger enabled. It can be done with an ifdef and two code variants, which I already do, but having unified interface is better, I think. ;)
No matter what I do now, I cannot get the extension to reinstall. I played with the unreleased version, then ram into issues and tried to revert to the main version.
I’ve tried removing it using the GUI. I’ve tried removing it from %userprofile%.vscode\extensions and I’m out of ideas. What am I missing?
Hi
Even when you quit WinUAE you still have to push the stop button in VSCode to stop the debug session. Any chance we could get it closed automatically?
I thought I would try endcli/CloseWorkBench to free up some chipmem on the A500 config.
However, it didn't seem to improve how much memory I can allocated with AllocMem(... , MEMF_CHIP). The largest size I can allocate is 427k in both cases.
Am I doing something wrong or is that to be expected?
I vaguely remember that kick1.3 might not close the workbench all the way.
Is there any other way to improve the chipmem I can allocate?
Hi
I have an issue where the system call CloseWorkBench() fails. This call is used to close the main workbench screen. I do this to free up some memory. This normally only fails if some non drawer windows are open but none are. I assume something hidden is preventing the workbench screen to close?
I am profoundly impressed with this extension and it will definitely lead to me writing more Amiga code this year and going forward. I desperately wish there was a version for macos but in the meantime I went to use calloc and found that I could not compile because stdlib.h could not be found.
Is it a setting or is simply not there?
I've had to manually comment out references to the rexx/storage.h file since this file seems to me missing from the extension and my code would not compile otherwise
Hi,
I'm having some problems with debugger features:
I'm using custom CMake to generate .map, .elf and hunk files in similar manner as your makefile, but perhaps I've messed up something. I'll try to debug it myself but I need a little guidance from you. How does vscode know at which address is given static/global/local variable?
-m68000 -fomit-frame-pointer -nostdlib -Wno-unused-function -Wno-volatile-register-var -fno-tree-loop-distribution -flto -fwhole-program -Wno-old-style-declaration-Wl,--emit-relocs,-Ttext=0-Wl,-Map=germz.map linker flag.${OBJDUMP} --disassemble -S ${GERMZ_EXECUTABLE} > germz.s to generate disassembly, where ${OBJDUMP} is objdump from your compiler suite.EDIT: The local variables are most of the time okay, the problem is with static/global vars.
I'm adding inline asm so I can write some fast code without giving up on the convenience of C/C++. However if I use too many registers, such as 8 data register then I run into problems. I haven't taken the time to examine what exactly is going on but was wondering if you know where I can find more information about how the compiler uses the registers. Seems likely there is a stack frame pointer in a6 but from the problems I'm getting it seems like one of the data registers is also in use. I'll look more closely if you've no idea what I'm doing wrong.
I'm maintaining game engine/framework/library/whatever as seen here https://github.com/AmigaPorts/ACE . I use some low level stuff in my code via macros in a way that is compatible with Bebbo's GCC, (kinda) VBCC and x86 (logic only).
For example putting function args in registers:
#elif defined(__VBCC__)
#define REGARG(arg, reg) __reg(reg) arg
#elif defined(__GNUC__)
#define REGARG(arg, reg) arg asm(reg)which is used like:
void INTERRUPT keyIntServer(
REGARG(volatile tCustom *pCustom, "a0"),
REGARG(volatile void *pData, "a1")
);Unfortunately GNUC variant syntax doesn't work with your compiler. In fact I tried some variations and I wasn't able to make it run. Also __far:
#define REGPTR volatile * const
extern tCustom FAR REGPTR g_pCustom; // FAR translates to __far
extern tRayPos FAR REGPTR g_pRayPos;
I've tried putting that __far everywhere but it didn't work. What's the correct syntax in both cases?
Hi,
I've installed this plugin on a Windows 7 system with GIT Bash package. After slight modifications to tasks [removed amiga.bin-path, added static paths] I was able to run compilation but it fails:
process_begin: CreateProcess(NULL, m68k-amiga-elf-gcc -g -MP -MMD -m68000 -Ofast -nostdlib -Wall -Wno-unused-function -Wno-volatile-register-var -fomit-frame-pointer -fno-tree-loop-distribution -flto -fwhole-program -fno-exceptions -Wa,-g -xassembler-with-cpp -c -o obj/gcc8_a_support.o C:/Users/rafamiga/OneDrive/tmp/amiga/amiga-1st/support/gcc8_a_support.s, ...) failed.
It's obvious paths inserted here [and in fact everywhere] adhere to Windows convention while GIT Bash [MINGW64] expects something like /c/Users...
I'm no VS guru nor any kind of expert but can you think of a way to tweak this in either VS or plugin settings?
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.