> IMHO not much really needs to change, I think how reposts and such are handled right now is great.

This is where I am at. The notion that HN posts die off relatively quickly (i.e. within ~48h) is potentially a very important part of what gives them the quality they have today.

If every post was eternal and duplicates were always closed, you'd wind up with effectively perma-stickied comments in these god threads that become the law for that topic until the end of time.

Exactly, it's a huge chore to keep those kind of mega threads up to date and relevant too. And you get a lot of bikeshedding on other topics with people complaing that this should be in the mega thread, etc.

Not if you sort them upside down, newest first. Old comments increasingly pushed away to deeper pages, becoming an archive.

You could have an AI to summarize the last N comments, giving an overview on the most recent opinion.

Thanks, that's helpful. The one missing piece is a good way of threading the beads together.

I do this informally: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que..., but that's all just prep for one day fitting it nicely into the software.

It's super tricky because how do you decide what constitutes a 'topic'?

For example, are all the various ChatGPT threads related? What about coverage of an event (e.g. there are six or seven threads regarding the recent earthquake in Turkiye—all empty apart from https://news.ycombinator.com/item?id=34672314 —however I've seen it where basically the same story is being carried across several outlets and discussions are on the same topics, just one post gets more traction than the other).

It's easy enough to link articles which refer to the same exact source and mark duplicates, but harder to link posts which rely on different sources to cover the exact same topic.

I agree!