Continuous Scroll

30 Jan 2008 - 6:44am
6 years ago
27 replies
3758 reads
Etkin Ciftci
2007

As supported in the Sympony Open-Source PHP Web Framework, "Ajax Pagination"
allows the client browser to load the content gradually as the user scrolls
down to the bottom of the web page. Once the vertical scrollbar reaches the
bottom it triggers a new page to load. Unlike its common use in separate
frames(i.e. the inbox of Hotmail Live), ajax pagination(weird name) attaches
a new full page to the end of the current one. So, we end up with a
continuous scroll bar and therefore a continuous content.

http://www.symfony-project.org/demo/pager.html

I am very excited about the possible usages of this method. It may totally
revolutionize the way we interact with web browsers. No more "previous
entries" buttons(i.e. techcrunch), no more page 1.2.3 links (i.e. google
results page)...the method proposes an instant continuity upon request for
the long web content.

On the other hand ajax pagination assumes "if the user scrolled down to the
bottom of a page, he should be asking for more." Do you agree with such an
assumption?

Etkin Ciftci
Interaction Designer
Bilende, Istanbul

Comments

30 Jan 2008 - 8:24am
jonesabi
2006

V. Interesting ajax implementation.

Right now, when I scroll down to the bottom of a page it also means
that I might be looking for information I assume is in the footer
(company contact info, address, help links).

Also, what does it mean for CTRL F style page searches?

Thanks for giving the list a lot to think about :)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

30 Jan 2008 - 8:17am
ELISABETH HUBERT
2007

I liked the demo. It didn't jump around at all and was actually
fairly elegant. However, I'm not sure that I agree that the user is
looking for more content if they scroll to the bottom of the page.
I'm assuming this is another case of "it depends". Depending on
the type of site I'm on and what type of content I'm reading about
or looking at I may be waiting for more. That being said in this case
(and I'm wondering if in other cases) it didn't hurt to have to
extra content appear. Thoughts?

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

30 Jan 2008 - 8:31am
Mark Schraad
2006

This is cool. The only downside is that users often use the scroll bar as an indicator of the longth of the page. This pretty much skews what is an important bit of feedback to the user.

Mark

On Wednesday, January 30, 2008, at 09:05AM, "Etkin Ciftci" <etkin at etkin.info> wrote:
>As supported in the Sympony Open-Source PHP Web Framework, "Ajax Pagination"
>allows the client browser to load the content gradually as the user scrolls
>down to the bottom of the web page. Once the vertical scrollbar reaches the
>bottom it triggers a new page to load. Unlike its common use in separate
>frames(i.e. the inbox of Hotmail Live), ajax pagination(weird name) attaches
>a new full page to the end of the current one. So, we end up with a
>continuous scroll bar and therefore a continuous content.
>
>http://www.symfony-project.org/demo/pager.html
>
>I am very excited about the possible usages of this method. It may totally
>revolutionize the way we interact with web browsers. No more "previous
>entries" buttons(i.e. techcrunch), no more page 1.2.3 links (i.e. google
>results page)...the method proposes an instant continuity upon request for
>the long web content.
>
>On the other hand ajax pagination assumes "if the user scrolled down to the
>bottom of a page, he should be asking for more." Do you agree with such an
>assumption?
>
>Etkin Ciftci
>Interaction Designer
>Bilende, Istanbul
>________________________________________________________________
>*Come to IxDA Interaction08 | Savannah*
>February 8-10, 2008 in Savannah, GA, USA
>Register today: http://interaction08.ixda.org/
>
>________________________________________________________________
>Welcome to the Interaction Design Association (IxDA)!
>To post to this list ....... discuss at ixda.org
>Unsubscribe ................ http://www.ixda.org/unsubscribe
>List Guidelines ............ http://www.ixda.org/guidelines
>List Help .................. http://www.ixda.org/help
>
>

30 Jan 2008 - 9:41am
Meredith Noble
2010

> Once the vertical scrollbar reaches the
> bottom it triggers a new page to load. Unlike its common use in
separate
> frames(i.e. the inbox of Hotmail Live), ajax pagination(weird name)
> attaches a new full page to the end of the current one. So, we end up
with
> a continuous scroll bar and therefore a continuous content.

