debugpy VS wasm3

debugpy description from it's repository:

I'd love to see what it can do with https://github.com/microsoft/debugpy - and especially how it would handle vendored dependencies.

I read that nvim-dap-python uses debugpy which has all these configuration parameters of which one of them is `cwd` which represents the "absolute path to the working directory of the program being debugged".

Hey all, I'm not sure if this is normal or not but, when debugging with nvim-dap, nvim-dap-ui, and nvim-dap-python (which uses debugpy) over a remote session, dap runs exactly once and then closes the entire process. This even happens when I just call `require("dap").restart()`. It's pretty frustrating to deal with and is not something that I was used to when I used to debug with vimspector, which let me restart the debug remotely as many times as I wanted without killing the application that I was connected to. Is this happening for anyone else? How do I disable it? I didn't see anything in debugpy's configuration options. I'm rereading dap's vim :help but haven't found any mention of auto-close / terminate logic yet.

-- Debug adapter plug-in. Debug anything in Neovim { "mfussenegger/nvim-dap", config = function() vim.keymap.set("n", "d", ":DapContinue") vim.keymap.set("n", "dl", ":DapStepInto") vim.keymap.set("n", "dj", ":DapStepOver") vim.keymap.set("n", "dh", ":DapStepOut") vim.keymap.set("n", "dz", ":ZoomWinTabToggle") vim.keymap.set( "n", "dgt", -- dg as in debu[g] [t]race ":lua require('dap').set_log_level('TRACE')" ) vim.keymap.set( "n", "dge", -- dg as in debu[g] [e]dit function() vim.cmd(":edit " .. vim.fn.stdpath('cache') .. "/dap.log") end ) vim.keymap.set("n", "", ":DapStepOut") vim.keymap.set("n", "", ":DapStepOver") vim.keymap.set("n", "", ":DapStepInto") vim.keymap.set( "n", "d-", function() require("dap").restart() end ) vim.keymap.set( "n", "d_", function() require("dap").terminate() require("dapui").close() end ) end, lazy = true, }, -- A default "GUI" front-end for nvim-dap { "rcarriga/nvim-dap-ui", config = function() require("dapui").setup() -- Note: Added this dd duplicate of because somehow the -- mapping keeps getting reset each time I restart nvim-dap. Annoying but whatever. -- vim.keymap.set( "n", "dd", function() require("dapui").open() -- Requires nvim-dap-ui vim.cmd[[DapContinue]] -- Important: This will lazy-load nvim-dap end ) end, dependencies = { "mfussenegger/nvim-dap", "mfussenegger/nvim-dap-python", -- Optional adapter for Python }, }, -- TODO: Make sure that debugpy is installed. Otherwise, disable -- Reference: https://github.com/mfussenegger/nvim-dap-python#installation -- { "mfussenegger/nvim-dap-python", config = function() require("dap-python").setup("/path/to/python/here") -- XXX: Replace this with your preferred Python, if wanted -- An example configuration to launch any Python file, via Houdini -- table.insert( -- require("dap").configurations.python, -- { -- type = "python", -- request = "launch", -- name = "Launch Via hython", -- program = "${file}", -- python = "/opt/hfs19.5.569/bin/hython" -- -- ... more options, see https://github.com/microsoft/debugpy/wiki/Debug-configuration-settings -- } -- ) end, dependencies = { "mfussenegger/nvim-dap", "nvim-treesitter/nvim-treesitter", }, }, -- Remember nvim-dap breakpoints between sessions, ``:PBToggleBreakpoint`` { "Weissle/persistent-breakpoints.nvim", config = function() require("persistent-breakpoints").setup{ load_breakpoints_event = { "BufReadPost" } } vim.keymap.set("n", "db", ":PBToggleBreakpoint") end, }

After testing I have the same situation. I reported this issue on GitHub, you can follow up.

vim.fn.sign_define('DapBreakpoint', {text='🛑', texthl='', linehl='', numhl=''}) dap.configurations.python = { { -- The first three options are required by nvim-dap type = 'python'; -- the type here established the link to the adapter definition: `dap.adapters.python` request = 'launch'; name = "Launch file"; -- Options below are for debugpy, see https://github.com/microsoft/debugpy/wiki/Debug-configuration-settings for supported options program = "${file}"; -- This configuration will launch the current file if used. pythonPath = function() -- debugpy supports launching an application with a different interpreter then the one used to launch debugpy itself. -- The code below looks for a `venv` or `.venv` folder in the current directly and uses the python within. -- You could adapt this - to for example use the `VIRTUAL_ENV` environment variable. return os.getenv("VIRTUAL_ENV") .. "/bin/python" or '/usr/bin/env python' end; }, }

You should take a look at DAP (Debug Adapter Protocol). There exist a Python module implementing this protocol (debugpy) and you can probably find some client applications that can hook into it (like this neovim plugin).

As long as this is happening, might as well try some of my favorites: https://github.com/wasm3/wasm3, https://github.com/WebAssembly/wabt, https://github.com/bytecodealliance/wasmtime

This means that newly created wasm blobs will stop being able to run in wasm3.

On a side note, I can't help feeling sorry for the people that advocate for C over C++ when I see commits like https://github.com/wasm3/wasm3/commit/121575febe8aa1b544fbcb...

