What's the difference?

It's the same algorithm, but it only considers upvotes by users who registered before (checking the code...) December 13, 2008. That's astonishingly long ago—only (checking the data...) 1.5% of HN accounts existed back then. Yet the "classic" frontpage is not that different from the main frontpage. This was the main conclusion when pg launched the feature: https://news.ycombinator.com/item?id=607271, and again in 2011: https://news.ycombinator.com/item?id=2073513.

What that tells me is that the forces creating the HN front page don't have much to do with changes in the userbase over time. That's a major finding, and I think to most people (me included, and pg probably included) a counterintuitive one.

It's an interesting result - it only counts upvotes by users from before December 2008 who are still active on the site

After all, if you changed HN into a My Little Pony discussion site you'd lose 98% of old users, retaining the 2% who like ponies - and thereafter, 100% of remaining old users would like ponies.

That's a good point. But there are a lot of users from 2008 still active on the site. It would be interesting to see how the attrition rate has changed over time, but I've never looked into that.

This is an excellent find!

My memory is hazy but I probably found out about HN via Slashdot or via Michael Arrington's HN post [0] from Mar. 10, 2008; so I have been reading HN since before Dec. 13, 2008 and still come back because of some really good conversations that can be had here, compared to elsewhere on the Internet.

0: https://techcrunch.com/2008/03/10/little-known-hacker-news-i...

The stories seem a bit different - maybe it's an older rating algorithm?