Hacker News Reader

Comment by raldi | Hacker News Reader

It's also a lie because "everything" on reddit is still not in fact searchable; buried in the comments on the announcement post, they admit that pre-2020 comments are still not indexed:

https://www.reddit.com/r/reddit/comments/u3oz2x/whats_up_wit...

View on HN

LOL!

Well time to continue using "site:www.reddit.com" on Google, then. I find the old search from old.reddit.com to be way better than the newer search (but I haven't tried this newest search linked in the thread).

Interesting - I did not even know that old.reddit used a different search. I rarely even use the search (occasionally I will use it to search a specific subreddit and assume it is only searching post titles). I pretty much exclusively use old.reddit + customized reddit enhancement suite and on mobile use the app Reddit is Fun. Anytime I have to go back to vanilla new reddit it feels like a totally different site. I'm not even really changing the default subscriptions all that much either at all - often I browse without being signed in. The part I find the most odd is that even without signing in the "hot" algo seems to organize top posts occasionally differently between the two (old and new).

LOL This is ridiculous. The majority of Reddit is Pre-2020 !

Or as I like to call it - The Before Covid times. BC. Surely there's no other acronym that uses those two letters. 2018 BC was a good year for me personally

Wouldn't that be 1BC?

This is likely intentional though.

Reddit's engagement model relies on new posts with which to comment on. Being able to easily find previous content means less reposting, and less opportunities to give users a place to engage. A standard, working search function would only hurt reddit's growth metrics.

It's so obvious to the userbase that you have certain users complaining about reposts, or even subreddits that intentionally flag for original content (OC).

In other words, the most valuable dataset (the majority of the data) is not yet searchable.

I expect post-2020 is the majority of the data given Reddit's exponential growth. But definitely pre-2020 is more valuable.

Any data to suggest that the past 2 years have generated the majority of reddit's data? I've been a user for a decade and it doesn't feel like this is the case to me.

Hi Raldi, when did last work with Reddit. I can't remember, genuinely curious.

2008-2011

Reply on news.ycombinator.com