As a developer I hoard many links and also would like to get notifications on different tasks being finished (script/RSS/IoT) so I've built a chat web app for myself where you can create a channel for each specific notification or links to read later.
The web app is a PWA so I can install it on my Android Mac and PC via Chrome browser without app/play store and access from anywhere.
I took it further by adding Webhoom API for the notification and a Chrome extension to share links from my browser to the channels.
People liked it (didn't try to market so much) but they wanted encryption so I have added end to end encryption as well.
I'm currently working on (on my spare time) big files support up to 4GB from any browser including live preview and video streaming for well known file types.
This reminds me Yahoo! Answers clones 10 years+ ago. To get traffic to website and cheat the search engine they would have index the Yahoo! answers website for specific niche category and create a garbage website with questions and answers not crediting the source and cramp the website with Ads everywhere to earn massive revenue.
In 2017 I built Twitark - https://github.com/pantchox/Twitark
Its a Node.js CLI tool that connects to Twitter Stream API to archive tweets and trends (all countries)
You can connect to Twitter stream API which is about %1 of whole twitter (was mentioned in their forums) and you can connect to Twitter Filter API with predefined hashtag monitoring or users monitoring and get all these related tweets.
Twitark creates daily rotated zip archives for each day structured in folders where each minute of tweets is set is a JSON file.
An archived day of Twitter stream API is about ~2GB per day.
I use this to analyze big data and trends with Twitter, also for past data.
So far it is working very well and I've been working on database adapters such as Postgres, Sqlite which are on different branch (not yet committed to master) so you can write directly to your database instead of zip files.
The second open source project I released is actually being used in Twitark.
It called Joebear (play of words of Jobber) https://github.com/pantchox/joebear
Its a Node.js serial queue manager, using old school Node.js event emitter.
The goal of Joebear is to receive a list of jobs and run one by one, if one fails there is ability to reset or stop at any given time.
In Twitark it is being used to monitor the Twitter Trends API.
I still hope more people will make use of my open source projects, I really invested time to write the documentation as well, but I don't have many stars on Github.
I love SQLite, I do all of my projects prototyping before I move to postgres which is also awesome.
One if the major problems in facing is that heroku for example in every build will override/delete the SQLite DB since it treats it the same as every file - which is a shame.