It can, wasm3 is a wasm interpretor ported to a lot of bare metal microcontrollers: https://github.com/wasm3/wasm3

On other benchmarks I'm seeing numbers closer to 20% slower, e.g. https://github.com/wasm3/wasm3/blob/main/docs/Performance.md and https://github.com/second-state/wasm32-wasi-benchmark. It's numerical code, which is the best case scenario for a native binary. It's much closer on an average web app or server workload, e.g. https://krausest.github.io/js-framework-benchmark/current.html - you can find WASM frameworks that beat most JS frameworks on there, but that is not as impressive considering the state of the JS ecosystem. Overall, it's already under 50%, and there is still plenty of room for improvement.

Can miniwasm share memory with the host? wasm3 doesn't allow this[1] and requires you to allocate VM memory and pass it to the host, but that has several downsides (some buffers come from external sources so this requires a memcpy; the VM memory location isn't stable so you can't store a pointer to it on the host; etc.).

I'm really interested in a fast interpreter-only Wasm VM that can allow the host to share some of its memory with the VM.

[1]: https://github.com/wasm3/wasm3/issues/114

> This product family will never run Linux, as developers will need to develop a new firmware from scratch. This is obviously not a problem for the earbuds, but a big limitation for the player

Perhaps not Linux, but I suspect there would be a place here for a Unix-like platform that feels familiar. If for example we could get wide-adoption of something like the JVM or wasm3 [0] on these platforms, code could become quite portable, despite wildly different architectures.

For example, Apache's NuttX [1] (that I first learned from Lupyuen, a guy making great progress working with Pine64 products).

> Processing wise, this chip is well sufficient for TWS headphones, but very inadequate for an audio player. It will not drive a good screen, it nor run high-resolution flacs or (probably?) support a high-quality, high-bandwidth codec. In fact, a first generation iPod Nano (retailing for $149 in 2006) had 16MB RAM, so over 16 times what the PinePod would offer. In fact, even the features of any custom firmware are limited from so little memory

I wouldn't call it time just yet. Displays can be interacted with intelligently (to reduce pixel bandwidth) and ultra high quality audio codecs offer diminishing returns, especially when you don't have a DAC or headphones to make the most of them.

My advice to Pine64 would be this:

1. Consolidate your product lines. The Pinebook is just a slower Pinebook Pro, just go with the Pinebook Pro. The PineTab is just a Pinebook without the keyboard, again I would consolidate this with the Pinebook Pro and just make the keyboard detachable.

2. The SBCs should just go straight into the device, thus creating a clear upgrade path for future products. If you want a PineBook Pro running Quartz, just swap the boards (of course with daughter boards for USB expansion, display driver, power, etc).

3. Don't be afraid to kill off products. The Pinebook and PineTab have never seen a new release. The PinePhone appears to be taking a back seat to the PinePhone Pro. The PineCube is basically DoA due to the processing power struggling to process the camera image.

More generally, try to do fewer things, but do them well.

[0] https://github.com/wasm3/wasm3

[1] https://nuttx.apache.org/docs/latest/

[2] https://lupyuen.github.io/articles/sensor

I think SIMD was a distraction to our conversation, most code doesn't use it and in the future the length agnostic, flexible vectors; https://github.com/WebAssembly/flexible-vectors/blob/master/... are a better solution. They are a lot like RVV; https://github.com/riscv/riscv-v-spec, research around vector processing is why RISC-V exists in the first place!

I was trying to find the smallest Rust Wasm interpreters I could find, I should have read the source first, I only really use wasmtime, but this one looks very interesting, zero deps, zero unsafe.

16.5kloc of Rust https://github.com/rhysd/wain

The most complete wasm env for small devices is wasm3

20kloc of C https://github.com/wasm3/wasm3

I get what you are saying as to be so small that there isn't a place of bugs to hide.

> “There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies, and the other way is to make it so complicated that there are no obvious deficiencies. The first method is far more difficult.” CAR Hoare

Even a 100 line program can't be guaranteed to be free of bugs. These programs need embedded tests to ensure that the layer below them is functioning as intended. They cannot and should not run open loop. Speaking of 300+ reimplementations, I am sure that RISC-V has already exceeded that. The smallest readable implementation is like 200 lines of code; https://github.com/BrunoLevy/learn-fpga/blob/master/FemtoRV/...

I don't think Wasm suffers from the base extension issue you bring up. It will get larger, but 1.0 has the right algebraic properties to be useful forever. Wasm does require an environment, for archival purposes that environment should be written in Wasm, with api for instantiating more envs passed into the first env. There are two solutions to the Wasm generating and calling Wasm problem. First would be a trampoline, where one returns Wasm from the first Wasm program which is then re-instantiated by the outer env. The other would be to pass in the api to create new Wasm envs over existing memory buffers.

See, https://copy.sh/v86/

MS-DOS, NES or C64 are useful for archival purposes because they are dead, frozen in time along with a large corpus of software. But there is a ton of complexity in implementing those systems with enough fidelity to run software.

Lua, Typed Assembly; https://en.wikipedia.org/wiki/Typed_assembly_language and Sector Lisp; https://github.com/jart/sectorlisp seem to have the right minimalism and compactness for archival purposes. Maybe it is sectorlisp+rv32+wasm.

If there are directions you would like Wasm to go, I really recommend attending the Wasm CG meetings.

https://github.com/WebAssembly/meetings

When it comes to an archival system, I'd like it to be able to run anything from an era, not just specially crafted binaries. I think Wasm meets that goal.

https://gist.github.com/dabeaz/7d8838b54dba5006c58a40fc28da9...

It allows WebAssembly to be programs which are run using a runtime on the command line, like Wasmtime, Wasmer, Wasm3, etc. Sometimes, you want to have a program which acts like a server, in that it can receive connections and send responses. This is what the patch for Tokio does.