wip: add RFC about apps

design/008-renku-apps/008-renku-apps.md

@@ -0,0 +1,89 @@
+- Start Date: 03-06-2022
+- Status: Proposed
+
+# Supporting Apps on Renku
+
+## Summary
+
+Renku projects can currently be homes for sessions, datasets, and workflows. This proposal introduces a new kind of session, the _App_ session (in contrast to the existing _Dev_ sessions), so projects can host interactive applications associated with the work and contents of the project.
+
+## Motivation
+
+One common outcome of a data-science project is an interactive application, such as a dashboard, that supports flexibly viewing and analyzing data. The infrastructure necessary for this is already available with sessions, but, since sessions currently implicitly assume a more code-development-oriented use case, there are many UX improvements that could be realized by directly supporting applications as project outcomes.
+
+### Status Quo
+
+To understand the problems, let us consider how a project can offer an app today. One way would be as follows:
+
+- Start a session and install streamlit into the session
+- Develop the app
+- Install jupyter server proxy, configure it for streamlit, and and modify the Dockerfile to use it
+- Modify the renku.ini to set the streamlit end point as the default
+- Update the README.md with a description of your app and provide a badge to start it.
+
+(For an example, see https://renkulab.io/projects/renku-stories/alphapept-gui-streamlit)
+
+Looking at this, there are several problems with the status quo that we can identify:
+
+- By default, the developer will also end up in streamlit when they start a session. Extra work is necessary on the part of the developer to make their own experience using Renku smooth.
+- What happens if we want to offer two apps? That becomes cumbersome.
+- The app session still contains the dev tools (e.g., JupyterLab) and a casual user can end up in the dev environment by accident.
+- An app should not have autosave branches because any changes made are intended to be transient
+- How can I search for projects that have apps? The fact that there is an app in this project is invisible to the search.
+
+By supporting apps as a first-class citizens, we can address these problems and more, improving the UX for both developers and users of apps within Renku.
+
+## Design Detail
+
+The proposal to address this gap is to introduce a new project-level entity -- the **App** -- with metadata and operations specifically for those use cases.
+
+The app metadata needs to let users search for apps and understand at a high level what an app does. It also needs to let developers describe the resource requirements for running the app and how the infrastructure should start the app.
+
+### Metadata
+
+An app needs the following metadata:
+
+| Field        | Purpose                                                              |
+|--------------|----------------------------------------------------------------------|
+| Name*        | The name of the app                                                  |
+| Description* | A short description explaining what the app does                     |
+| Dockerfile*  | A path to a dockerfile for the container that runs the app           |
+| Resources*   | Information matching the _renku "interactive"_ section of renku.ini  |
+| Image        | (optional) A path to an image that represents the app                |
+| Readme       | (optional) A path to a markdown file with more details about the app |
+| Tags         | (optional) Keywords that categorize the app                          |
+
+
+The metadata for apps should be indexed by the knowledge graph (KG), just as is the case for _datasets_ and _workflows_.
+
+### Wireframes
+
+**TODO**
+
+
+## Drawbacks
+
+There are no drawbacks to supporting apps, since Renku users are already creating solutions themselves. By making apps first-class citizens, we can provide better support activities that users are already doing in an ad-hoc way.
+
+**TODO**
+> There are tradeoffs to choosing any path, please attempt to identify them here.
+
+## Rationale and Alternatives
+
+**TODO**
+
+> Why is this design the best in the space of possible designs?
+
+> What other designs have been considered and what is the rationale for not choosing them?
+
+> What is the impact of not doing this?
+
+## Unresolved questions
+
+**TODO**
+
+> What parts of the design do you expect to resolve through the RFC process before this gets merged?
+
+> What parts of the design do you expect to resolve through the implementation of this feature before stabilisation?
+
+> What related issues do you consider out of scope for this RFC that could be addressed in the future independently of the solution that comes out of this RFC?