Daily has gone open source

Ido Shamun on July 08, 2019

We have shared some hints and clues along the way but it's time to announce that Daily is now open source! 🎉 dailyn... [Read Full]
markdown guide

I installed the extension right after reading this post. Awesome work!

I do have some questions for you as a business focus most on curating content:

  1. How do you curate and tagging content? Look up for keywords automatically? I did see an entry on Daily having so much tags that its title gone invisible.
  2. What do you plan to optimize personal experiences? From my experience, all curated news sources don't work: Medium, DEV feed, Pocket discover, newsletter digest, blog news feeds, subreddits... For 10 articles in a news feed, less than 1 worth my glance, and there's a small percent that worth saving to reading list. I'd like to specify an expectation like this:

Hey I definitely want to hear about:

  • Article targeting novice coder, explaining to them confusing terms with different meanings in different (non-)programming cultures.
  • Release annoucement of version 1.0 of the projects I'm interested in (don't notify about verions 0.x)
  • Particular programming subject: Async in Python...
  • ...

Thanks for reading my rant :)


First of all, thank you for dedicating the time using the extension and commenting here. Currently we curate content by hand picking the sources and we fetch tags from the blog post tags if exist or meta keywords. I am familiar with the issue you raised with the tags and I need to fix it by limiting the maximum amount of tags per post.
Regarding the personalization, we are looking to integrate a recommendation system and also other sources of content, not just blog posts. This way you will truly be able to get all the content you want and need.

Hopefully it answers your question :)


limiting the maximum amount of tags per post

I think you shouldn't do that if you ever want to provide searching feature. Just pick (some of) the followed tags to show up and display a More... button at last.

Btw, ideally this is the way I want to set up expectations by tags:

#explain #newbie

#release #production any(#guix #entropic #redox)

#async #python

#webdev #python

#security any(#nodejs #python)

Your list of tags should be curated and organized to ensure no duplicate. Then comes a suggesting system to help user type the right tag name.

It is very hard to curate the tags as well. When you curate content from multiple sources, each source has its own "native" tags and some tags such as release don't even exist there. It will require a sophisticated system for automatically tagging the content. Unless I'm missing something and you have an idea :P

Your tags is under your decision. Just make sure they're unique, e.g. if you decided to use #async and #python, then you would translate "native" #asyncio to them.

Of course I will miss some news, it always happens anyway. And it doesn't matter! They're news I want to know, not what I need. If I need, I will seek them. If I'm in freetime, I will surf the source directly.

So in my opinion, you just have to keep all things relevant, not all relevant things.

Thank you for pointing it out :)
We still have a lot of work on curating correctly the content and as you mention keep all things relevant (or at least most of them). Hopefully, now that we are open source the community will also help us to accomplish it.


Great news!
What about a Safari extension, or an Electron app?


We have an open issue about the Safari extension:

We are looking for contribution :)

code of conduct - report abuse