sml VS stb

Anyway the need is not complicated, I need both entry and exit functions from every state, optionally allowing these functions to be coroutines (that's special sauce but for later), and an explicit state transition table which also have a way to say that a state can be accessed at any point. Also allow states to be state machines, optional FSM hierarchies. If you ignore the coroutine stuff it's pretty standard features these days, except that - Boost.MSM is quite archaic now (it was so novel when it was first released...) although it allows most of the features I talked about, I just think it will complicate my code unneecessarilly; - Boost-Ext.SML (not Boost) is almost perfect except it doesnt have entry/exit functions on states for some reason. Also last time we (as in in livestream) tried it in prototypes it didnt compile on msvc XD - Boost-Ext.SML2 is even better but still doesnt have entry/exit functions although it's in the plans.

Hello all. Being a huge fan of state machine and coroutines, I have been browsing around for examples of what other people do combine these two. I have been using boost-ext/sml for quite many projects and are quite happy about that. But when transitioning to code that relies on coroutines, I would like to write entry/exit/actions/guard methods that uses coroutines and where I can co_await on awaitables from Asio and more recently "Boost Async".

boost-ext/sml: quite modern way of doing state machines using a DSL

I recently was troubleshooting a crash that backtraced through the boost::sml library [0]. The crash didn't actually have anything to do with the library, but it was used as the core event loop.

The backtrace -- as in, just the output from running `bt` in GDB -- was over a thousand wrapped lines long. There were ~5 stack frames that took up 200 lines of console each to print just the function name. That product's debug builds recently hit the 2GB line, which is enough that old versions of binutils complain.

I don't know what the solution is. There's some really neat stuff you can do with template metaprogramming, and in stripped release builds it compiles down extremely tiny. Plus the code is very clean to read. But it does feel like there isn't any kind of central vision for the C++ debugging experience, and bad interactions between highly-complex modern C++ typing, the compiler, and the debugger are probably only going to get worse unless somebody (the ISO committee? Vendors?) thinks really hard about debugging support.

[0]: https://github.com/boost-ext/sml

I'm a big fan of MSM but what you're experiencing is pretty normal for template-heavy libraries built on C++03 machinery (emulation of variadic templates is the usual culprit). It's probably not the answer you're hoping for, but the real solution is to switch to a library with more modern foundations. (I've been happily using [Boost::ext].SML for a few years but I'm reluctant to strongly recommend anything in particular since I haven't re-explored the problem space since I found it.)

The Boost one that you mention, seems to be one that it's kinda old. A "successor" of sorts is Boost SML. I've not used it yet, but certainly the first impressions are very good.

Or Boost.Sml

At our company, we rely a lot on https://github.com/boost-ext/sml

I'm a big fan of boost::sml for representing state machines.

https://github.com/boost-ext/sml (State Machine DSL and backend for perfomance)

Have you considered not using an engine at all, in favor of libraries? There are many amazing libraries I've used for game development - all in C/C++ - that you can piece together:

* General: [stb](https://github.com/nothings/stb)

Great to see more accessible references on font internals. I have dabbled on this a bit last year and managed to have a parser and render the points of a glyph's contour (I stopped before Bezier and shape filling stuff). I still have not considered hinting, so it's nice that it's covered. What helped me was an article from the Handmade Network [1] and the source of stb_truetype [2] (also used in Dear ImGUI).

[1] https://handmade.network/forums/articles/t/7330-implementing....

[2] https://github.com/nothings/stb/blob/master/stb_truetype.h

So I read through the materials on mesh shaders and work graphs and looked at sample code. These won't really work (see below). As I implied previously, it's best to research/discuss these sort of matters with professional graphics programmers who have experience actually using the technologies under consideration.

So for the sake of future web searchers who discover this thread: there are only two proven ways to efficiently draw thousands of unique textures of different sizes with a single draw call that are actually used by experienced graphics programmers in production code as of 2023.

Proven method #1: Pack these thousands of textures into a texture atlas.

Proven method #2: Use bindless resources, which is still fairly bleeding edge, and will require fallback to atlases if targeting the PC instead of only high end console (Xbox Series S|X...).

Mesh shaders by themselves won't work: These have similar texture access limitations to the old geometry/tessellation stage they improve upon. A limited, fixed number of textures still must be bound before each draw call (say, 16 or 32 textures, not 1000s), unless bindless resources are used. So mesh shaders must be used with an atlas or with bindless resources.

Work graphs by themselves won't work: This feature is bleeding edge shader model 6.8 whereas bindless resources are SM 6.6. (Xbox Series X|S might top out at SM 6.7, I can't find an authoritative answer.) It looks like work graphs might only work well on nVidia GPUs and won't work well on Intel GPUs anytime soon (but, again, I'm not knowledgeable enough to say this authoritatively). Furthermore, this feature may have a hard dependency on using bindless to begin with. That is, I can't tell if one is allowed to execute a work graph that binds and unbinds individual texture resources. And if one could do such a thing, it would certainly be slower than using bindless. The cost of bindless is paid "up front" when the textures are uploaded.

Some programmers use Texture2DArray/GL_TEXTURE_2D_ARRAY as an alternative to atlases but two limitations are (1) the max array length (e.g. GL_MAX_ARRAY_TEXTURE_LAYERS) might only be 256 (e.g. for OpenGL 3.0), (2) all textures must be the same size.

Finally, for the sake of any web searcher who lands on this thread in the years to come, to pack an atlas well a good packing algorithm is needed. It's harder to pack triangles than rectangles but triangles use atlas memory more efficiently and a good triangle packing will outperform the fancy new bindless rendering. Some open source starting points for packing:

https://github.com/nothings/stb/blob/master/stb_rect_pack.h

https://github.com/ands/trianglepacker

The STB headers are mostly built like that: https://github.com/nothings/stb

You could also add an optional 'convenience API' over the lower-level flexible-but-inconvenient core API, as long as core library can be compiled on its own.

In essence it's just a way to decouple the actually important library code from runtime environment details which might be better implemented outside the C/C++ stdlib.

It's already as simple as the stdlib IO functions not being asynchrononous while many operating systems provide more modern alternatives. For a specific type of library (such an image decoder) it's often better to delegate such details to the library user instead of circumventing the stdlib and talking directly to OS APIs.

My stuff for instance:

https://github.com/floooh/sokol

...inspired by:

https://github.com/nothings/stb

But it's not so much about the build system, but requiring a separate C/C++ compiler toolchain (Rust needs this, Zig currently does not - unless the proposal is implemented).

STB Libraries: https://github.com/nothings/stb

stb_ds is also very popular.