8 ms·

Dear God I hate this anti-pagination nonsense. I loathe infinite scroll. It puts you at the mercy of the site doing browser caching right (no one does) otherwi

by cletus 15y ago

Dear God I hate this anti-pagination nonsense. I loathe infinite scroll. It puts you at the mercy of the site doing browser caching right (no one does) otherwise you click on a link, click back and then have to start from the beginning again. Even Apple gets this wrong (you'll reach a point going through the Top Charts in the App Store where pressing Back reduces you back to the first 10 results). Pages can be bookmarked easily, which can be incredibly important. Paging is also easier for crawling (otherwise it requires some JS evaluation or other hackery). DZone is a good example of a site that annoys the crap out of me with infinite scroll. Seriously, cut that shit out.

jonursenbach 15y ago
It's especially annoying when people have site footers with important links, but infinite scroll just above it, rendering it impossible to click on any link in said footer. Drives me fucking insane.
- funkah 15y ago
  Yeah, the only way to do that right would be to give the infinite scroll a container with a fixed height, and drive the infinite scroll based on the position of the container rather than the entire page.
- citricsquid 15y ago
  > Drives me fucking insane. I share exactly the same feelings. Want to create a Facebook page: http://i.imgur.com/vZ5LD.png http://i.imgur.com/vZ5LD.png good luck! It's like the game where you have to click a button but whenever you hover over the button it moves.
- fpgeek 15y ago
  Now try getting to that footer on your smartphone (because you wanted some feature of the desktop site that wasn't in the mobile version, for example).
lukifer 15y ago
There are occasionally use cases for infinite scroll, but I agree that most of the time, it's a design anti-pattern. My preferred solution to pagination: by default, load a large number of results per page. The markup overhead is minimal with gzip, database hits can be mitigated with caching, and images can be lazy-loaded if necessary. Best of both worlds, and friendlier to users and search engines alike. Also, there's no reason pagination can't work inline (ie, link to "Load 10 More Results"), which also gives you the best of both.
gcb 15y ago
Even more fun on mobile. Load 10 scrollfuls of content, and your phone halt to a crawl
joe_the_user 15y ago
I hate pro-infinite scroll nonsense. But the argument that your "page 867" means nothing remains. What's the solution? I'm not sure and it probably depends on the application involved. One idea in search might be to return fifty items and give ten suggested refinements by key word or otherwise.
- swah 15y ago
  I think its obvious... just limit the results to a sensible number.. perhaps randomize the last half...
- GreenNight 15y ago
  "page 867" means nothing, but so does scrolling down for half an hour. And the first one is faster.
azov 15y ago
Indeed. Pagination is one of those things not broken. Please don't fix it!
seiji 15y ago
Pages can be bookmarked easily, which can be incredibly important. Pages can be bookmarked easily only sometimes! Everybody seems to have adopted the same database-centric but user-hostile (and cache/crawler-hostile) approach to pagination of posts (see: any blog platform, engadget, tumblr, etc). Their brain damaged flow goes: You make a new post. The new post appears at the top of the front page. But, the front page has a limited number of story slots. So, the oldest story on the front page gets pushed to page two. But page two has limited slots, so the oldest story on page two goes to page three. Repeat for every page. Some sites have tens to hundreds of thousands of "pages." (Preemptive note to future comment haters: I know the site doesn't "push" articles to every page because it's all just database queries, but the effect is the same.) Every time you make a new post, you invalidate all thousand (ten thousand? 100k? million?) older pages. It's absolutely moronic. Google has no chance of keeping up. You make one new post and Google has to re-index thousands of pages. The proper solution is to make pages "fill up" then have your root page point to the highest numbered page. Example: http://omgpagination.tumblr.com/ http://omgpagination.tumblr.com/ would logically be page 600, then the "Older Posts" button would link to page 599. Instead, everybody right now makes / always be page 1, "Older Posts" always links to page 2, etc. But those URLs aren't content stable and every update invalidates your cache of the entire site. Your oldest posts on your site should always be on page 1, not a moving target of page {TOTAL_POSTS / POSTS_PER_PAGE}. In short: be a little more clever and do the right thing for your caches. What's good for your caches is inevitably good for your users too.
- gry 15y ago
  Pagination is useful when a user needs to understand time and space for a given context. Infinite scroll is useful when the time and space is now for the context. As noted by both parents, implementations for both could be better.
  - seiji 15y ago