Thanks for bringing this up again, Etkin.

Three real-world examples I've found are:

The Humanized Reader
http://humanized.com/reader/

Google Reader (I use this daily and love it, for the most part)
http://reader.google.com

A9
http://a9.com/

I have been referring to it as "infinite scroll" -- not sure whether I
made that term up or if I heard it somewhere.

I'm interested to know what other good examples of this people have
found.

Overall, I agree that its value entirely depends on the user's
expectations for the page. It works well for me for RSS viewing because
I have no desire to do anything other than keep reading when I get to
the bottom of a page of articles. The trickiness comes in with other
features -- for example, Google Reader has to guess that if I've
scrolled past something, that means I've "read" it and its state should
therefore be changed, so I won't see that article again. This isn't
always necessarily a safe assumption, and it can be quite irksome at
times.

Meredith

-----------------------------------------------------------------------
Meredith Noble
Information Architect, Usability Matters Inc.
416-598-7770, ext. 6
meredith at usabilitymatters.com

30 Jan 2008 - 10:30am
Jeff Seager
2007

I agree with both Abi and Mark about this. I'm impressed that it
remains accessible without scripts, and I like that a lot.

I think you'd want to use this very selectively. In most cases,
it's an advantage to me that my screen contains a finite amount of
information I can sort through and use. My specific usage exception
might be for archived material in a wiki, news site or the like --
cases in which you're unlikely to want anything at the very bottom.

No, maybe not even for that. Try this: reload the page (with
javascript enabled), and use your browser's search function to find
"PHP Throwdown" (which is several "pages" down). You won't find
it because it isn't loaded yet, and it won't be loaded until you
scroll down to it. As often happens with javascript, some basic
functionality is denied. Search is a very important feature to toss
aside without good reason.

Also, this script may make the assumption that the user who enables
javascript has a fast connection. I'm on a T1 line now, and the load
time was noticeable. Anybody care to test this on a dial-up
connection?

Don't get me wrong, I think it's _very_ cool. But these are
accessibility problems that can come back to bite you later, if not
considered at the start.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

30 Jan 2008 - 9:24am
kimbieler
2007

It's a neat implementation. However, there's something vaguely
Sisyphian about scrolling and scrolling and never getting to the bottom.

I kind of like the "page 1 | 2 | 3" links because they tell me how
much content to expect. I'd prefer the Ajax pagination if it gave me
some indicator about how long it was going to keep adding content.
Another thought: with shorter pages, you keep the navigation and
identity information visible at the top of the screen.

There are probably some applications for which this is an
improvement, however I'm not seeing it as a wholesale replacement.

-- Kim

+ + + + + + + + + + + + + + + + + +
Kim Bieler Graphic Design
www.kbgd.com
c. 240-476-3129
+ + + + + + + + + + + + + + + + + +

30 Jan 2008 - 11:49am
Alexander Baxevanis
2007

How about the iPhone approach, which is used for example in the mail
client: when you reach the bottom of what's already loaded, you need
to click in order to load more items, and you get an indication of how
many items there are.

I think in some applications it's important to be able to load more
items without losing the context of the original items. For example, I
spent the last couple of hours looking for photos in a stock images
website in order to illustrate a presentation. If I was able to have
that sort of "infinite scroll", I could for example easily compare a
thumbnail that would have normally been in the first page with one
that would have been in the second page, without having to switch
pages.

As for the browser built-in search, while it's a great tool, it's a
shame that it's not more integrated with the design of a webpage. It
would be great if the page could get an "onsearch" event when the user
invokes search to make sure all the content is loaded. It would also
be a nice idea to be able to exclude certain areas of the page from
searching, or demote the significance of a match in these areas, e.g.
if I'm looking for the world "previous", a navigation label named
"previous document" will create a spurious match.

