OK, here's a huge topic for you! I've already discussed this with Frank, but while playing with the very good looking adoc Studio, I have some thoughts I wanted to write down. Let's go.
In my company, we write hundreds of texts a day from different authors. Some are better, some are worse. To guarantee consistent quality, we use a so-called prose linter, among other things. In our case, this is Vale, but the manufacturer doesn't matter for now.
But first, let's take a step back. Nowadays, texts can be optimized in two ways: By an AI and by a rule-based algorithm. In my opinion, both are justified.
First things first, AI can optimize texts perfectly. I think everyone agrees on that. But that always requires online access and the confidence to send your texts to a machine like ChatGPT. Or you can set up your own LLM. Have fun with that ;-) But I know from Frank that a connection to some AI is already on the wish list. That's good enough for me and I'll tick the box.
Next to that is a rule-based approach, which I think is extremely important and weighty. Back in the day (Frank, I'm sure you remember), there was the WWB (sigh). It set the standard for all subsequent tools - and it set the bar to the highest level.
But here is finally my suggestion:
- Pick a prose linter and use it as the base.
- Give it some dozen rules that everyone can use (a space after a period, etc.).
- Let the users extend it by language (very important, since even we write in 10 languages).
- Put the tool on its own tab in the sidebar in adoc Studio.
- Then the user can click on an error or a warning and it will be highlighted in the editor.
I know you are very smart & creative and have understood where the journey can (or should?) take you.
Finally, a few links to tools and further information:
Disclaimer, since Frank knows me: I make no promises ;-)
