Compile error

[gardotd426]

❯ ninja -C build                                                        

 ninja: Entering directory `build'
[7/12] Compiling C++ object decompress-cpu.p/decompress-cpu.cpp.o        ../decompress-cpu.cpp: In function ‘int main(int, char**)’:
../decompress-cpu.cpp:89:21: warning: comparison of integer expressions of different signedness: ‘__gnu_cxx::__alloc_traits<std::allocator<char>, char>::value_type’ {aka ‘char’} and ‘std::size_t’ {aka ‘long unsigned in ’} [-Wsign-compare]
   89 |     if (out_data[i] != (i & 127)) {
[9/12] Generating gdeflate_shader.spv.h with a custom command
../third_party/gdeflate-shader/gdeflate.comp
[12/12] Linking target decompress-gpu

[gardotd426]

Wierdly enopugh it still compiles. It would be helpful to add a --help to any of the commands, because running ./compress img.png tells me how much it's initial raw size was and how much it was compressed, but it's not compressing the original image in place and it's not seemingly copying the compressed version anywhere. So I have no idea where the compressed file is going..

But even running decompress_gpu and decompress_cpu on a regular screenshot shows the INSANE difference:'

compressed throughput: 0.811209
uncompressed throughput: 195.453651

CPU:

compressed throughput: 0.013387
uncompressed throughput: 3.225473