Cheers,
Alex

On Jan 30, 2008 3:24 PM, Kim Bieler <kimbieler at mindspring.com> wrote:
> It's a neat implementation. However, there's something vaguely
> Sisyphian about scrolling and scrolling and never getting to the bottom.
>
> I kind of like the "page 1 | 2 | 3" links because they tell me how
> much content to expect. I'd prefer the Ajax pagination if it gave me
> some indicator about how long it was going to keep adding content.
> Another thought: with shorter pages, you keep the navigation and
> identity information visible at the top of the screen.
>
> There are probably some applications for which this is an
> improvement, however I'm not seeing it as a wholesale replacement.
>
>
> -- Kim
>
> + + + + + + + + + + + + + + + + + +
> Kim Bieler Graphic Design
> www.kbgd.com
> c. 240-476-3129
> + + + + + + + + + + + + + + + + + +
>
>
>
>
> ________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>

30 Jan 2008 - 12:44pm
Robert Hoekman, Jr.
2005

> This is cool. The only downside is that users often use the scroll bar as
> an indicator of the longth of the page. This pretty much skews what is an
> important bit of feedback to the user.

My own observations (of dozens of people using "infinite scrolling" on
various sites) are that not only is this true, but that users will often
relentlessly try to "reach the end" of the page. They don't realize that the
page is loading new results on purpose—they think that the results simply
have not finished loading. As such, they think the page is poorly
constructed (slow) and the design pattern is often seen as annoying.

I have video of people pounding on their mouse buttons in an attempt to get
to the end of a page that will simply never end.

I'm all for innovation, but pagination is such a standard thing that it will
take a lot of undo-ing to convince people that infinite scrolling is better,
especially considering the technical repercussions (such as Find
functionality).

-r-

30 Jan 2008 - 1:46pm
Shekhar P Bagawde
2007

How about this AJAX paging example from 'Componentart', most of you might
have already seen this-
http://www.componentart.com/webui/demos/demos_technology_showcase/web-services_ajax/ajax_withClientSideCaching/default.aspx

- Each action retrieves the required set of records from the server without
reloading the entire page;
- Retrieved pages are cached on the client, and the slider coloration
indicates pre-loaded areas.

cheers,
Shekhar

On Jan 30, 2008 1:44 PM, Robert Hoekman, Jr. <robert at rhjr.net> wrote:

> > This is cool. The only downside is that users often use the scroll bar
> as
> > an indicator of the longth of the page. This pretty much skews what is
> an
> > important bit of feedback to the user.
>
>
> My own observations (of dozens of people using "infinite scrolling" on
> various sites) are that not only is this true, but that users will often
> relentlessly try to "reach the end" of the page. They don't realize that
> the
> page is loading new results on purpose—they think that the results simply
> have not finished loading. As such, they think the page is poorly
> constructed (slow) and the design pattern is often seen as annoying.
>
> I have video of people pounding on their mouse buttons in an attempt to
> get
> to the end of a page that will simply never end.
>
> I'm all for innovation, but pagination is such a standard thing that it
> will
> take a lot of undo-ing to convince people that infinite scrolling is
> better,
> especially considering the technical repercussions (such as Find
> functionality).
>
> -r-
> ________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>

--
cheers,
Shekhar

30 Jan 2008 - 2:05pm
bminihan
2007

I really like this mode of browsing long lists, it would be really useful in
things like discussion forums. I agree with the caveats that others
mention, as well as the lack of bookmark-ability for pages deeper than the
first. Much of my work involves finding reference information for various
technologies, and I bookmark almost all of them (I rarely go back after a
few days, but still...), so there should be a way of recognizing where the
user is when they scroll down. Perhaps when each page loads you could
append a #page5 to the URL and a " - Page 5" to the title, so if you want
to, you could bookmark and come back to the spot you left.

Otherwise, great concept =]

Bryan
http://www.bryanminihan.com

-----Original Message-----
From: discuss-bounces at lists.interactiondesigners.com
[mailto:discuss-bounces at lists.interactiondesigners.com] On Behalf Of Etkin
Ciftci
Subject: [IxDA Discuss] Continuous Scroll

On the other hand ajax pagination assumes "if the user scrolled down to the
bottom of a page, he should be asking for more." Do you agree with such an
assumption?

30 Jan 2008 - 2:03pm
dmitryn
2004

My first reaction on seeing this demo was to interpret the navigation
controls as pertaining to an embedded video, rather than an actual live
search listing.

