SystemSculpt
commented
3 weeks ago I totally understand; I should have made it clear that Marker is just a temporary "let's make sure everything works properly" type of deal at first, with much more intuitive/automatic/built-in solutions coming in the next iteration
Marking this as enhancement and leaving it open until this is solved and what I said above comes into production :)
I just upgraded to version 0.6.1 and it was excited by the release notes, but then you lost my interest when I realized that I had to be running something called 'Marker'. I haven't got the technical ability to run my own instance of Marker, and the paid-for service seems to cost $25 a month. Obsidian has a PDF extractor plugin**. Is it not possible for your plugin to make use of the capabilities we already have to interpret PDFs? Requiring the use of some other highly technical service means that your features aren't nearly so accessible to the general user. Otherwise, incredible job you're doing.
(** I obviously know very little about this but in the few minutes I spent looking into this it seems that Marker offers very high quality conversion of PDFs using advanced AI models. And forgive me if I'm mistaken but in order to provide the text from a PDF for what SystemSculpt is doing, I would assume that you would need just the raw text. The fine detail of the PDF's format and layout isn't going to be that useful. The 'Text Extractor' community plugin seems to provide that without the hassle or expense of a third party service)