interest in a research software story about phoenix libraries

[vincent-pollet]

What is the topic of the new content?

Phoenix - An ecosystem of minimalist, modular and extensible libraries for transverse topics in data acquisition and analysis software in C++, Python and Rust.

Description of the new content

Would there be interest for a research software story about the phoenix libraries?

The research story will describe the motivation for the phoenix libraries: a set of modular and extensible tools, consistent in C++, rust and python, ranging from common utilities (parsing configuration, arguments, handling paths) to more complex yet transverse topics such as data serialization, data exchange, interactions mocking/testing, packaging and CI/CD... The libraries are designed to be used as back-ends, easily replaced/updated as technology evolves, therefore enabling sustainable and maintainable development of research software across long periods of time.

Section within RSQKit for the new content

The research software stories section.

[sparkslabs]

Hi @vincent-pollet That would absolutely be of interest.

The key thing about research software stories is that their primary purpose within RSQKit is to allow sharing of what works in practice. The process of writing them is surprisingly introspective, and IMO a useful exercise. If I understand the context of the software correctly, your libraries are what we would classify that as research software infrastructure - and specifically appear to assist with enabling best practices, as well as having their own best practices in development.

In case helpful - while you can just follow the template in the repository, there's 2 extra approaches which might be useful - which I mention below. If you create this story as a PR and submit it, please feel free to assign it to me.

Context: Last November I gave an EVERSE seminar ( youtube link ) about research software stories. This is lengthy. (Probably too lengthy). However, it was really a few brief 10 min presentations given together (Intro, What, Why, Template walkthrough), followed by a guided creation session. There's also the plain notes and a a summary overview

First easy way of writing a version 0

The reason I mention this - the demo of the guided interview based creation approach for a version 0 is an approach to consider. (links below are timestamps into the video.

• Record an interview based on the template
• Transcribe it
• Ask an LLM to fill in the template based on the conversation to produce a version 0

Second easy way of writing a version 0

And in particular this -- I've also created since that mechanism where by you can ask an LLM to act as the role of interviewer/summariser. These are here - with some very brief instructions:

• https://github.com/sparkslabs/EVERSE-Seminar-Nov2025-ResearchSoftwareStories/tree/main/GPT-PromptTools

It's worth noting that this doesn't attempt to do the writing for you - it asks you questions, you provide answers and it stitches them together to give you a version 0 draft. As such it paraphrases as little aspossible and tries to use interviewee text and language. There's an example of a story created this way.

It goes without saying that obviously you can just follow the template since it's relatively self-explanatory at this stage.0

However, you choose to do the writing, many thanks for the offer - it's greatly appreciated!

[vincent-pollet]

Hi @sparkslabs, thanks for all the pointers! Here is the pull request: #599 (I don't think I have rights to assign it to you)

[sparkslabs]

Hi @vincent-pollet - many thanks for this, will review and ping early in the coming week. I've assigned myself.