I would suspect that, unless this sort of control gains widespread usage,
many people would make the same mistake.

Dmitry

On Jan 30, 2008 11:46 AM, Shekhar Bagawde <shekhar.bagawde at gmail.com> wrote:

> How about this AJAX paging example from 'Componentart', most of you might
> have already seen this-
>
> http://www.componentart.com/webui/demos/demos_technology_showcase/web-services_ajax/ajax_withClientSideCaching/default.aspx
>
> - Each action retrieves the required set of records from the server
> without
> reloading the entire page;
> - Retrieved pages are cached on the client, and the slider coloration
> indicates pre-loaded areas.
>
> cheers,
> Shekhar
>
>
> On Jan 30, 2008 1:44 PM, Robert Hoekman, Jr. <robert at rhjr.net> wrote:
>
> > > This is cool. The only downside is that users often use the scroll bar
> > as
> > > an indicator of the longth of the page. This pretty much skews what is
> > an
> > > important bit of feedback to the user.
> >
> >
> > My own observations (of dozens of people using "infinite scrolling" on
> > various sites) are that not only is this true, but that users will often
> > relentlessly try to "reach the end" of the page. They don't realize that
> > the
> > page is loading new results on purpose—they think that the results
> simply
> > have not finished loading. As such, they think the page is poorly
> > constructed (slow) and the design pattern is often seen as annoying.
> >
> > I have video of people pounding on their mouse buttons in an attempt to
> > get
> > to the end of a page that will simply never end.
> >
> > I'm all for innovation, but pagination is such a standard thing that it
> > will
> > take a lot of undo-ing to convince people that infinite scrolling is
> > better,
> > especially considering the technical repercussions (such as Find
> > functionality).
> >
> > -r-
> > ________________________________________________________________
> > *Come to IxDA Interaction08 | Savannah*
> > February 8-10, 2008 in Savannah, GA, USA
> > Register today: http://interaction08.ixda.org/
> >
> > ________________________________________________________________
> > Welcome to the Interaction Design Association (IxDA)!
> > To post to this list ....... discuss at ixda.org
> > Unsubscribe ................ http://www.ixda.org/unsubscribe
> > List Guidelines ............ http://www.ixda.org/guidelines
> > List Help .................. http://www.ixda.org/help
> >
>
>
>
> --
> cheers,
> Shekhar
> ________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>

30 Jan 2008 - 2:24pm
bminihan
2007

There are many common scenarios where people only care about the first page
of results (which is why many SEO efforts are geared toward getting a site
listed above the fold on Google/Yahoo, etc).

I suspect there is almost a subconscious drive for many folks to see the
"whole first page" of results to know what they're up against, have a quick
pass through the page, before deciding whether to refine their search or
keep digging through.

In any kind of record listing, there are (at least) two big factors at stake
*: the assumption that the results contain what you want, and the
assumption that you entered the right query/keywords/selected the right
button to get what you want. If the 1st factor were the only one at play,
improving the ability to get to each page effortlessly would be the
important. If the 2nd is in play at all, you start interfering with a
person's ability to make that all-important decision - forge ahead or do
another search.

I'm thinking of that I Love Lucy episode in the candy factory where the
conveyor belt keeps spitting out candy to wrap and they can't keep up. If
you keep getting results no matter how much you scroll, your 3 second
decision about the 1st page of results turned into an endless loop getting
what you need to make that decision.

* I'm not a psychologist, just shooting from the hip, here...

Bryan
http://www.bryanminihan.com

-----Original Message-----
From: discuss-bounces at lists.interactiondesigners.com
[mailto:discuss-bounces at lists.interactiondesigners.com] On Behalf Of Robert
Hoekman, Jr.
Cc: discuss at ixda.org; Etkin Ciftci
Subject: Re: [IxDA Discuss] Continuous Scroll

I have video of people pounding on their mouse buttons in an attempt to get
to the end of a page that will simply never end.

30 Jan 2008 - 2:50pm
Jeff Seager
2007

Shekhar: "How about this AJAX paging example from 'Componentart'
..."