    You are quite right. Time and space is the important point here. There's no way of saying "Give me page 6 as it was when I bookmarked it two weeks ago."
jQueryIsAwesome 15y ago
I think that infinite scroll should have infinite scroll in the upperside as well, meaning that the posts on the top should deleted as you scroll down. That would also make viable creating links and bookmarking based on your current position.
- spullara 15y ago
  Yep, this is the only solution in combination with timestamps rather than page numbers.
mjwalshe 15y ago
Yep you wouldnt belive how much work on a big site doing pagination and faceting (pagnations big brother)an doing it RIGHT! (wel in my case getting developers to go back and fix stuff they broke/messed up the first time) Google thinks pagnation is important else why would they produce and suport the new rel tags prev and next.
guynamedloren 15y ago
> Even Apple gets this wrong (you'll reach a point going through the Top Charts in the App Store where pressing Back reduces you back to the first 10 results) Apple gets a lot wrong. Personally, I think their website is horribly difficult to navigate - specifically their web store and web-based app store. The site is chock full of beautiful graphics and plenty of information, but I can't believe how difficult it is to buy the damn thing once I've made up my mind. The call-to-action buttons (ie "BUY THIS RIGHT NOW") are tiny, nearly invisible, or nonexistent. Not what you'd expect from the tech powerhouse that is Apple. I'd love to do a writeup on this one day. I feel like I'm crazy finding the website so difficult to navigate, but there's no way I'm alone here.
- jiggy2011 15y ago
  I think that's partly on purpose, it's sort of the same in their brick & mortar stores too. There's lots of gadgets and fancy stuff to play with but no obvious sales funnel. I think it's because if you decide your going to buy a mac then your going to buy a mac and apple get a cut regardless of where you buy it so there's no incentive to have to convert you right that second before you walk out the door. The entire apple website is basically just computer porn, it's designed to create a lasting emotional impact rather than to feel like a shop.
  - guynamedloren 15y ago
    This is powerful stuff. I hate to admit it because it goes against almost all conventional ecommerce thinking, but I believe you're right.
firefoxman1 15y ago
Yeah, his forum argument made a good point, but what about 200-page threads? How am I supposed to get to page 200 without scrolling?
dybber 15y ago
Facebook seems to have found a way to do this the right way in their "profile timelines": * The right side contains a scrollbar annotated with years and months, so you can efficiently navigate to posts from a particular period. * You can deep link to a specific months: https://www.facebook.com/<profile>/timeline/2008/11 https://www.facebook.com/<profile>/timeline/2008/11 * The center line visualizes that the feed continues and that the user can keep scrolling (until they reach their birth) * Displays a loading bar while new posts are fetched * No footer and a static header. Two problems: * Forward/backward buttons doesn't work unfortunately. * Webspiders does not make sense for such a closed system, so I'm not sure how they would be able to index such a site. --- I would really love if other sites would adapt to this scheme of numbering pages by their year and month. Try navigating through which commits where introduced a year ago in a Github repository: https://github.com/mirrors/linux-2.6/commits/master https://github.com/mirrors/linux-2.6/commits/master - it ends in a binary search through page numbers! (You could use the CLI, but that is not the point here. Say you need to link to a particular line in an old commit)
SideburnsOfDoom 15y ago
The thing I don't like is that when you have both scrollbars and pagination, there are two very different ways of getting more content onto the screen. That's one to many. It doesn't matter that they reflect different things - client-side display of already-retrieved content and interaction with a server for new content. That's a detail that a user of software should not have to react to. Infinite scroll is one response, maybe there are others?
Shorel 15y ago
I love scrolling and hate all that extra pagination. My browser handles correctly what you just described while browsing huge pages, except may be the bookmarks. Or may be we simply visit different sites, I love to read books and wikis online. I won't say that you change your browser, as that path leads to endless religious wars, but bear in mind that there's people like me with the exact opposite mindset to yours.
ChrisNorstrom 15y ago
http://www.chrisnorstrom.com/2011/04/invention-infinite-scroll-pagination-hybrid-ui/ http://www.chrisnorstrom.com/2011/04/invention-infinite-scro... I feel your pain. So I came up with a hybrid pagination + infinite scroll solution last year. I can't implement it myself but I'd love to see someone try it. Basically it puts in pagination links in between the content loads so you can easily skip ahead or go far back without having to scroll your fingers off.

