Hacker News Reader

I just use real URLs that resolve when you navigate back to them... https://www.zadieheimlich.com (terrible use case as clicking a heading takes you to the same thing you see currently on the homepage but you get the idea.)

View on HN

I pondered doing something like that, however there's a performance impact with that version right? So if I want to let a user scroll for 5 pages (30 records each), click on something, and then go back to a URL that "resolves". That url is going to have to render all 5 pages of stuff to get the look right. That might not be that big of a deal for a bunch of apps, but for us had a cost I didn't want to pay. So keeping it all cached on the client and just picking up scrolling where the user had left off seemed like a good solution right now.

If you had a publicly visible site that you wanted to be crawled, then having a real url would be pretty important.

But for something like Highrise, where it's dark to google and crawlers, your solution makes sense.

Reply on news.ycombinator.com