25.0222 (0208) Doc URLs Stack URLs Lab notes (Gdrive), Git

4 Non-stack doc/API/AI

Create the non-MERN docs

Phase #2 Doc org/content

This where the tech writer, product owner, marketing etc write the text/diagrams.

This wiki and the Gdrive lab notes are the basis for much of the "starter" text.
API text. Detailed tech descriptions of API usage (I typically wrote much of the original on my previous projects after extensive Postman testing). Includes
- API comments in code and swagger.yaml
- API code boxes (try) in dev site
- Postman collections/docs, etc
- All of the above for GraphQL
Web site text (non-API text) for text sites such as
- Quick start (one of the first docs I like to create on any project)
- User guide
- Dev site (developer.xxx.com)
- Main webpage (marketing)

Use AI to

1 create scaffolding for
- QS
- GS
- dev website
- concepts -main webpage
2 create the content
- search / scrape anything possible
  - from the Gdrive docx's
  - from github code (let AI describe it)
- ask AI to create org, concepts, etc (just for initial version)
3 redo the content
- reorg
- write the final intro/concepts material

25.0222 (0214 reorg) (Gdrive)

Non-stack doc deployments

1 API/API-doc ecosystem (Copilot). Use AI (Copilot) wherever possible.
- Build MERN
- Add API
- Build API docs scaffolding (with AI?)
- Build text doc sites scaffolding (with AI?)
Phase #2 docs.
- AI-create docs.
- Use LLM reasoning to search public and scrape* private text sources.
- Text/diagrams written by the tech writer, product owner, marketing manager, etc (wiki, Gdrive, API test, web site text).
Phase #3 chatbot.
- Copilot-style chat bot using LLM/RAG.
- Scraped* from phase #2 docs.

* "scrape" = use AI tools to crawl over info sources (planning docs, docx's, lab notes) to collect info.

About using AI: If you automate too much of dev and writing with AI, then you might end up like someone with an AI driven (battery powered) car:

You have no control over whats under the hood.
Only a real expert can fix anything.
The AI may fail (and leave you stranded or worse).

I want to use only as much AI as currently is reasonable (with eyes on the future).

NOTES / TODO

My approach to creating docs and tutorials

from https://github.com/terrytaylorbonn/auxdrone/wiki/Docs-tutorials

First create lab notes (docx's in Gdrive) during hands-on exploration of topics.
After organization and content have stabilized (requires time):
- Convert the lab notes into user docs (docx's).
- Publish docx user docs as webcontent.
- Train the LLM on the docx, swagger, UI, etc content (see 5 AI LLM (TODO)).

My tutorial style:

Start with Youtube video, templates (from Vercel, etc), or simple GPT/Copilot
GIT code backups
The goal of lab notes is enabling to reproduce later
Eventually mix the content of docx's (future)
Top down approach to creating demo content (see "CREATING (REPLICATING) / DOCUMENTING DEMOS" below).
- First deploy a minimal viable app.
- Then fill out the app in stages (deploy often to verify).
- Use AI wherever possible.

CREATING (REPLICATING) / DOCUMENTING DEMOS

(need to update this section)

Docx files in the Gdrive (these are for now basically my lab notes)
Git repos

The following summarizes the API sandbox approach to documentation.

1 Updated Gdrive docx's as the source of truth

Youtube videos do not get updated. You need the details, updated regularly, in print. In this wiki the docs are currently Gdrive docx's that are my lab notes (that could be used to create end user docs).

2 GIT code for each chapter

Copy and paste (where appropriate), not endless (error prone) typing.
Explain code without typing it out.
Comment out code that is temporarily not needed.

3 Code deployed in stages

First deploy the minimal functional code (stage1):

Deploy frontend React to Render 1
Deploy backend Node/Express to Render 2
Setup MongoDB

Then deploy for stage 2, 3, until finished.

4 "Drill down" into the code

Understand the top level, then drill down into the lower levels for the backend (in this example from #901 Express, Node, and Mongo)...

... and the frontend (React).

5 Use AI wherever possible

6 Intro the ecosystem and Youtube authors/videos

Notes2

I'd reinterpret some common concepts. For example: An detailed explanation (draft) of an MVC implementation example that explains the confusing terminology (see #901 for details).