by cletus 15y ago

jonursenbach 15y ago

It's especially annoying when people have site footers with important links, but infinite scroll just above it, rendering it impossible to click on any link in said footer. Drives me fucking insane.

funkah 15y ago
Yeah, the only way to do that right would be to give the infinite scroll a container with a fixed height, and drive the infinite scroll based on the position of the container rather than the entire page.
citricsquid 15y ago
> Drives me fucking insane. I share exactly the same feelings. Want to create a Facebook page: http://i.imgur.com/vZ5LD.png http://i.imgur.com/vZ5LD.png good luck! It's like the game where you have to click a button but whenever you hover over the button it moves.
fpgeek 15y ago
Now try getting to that footer on your smartphone (because you wanted some feature of the desktop site that wasn't in the mobile version, for example).

lukifer 15y ago

There are occasionally use cases for infinite scroll, but I agree that most of the time, it's a design anti-pattern. My preferred solution to pagination: by default, load a large number of results per page. The markup overhead is minimal with gzip, database hits can be mitigated with caching, and images can be lazy-loaded if necessary. Best of both worlds, and friendlier to users and search engines alike. Also, there's no reason pagination can't work inline (ie, link to "Load 10 More Results"), which also gives you the best of both.

gcb 15y ago

Even more fun on mobile. Load 10 scrollfuls of content, and your phone halt to a crawl

joe_the_user 15y ago

I hate pro-infinite scroll nonsense. But the argument that your "page 867" means nothing remains. What's the solution? I'm not sure and it probably depends on the application involved. One idea in search might be to return fifty items and give ten suggested refinements by key word or otherwise.

swah 15y ago
I think its obvious... just limit the results to a sensible number.. perhaps randomize the last half...
GreenNight 15y ago
"page 867" means nothing, but so does scrolling down for half an hour. And the first one is faster.

azov 15y ago

Indeed. Pagination is one of those things not broken. Please don't fix it!

seiji 15y ago

Pages can be bookmarked easily, which can be incredibly important. Pages can be bookmarked easily only sometimes! Everybody seems to have adopted the same database-centric but user-hostile (and cache/crawler-hostile) approach to pagination of posts (see: any blog platform, engadget, tumblr, etc). Their brain damaged flow goes: You make a new post. The new post appears at the top of the front page. But, the front page has a limited number of story slots. So, the oldest story on the front page gets pushed to page two. But page two has limited slots, so the oldest story on page two goes to page three. Repeat for every page. Some sites have tens to hundreds of thousands of "pages." (Preemptive note to future comment haters: I know the site doesn't "push" articles to every page because it's all just database queries, but the effect is the same.) Every time you make a new post, you invalidate all thousand (ten thousand? 100k? million?) older pages. It's absolutely moronic. Google has no chance of keeping up. You make one new post and Google has to re-index thousands of pages. The proper solution is to make pages "fill up" then have your root page point to the highest numbered page. Example: http://omgpagination.tumblr.com/ http://omgpagination.tumblr.com/ would logically be page 600, then the "Older Posts" button would link to page 599. Instead, everybody right now makes / always be page 1, "Older Posts" always links to page 2, etc. But those URLs aren't content stable and every update invalidates your cache of the entire site. Your oldest posts on your site should always be on page 1, not a moving target of page {TOTAL_POSTS / POSTS_PER_PAGE}. In short: be a little more clever and do the right thing for your caches. What's good for your caches is inevitably good for your users too.

gry 15y ago
Pagination is useful when a user needs to understand time and space for a given context. Infinite scroll is useful when the time and space is now for the context. As noted by both parents, implementations for both could be better.
- seiji 15y ago
  You are quite right. Time and space is the important point here. There's no way of saying "Give me page 6 as it was when I bookmarked it two weeks ago."

