Not sure how to best share news about the app and project - so far I’ve been posting them to Patreon but I guess it would make sense to at least link them here. If anyone has any suggestion, please let me know.
For now here the latest one:
The original search engine in Joplin was pretty limited - it would search for your exact query and that is it. For example if you search for “recipe cake” it would return results that contain exactly this word in this order and nothing else - it would not return “apple cake recipe” or “recipe for birthday cake”, thus forcing you to try various queries.
Hmm, probably not. I don't know what these modules are using for full text search but even then the notes are only searchable from the apps, not from Nextcloud.
If the API is reachable by 3rd party apps some needs to write an app for nextcloud to speak with Joplin. Any further steps would be done in Nextcloud
Isn’t it that simple?
I think Laurent might be referring to the fact that Joplin doesn’t have a central server you could point NextCloud to. You’d have to host a Joplin server somewhere, as well. (Joplin’s API as such isn’t “closed” or anything, as far as I know. And the entire thing is open-source.)
That explains the misunderstanding , it would be self hosted on my NAS as Nextcloud.
The elastic search guy is about to write the documentation how to write a provider app, that is reason why I'm asking
Well, it applies less I guess. But you'll probably have to hack the clipper server to get everything to work because it's not currently made for that. It's really just meant to be a local service running on your computer, not a web http server as we generally understand it.
I've tried converting to a headless server at some point (here) but it wasn't quite working so I didn't continue.
it is a pity , so the only hope that you will find to document jex format in more detail, as required in multiple posts, starting in JEX Format and Joplin E2E Encryption Documentation #408 to enable development of deeper nextcloud integration.
but when you use Nextcloud to sync Joplin via WebDAV, then all the notes are in the Nextcloud file tree as pure text, almost pure Markdown. So, anything that can search there, would also find the Joplin notes. Am I missing something?
Integrating Joplin with Nextcloud in the way many users want is essentially creating a web client and this is a huge task, probalby much bigger than people realise. The JEX format in all this is just a tiny issue - someone who cares enough to implement this feature will also have the motivation to understand the JEX format. It should just take a few minutes and they can easily contact me if they want more info. And at that time I can even write the spec.
Basically, I'm not doing it now because in my experience (with other open source projects) it's very rare that it's eventually useful to someone, so it's not worth it. And also, the format is very simple (and partially spec in api.md).
, it would be self hosted on my NAS as Nextcloud.
, so the only hope that you will find to document jex format in more detail, as required in multiple posts, starting in 
.