@dfox, How does the 'algo' work ?

That's called proprietary code

I can't tell, it's top secret.

Probably @theNSA knows this.

@theNSA wow, It will take me 250 years to develop this advanced code.

@theNSA you got an error in your code there. Assuming a rant is an object, you forgot a '.' or if it's just a function that returns True, you forgot a '_'.

@theNSA You can’t just share private info like that. 😠
dfox won’t be happy.

That's the easy part
The magic happens inside isGood()
;)

📌

I think it just shows @linuxxx rants. The algo page is filled with them, for me at least 😄

David could've had made it up just to fuck with us

@dfox

Tags dont work in the original post

@Froot @linuxxx
if(rant.author == "linuxxx") {
rant.show();
}

@Hedgepig Meant to be that way!

@LinuxUser0001😊

@linuxxx but why?

@linuxxx sure, I was just tagging him

There’s a bunch of criteria and two types of content mixed in:

1. New rants (rants that have a few votes but are pretty new). These are sprinkled in.

2. Recommended rants from the last 45 days. These are the bulk of what you see. For any rant to be recommended (or new for that matter) it means you never went into it before and you never scrolled past it before, and it also means you didn’t have it get loaded a few times in the feed already. So simply put, you haven’t seen it yet.

Then, to decide if a rant is recommended and how highly it’s recommended to you (how high it should show up in your feed, we look at:
- how you like the tags
- how you like the author
- collaborative filtering - how do users you have similar tastes to like that rant you haven’t seen

I hope that clears it up a bit. Let me know if you have any other questions.

I’m also trying to find time to make algo more advanced and try to get some better results.

You have included AI in this to filter things out? @dfox

@dfox wow, good job, cheers for sharing!

Do you do any machine learning?

Edit: whoops didn't see the above question which is basically the same

@Hedgepig 😉