STM32 VSCode Extension Open-source Plan

[htala]

Does ever ST intend to open-source it's vscode extension?

I don't think it would serve any purpose to hold this project as close source as all the files are already minified version of original source code, so it would not serve any security/IP protection, and ST simply just loose the opportunity of help with other contributors. 

For years ST was preferred for their easy to use IDEs/toolchains compared to other competitors, but nowadays, seems ST does not care about this anymore. 

Using STM32CubeIDE is not optimal in best case scenario and just running super slow CubeMX code generator/Pin configurator is really time consuming. All these problems would be fixed instantly if ST moves to a better IDE instead of Eclipse and VSCode seems the best choice at this point of time. I don't think with the resources ST have, it would take more than 6 month to move all the functionality of STM32CubeIDE to VSCode including configuration/code generation and everything in CubeMX.

 

[mattias norlander]

The move from CubeIDE to VS Code is on-going.

It may seem like we are doing nothing. But VS Code is new to us, we have to have approach it with a "crawl, walk, run" method. Our first major milestone will be to get a maintainable (updateable) tool architecture in-place supporting VS Code. CubeIDE was released 3 times per year. With VS Code we are setting up a more aggressive continuous delivery strategy allowing monthly updates. We are completely re-doing years of tooling architecture based on lessons learned from the community forum feedback, key account customers and looking at competitor and partner solutions. The first step will take some time. Then we hope to see the pay-off of this architecture.

 

The first milestone should arrive in 2025 Q1/Q2. From a feature point of view it may not be super-cool. But it will bring a "updateability" allowing us to rapidly deploy new features and bug fixes. This will be a paradigm shift for customers and also inside ST. :)

We acknowledge that our first attempts in VS Code will not be perfect. But the updatability aspect will allow us to quickly act on your feedback. Things will improve in shorter iterations. 

 

 

Joint open-source development is not a possible short-term strategy given our aggressive target ahead of us.
But, it is something to look into for the future. Our strategy is not "black or white": we will neither be fully open-source, nor fully proprietary. During many years we have contributed a lot to Eclipse/CDT, CMSIS, GCC and Theia to mention a few open-source projects.

We are open for proposals! :)

 

Additionally we want to launch an "early access program" to provide new features before they are released to the mass-market. We want to do this to ensure that roadmap priorities, UX and performance meet customer expectations. Quickly collect feedback and improve. I hope we can launch this during 2025.