I saw a MySQL node and I think a SQLite node can be useful to store data without to setup a MySQL server.
To declare Credentials, we could just locate the SQLite database file.
In this node, we could :
Execute an SQL query
Insert rows in database
Update rows in database
An usage example :
I would like send a discord message each time a new item in my RSS feed.
But if I don’t know who is the last sended feed, i have to send all !
If I store GUID (Hash of URL) in SQLite database, I could use SQLite node to get the sended feed
Yes totally agree, that a node like it could be interesting for many use cases.
For your example use-case we had actually the following planned:
Create some kind of abstraction layer that allows nodes to read/write data to the database (it would use an own table in the database the one the user uses already anyway, so could be Postgres, SQLite, MySQL, …)
Create a node which allows to send in items, define a unique parameter (like id, date, …) and this node would then use what got created in 1, to see if that item got processed already earlier. If it did, it gets filtered out. If not, it passes through.
Btw. till that is in place you can use static-data. Here an example for actually exactly the same use-case:
It has not been reviewed yet, We have recently released the Community Node Repository feature though which will allow nodes to be created, published and installed without needing to wait for us to review the pull request.
There is a comment on the PR that explains why it was closed. We closed a lot of new node PRs when we launched the community node feature this allows node creators to publish nodes that can be used with self hosted instances quicker.
Just because a PR is closed it doesn’t mean that we won’t reopen it again in the future if it becomes a priority either.