Hacker News Reader

Comment by mrob | Hacker News Reader

The biggest problem I have with infinite scroll is that it breaks searching within a page. If I want to find something I have to scroll down to the bottom and then wait some arbitrary time until I think enough content has loaded. Even if I know roughly where the content I'm searching for is, I can't just click to a specific page. I have to guess how far I've scrolled and hope that it's far enough. I waste time and data transfer because I'm forced to overshoot if I want to reliably find things. I can't even bookmark a specific page, so I have to repeat this tedious process every time.

I've never seen a use of infinite scrolling that wouldn't be better replaced with lazy loading of images/video.

View on HN

If the information you're searching for would have been in the next page, searching within the page would have been similarly "broken" in a paginated UI.

The alternative to infinite scrolling isn't to load infinite data up front.

If you are presented with pagination, you have a hint that there's more content to load and how much more, 2 pages or 100 pages? I know immediately that I have to repeat my search and how many times. You aren't "warned" of infinite scrolling upfront, you don't know there's unloaded content unless you scroll down. Nor do you have any idea when it will end. It's mystery meat navigation.

But if I found that content via a search engine I would have been directed to the correct page to find what I was looking for.

I've had many occasions of clicking links from G to land on a page that should have what I want, the cached copy did but the landed page doesn't

That would be true of any site that updates its webpages at any time, regardless of static or dynamic webpages, and certainly regardless of pagination/infinite load.

The only difference really is that infinite load almost guarantees the page won’t have what you want on it, when you visit it.

Yep, exactly.

And with my first problem, I can use the Next buttom and Ctrl+F (features others on this thread mentioned).

Infinite load is not the right answer everywhere (anywhere?)

Eg: What if StackOverflow adopted it?

Pagination at least gives a hint as to what is loaded and if i need to repeat search again on next page. This is harder when you don't know what is actually loaded without scrolling.

>The alternative to infinite scrolling isn't to load infinite data up front.

Lazy loading implies the opposite? Regardless, it being broken up is very useful for everything around the site usage. IE google-linking, sharing urls, bookmarking, etc.

Infinite scroll is an optimization for first-time usage of a site, or for pages that are naturally read-once (ie twitter feed). They break everything surrounding n-use of the page.

The URL should change with either offset (which is dynamic) or ideally ID as cursor (which is static).

Wouldn't this break the back button? I wouldn't expect to have to go back several times just because I scrolled.

You just use history.replaceState

Isn't this better solved with a combination of bookmarking, tagging, good search, and resetting the person to the same spot on a page when it reloads? (facebook doesn't do the latter for instance, and it's massively frustrating to me)

The bit on Facebook is purposeful. Youtube does the same thing. It's about keeping the user engaged with another novel engagement whether ad or content driven. You may be frustrated and possibly rage quit, but the majority fall for it cyclically and spend more time on the service as a result.

As mentioned elsewhere in this thread: "As someone with ADHD, I immediately close any website that uses infinite scroll. I will become insanely focused on that site to the point that I tune out the outside world. What begins as "oh, I'll just look at this one article" turns into at least a half hour of wasted time before I'll be able to break out of it."

None of that will be as fast and simple as pressing Control-f and typing the search term. You don't even need to press enter. Searching within the page is a superbly useful and well-designed interface, and I would be very surprised if any custom web app solution could do better.

The alternative to infinite scrolling isn't everything on one page though, it's paging.

If only they had regular expression support in the built in find. A few times I've put that into web apps and found it quite useful.

I bet there's a browser extension for that.

Lazy loading can also break search within page.

Now you have two problems.

You should include same-sized placeholders for the unloaded content, so the page doesn't jump about when you search and push the search result off-screen.

Reply on news.ycombinator.com