move Motivations to its own doc, for #3

README.md

@@ -73,11 +73,7 @@ See [the documentation folder](docs).
 
 The "Swarm" name is in reference to the original key function of the UI: enabling a 'swarm' of GPUs to all generate images for the same user at once (especially for large grid generations).
 
-This project is built with a C# backend server to maximize performance while only minimally increasing code complexity. While most ML projects tend to be written in Python, that language is simply insufficient to meet the performance goals of this project (ie to provide a very fast and responsive multi-user-ready multi-backend service), notably it lacks "true" multithreading capabilities (due to Python GIL), which was deemed strongly necessary for StableSwarmUI (it must be able to use available CPU cores while serving user requests and managing internal data to be able to respond to all requests as quickly as possible). It is also hoped that building Stable Diffusion tools in C# will enable a wider range of developers to make use of Stable Diffusion (vs being limited to the Python ecosystem).
-
-The project was designed to be heavily modular, such that backends are fully separated from the middle-layer which is fully separated from the frontend UI, and all components are interswappable. This is to enabled extensibility and customization. For example, an extension can easily provide alternative backend generators (this project comes with several built-in, such as ComfyUI, Auto WebUI, StabilityAPI, ...) without having to edit anything else to work. The limitation of this approach is some tools may not easily be intercompatible, eg the StabilityAPI backend has only a select few limited inputs, vs the local backends that have a wider range, and so many parameters don't work with StabilityAPI.
-
-A completely custom HTML/JS frontend was built with the goal of allowing detailed and thorough customization of the UI (as opposed to eg being locked in to the way Gradio generates things).
+See [the motivations document](/docs/Motivations.md) for motivations on technical choices.
 
 # Legal
 

docs/Motivations.md

@@ -0,0 +1,7 @@
+# Motivations
+
+This project is built with a C# backend server to maximize performance while only minimally increasing code complexity. While most ML projects tend to be written in Python, that language is simply insufficient to meet the performance goals of this project (ie to provide a very fast and responsive multi-user-ready multi-backend service), notably it lacks "true" multithreading capabilities (due to Python GIL), which was deemed strongly necessary for StableSwarmUI (it must be able to use available CPU cores while serving user requests and managing internal data to be able to respond to all requests as quickly as possible). It is also hoped that building Stable Diffusion tools in C# will enable a wider range of developers to make use of Stable Diffusion (vs being limited to the Python ecosystem).
+
+The project was designed to be heavily modular, such that backends are fully separated from the middle-layer which is fully separated from the frontend UI, and all components are interswappable. This is to enabled extensibility and customization. For example, an extension can easily provide alternative backend generators (this project comes with several built-in, such as ComfyUI, Auto WebUI, StabilityAPI, ...) without having to edit anything else to work. The limitation of this approach is some tools may not easily be intercompatible, eg the StabilityAPI backend has only a select few limited inputs, vs the local backends that have a wider range, and so many parameters don't work with StabilityAPI.
+
+A completely custom HTML/JS frontend was built with the goal of allowing detailed and thorough customization of the UI (as opposed to eg being locked in to the way Gradio generates things).