jQueryIsAwesome 15y ago

I think that infinite scroll should have infinite scroll in the upperside as well, meaning that the posts on the top should deleted as you scroll down. That would also make viable creating links and bookmarking based on your current position.

spullara 15y ago
Yep, this is the only solution in combination with timestamps rather than page numbers.

mjwalshe 15y ago

Yep you wouldnt belive how much work on a big site doing pagination and faceting (pagnations big brother)an doing it RIGHT! (wel in my case getting developers to go back and fix stuff they broke/messed up the first time) Google thinks pagnation is important else why would they produce and suport the new rel tags prev and next.

guynamedloren 15y ago

> Even Apple gets this wrong (you'll reach a point going through the Top Charts in the App Store where pressing Back reduces you back to the first 10 results) Apple gets a lot wrong. Personally, I think their website is horribly difficult to navigate - specifically their web store and web-based app store. The site is chock full of beautiful graphics and plenty of information, but I can't believe how difficult it is to buy the damn thing once I've made up my mind. The call-to-action buttons (ie "BUY THIS RIGHT NOW") are tiny, nearly invisible, or nonexistent. Not what you'd expect from the tech powerhouse that is Apple. I'd love to do a writeup on this one day. I feel like I'm crazy finding the website so difficult to navigate, but there's no way I'm alone here.

jiggy2011 15y ago
I think that's partly on purpose, it's sort of the same in their brick & mortar stores too. There's lots of gadgets and fancy stuff to play with but no obvious sales funnel. I think it's because if you decide your going to buy a mac then your going to buy a mac and apple get a cut regardless of where you buy it so there's no incentive to have to convert you right that second before you walk out the door. The entire apple website is basically just computer porn, it's designed to create a lasting emotional impact rather than to feel like a shop.
- guynamedloren 15y ago
  This is powerful stuff. I hate to admit it because it goes against almost all conventional ecommerce thinking, but I believe you're right.

firefoxman1 15y ago

Yeah, his forum argument made a good point, but what about 200-page threads? How am I supposed to get to page 200 without scrolling?

dybber 15y ago

Facebook seems to have found a way to do this the right way in their "profile timelines": * The right side contains a scrollbar annotated with years and months, so you can efficiently navigate to posts from a particular period. * You can deep link to a specific months: https://www.facebook.com/<profile>/timeline/2008/11 https://www.facebook.com/<profile>/timeline/2008/11 * The center line visualizes that the feed continues and that the user can keep scrolling (until they reach their birth) * Displays a loading bar while new posts are fetched * No footer and a static header. Two problems: * Forward/backward buttons doesn't work unfortunately. * Webspiders does not make sense for such a closed system, so I'm not sure how they would be able to index such a site. --- I would really love if other sites would adapt to this scheme of numbering pages by their year and month. Try navigating through which commits where introduced a year ago in a Github repository: https://github.com/mirrors/linux-2.6/commits/master https://github.com/mirrors/linux-2.6/commits/master - it ends in a binary search through page numbers! (You could use the CLI, but that is not the point here. Say you need to link to a particular line in an old commit)

SideburnsOfDoom 15y ago

The thing I don't like is that when you have both scrollbars and pagination, there are two very different ways of getting more content onto the screen. That's one to many. It doesn't matter that they reflect different things - client-side display of already-retrieved content and interaction with a server for new content. That's a detail that a user of software should not have to react to. Infinite scroll is one response, maybe there are others?

Shorel 15y ago

I love scrolling and hate all that extra pagination. My browser handles correctly what you just described while browsing huge pages, except may be the bookmarks. Or may be we simply visit different sites, I love to read books and wikis online. I won't say that you change your browser, as that path leads to endless religious wars, but bear in mind that there's people like me with the exact opposite mindset to yours.

ChrisNorstrom 15y ago

http://www.chrisnorstrom.com/2011/04/invention-infinite-scroll-pagination-hybrid-ui/ http://www.chrisnorstrom.com/2011/04/invention-infinite-scro... I feel your pain. So I came up with a hybrid pagination + infinite scroll solution last year. I can't implement it myself but I'd love to see someone try it. Basically it puts in pagination links in between the content loads so you can easily skip ahead or go far back without having to scroll your fingers off.