Recent HN post: https://news.ycombinator.com/item?id=32495846
… and honestly what does “page 2” even mean on a site like this? It’s almost a completely arbitrary distinction because the amount of items on each page is just a magic number. And your “page 2” is nothing more than snapshot in time… sharing a URL for “page 2” to somebody else is pointless because they won’t see the same content you did. Bookmarking “page 2” is pointless since it will change almost every load. Having a search engine index “page two” is pointless because “page two” won’t even say the same stuff by the time it gets exposed in the search page (you see this sometimes with SEO results that link to a “stories” page). In fact you might as well forbid search engines from indexing anything after “page 1”
And once you start thinking about it you might just say “fuck it… everything is the same ‘page’”. Instead of a “page” you have a viewport into a infinite, constantly changing pane of content. But since it is impractical to load an infinite list of items at once, you’ve got to load chunks of it on demand. And suddenly you’ve invented infinite scroll.
Again, details matter here, but it is easily solvable.
Or wants to send a link right now to be opened right now ("Hey honey, which one of these things should I get?").
Links are inherently ephemeral: content changes and expires, and there's nothing that guarantees against that. Doesn't make them useless.
That's the usual default, but there are usually sort options along with pagination that can make an items place in a list more or less fixed.
Paginated pages are about as linkable as properly implemented infinite scrolling.
Maybe if you're linking to the page with the dragonfruit on it and a bunch of items get added between that and the banana it might be off by a page, but it'll be close enough to be useful.
Like:
exampe.com/products#offset-999