Hi André,
If you know anyone who would be willing to take this project on and maintain it for the next 10 years, the Skyline team would be eternally grateful. I would even be willing to help look for a way to fund such a "User Education" focused effort.

When I was at Microsoft in the 1990s, it felt like developers like me got assigned a host of support personnel: testers, product designers, release managers, user education, support, and marketing. Even in the two startups I joined before joining an academic lab we did better at filling these roles. It felt audacious, sitting down alone in a cubicle at UW to start a software project. No testers? No documentation team? An academic professor as my only "marketing"? In those early days I did my best to cycle through the roles I have listed above writing a lot of automated tests, doing all the product design, setting up my own release plan and infrastructure, starting the support board, writing tutorials, recording videos, and even in-person education with key adopters.

Now here we are 16 years later with over 15,000 startups a week, a team of 10 devoted to Skyline, including IT and a project manager. We do spend a lot of time on user education. Thanks for your appreciation of our tutorials. We are currently engaged in a major effort to improve our ability to keep the current as the software changes, and especially the Chinese and Japanese translations.

We are an open-source project, though, and we are grateful for the amount the community itself has contributed to our tutorials, Skyline-Online, and in-person education. Please let us know if you see this as an opportunity for you to contribute.

--Brendan

Hi André,

Can you elaborate on why chatbots would benefit from a single readme vs. gathering all the information from the many pages (documentation and tutorials) on skyline.ms? It does seem like a good idea to train a chatbot on all the content of the website if it can be done fairly easily and maintainably. Especially if it can reference the source material.

Not to mention all of our answered support requests and other accumulated tips and tutorial content that has not been promoted to an official Skyline tutorial, like some that are attached to our tutorial webinars.

Hi Brendan,
That is indeed a fascinating path Skyline has taken, and it will certainly continue! The material you have generated, and most importantly, the source code alone, would likely be a great basis for a centralized knowledge base. Unfortunately, it is not my area of expertise. Maybe there is someone out there can scrawl everything and set it up.
The reason, I was asking about a compact readme (this might answer Matt's question) is for the use of a non-central chatbot. For instance, Claude allows you to deposit and store files in projects that will be used as knowledge every time u start a chat there. Since it 'reads' the data every time you ask sth. it is limited by size otherwise it is too (token) expensive. For instance, if you are new to DIA-NN, you can just throw in the readme and can ask anything about proper usage, settings, etc. Works quite well. I am sure a compact readme might also be generated automatically for Skyline (mainly based on the code itself), but would need proper fact-checking afterward.