Hi, Shekhar ... As you might guess from my comments above, I'm in
part an accessibility advocate. As such, I like javascript as
progressive enhancement. AJAX is being used far too much (in my
opinion) for primary development with integrated enhancements. I hope
that distinction makes sense. Integrated enhancements are a barrier to
some users, and it's the biggest problem I have with what many people
call Web 2.0.

The Componentart implementation allows no useful content to be
delivered if scripts are denied, so because of our laws it isn't
anything I would be able to use in designing for a government website
(my main interest, these days) ... In other ways I like it, assuming
all intended users have good eyesight, good hand-eye coordination and
javascript enabled.

Very interesting example, Shekhar! Thanks for sharing it.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

30 Jan 2008 - 12:37pm
Vincent Robert
2008

Hi, I am new to this list so sorry that my first post have to disagree
with this implementation.

uI am really concerned about the user expectations. What I observed
from rapid usability tests is that users are confused by the initial
size of the page and as mch confused by the end size of the page.

"How a page so small can contain all the blog posts ? Oh, I can read
that I have to scroll to make them appear... But how many posts are
they in this page ?"

This does not seem like a natural reaction to me.

I think the problem with these "infinite" lists is that they are
not displaying something infinite. Time is infinite, number of blog
posts is not.

Let's rethink the whole thing from the beginning. I want to have a
single page that displays all my blog posts. I want that page to load
quickly although my blog currenlty holds 1877 posts. How can I do that
?

Maybe I can send to the user a page that has enough space for 1877
posts to be displayed as just couting them is trivial.

I can then use asynchronous loading to load posts that are in the
user viewport, like the first 13. Then my user can scroll in the page
wherever he wants and I will asynchronously load any content that its
viewport should be displaying.

I remember a similar widget capable of loading Yahoo search results.
The nice thing was that you were able to jump anywhere you want in
the result list and the results would just show up, meaning that you
could jump at the end of the list and get the last result without
waiting for the previous results to load.

I don't mean to be dogmatic here but I really think that web pages
should not change their size, unless the user manually activates some
expanding button like on the iPhone UI Alexander Baxevanis described.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

30 Jan 2008 - 2:58pm
Weixi Yen
2007

This has been around for a while, never caught on.

Biggest concern was bookmarking. You can in fact bookmark this type of
scrolling by adding event listeners, nobody tried it yet but it's easy
enough to do. Each back button would take you to a different anchor point
on the page. You can also email to friends depending on what part of the
page you are on, and they would be able to view the set of results you were
viewing when you sent the link.

More info here:
http://unspace.ca/discover/pageless/

On Jan 30, 2008 12:24 PM, Bryan Minihan <bjminihan at nc.rr.com> wrote:

> There are many common scenarios where people only care about the first
> page
> of results (which is why many SEO efforts are geared toward getting a site
> listed above the fold on Google/Yahoo, etc).
>
> I suspect there is almost a subconscious drive for many folks to see the
> "whole first page" of results to know what they're up against, have a
> quick
> pass through the page, before deciding whether to refine their search or
> keep digging through.
>
> In any kind of record listing, there are (at least) two big factors at
> stake
> *: the assumption that the results contain what you want, and the
> assumption that you entered the right query/keywords/selected the right
> button to get what you want. If the 1st factor were the only one at play,
> improving the ability to get to each page effortlessly would be the
> important. If the 2nd is in play at all, you start interfering with a
> person's ability to make that all-important decision - forge ahead or do
> another search.
>
> I'm thinking of that I Love Lucy episode in the candy factory where the
> conveyor belt keeps spitting out candy to wrap and they can't keep up. If
> you keep getting results no matter how much you scroll, your 3 second
> decision about the 1st page of results turned into an endless loop getting
> what you need to make that decision.
>
> * I'm not a psychologist, just shooting from the hip, here...
>
> Bryan
> http://www.bryanminihan.com
>
>
> -----Original Message-----
> From: discuss-bounces at lists.interactiondesigners.com
> [mailto:discuss-bounces at lists.interactiondesigners.com] On Behalf Of
> Robert
> Hoekman, Jr.
> Cc: discuss at ixda.org; Etkin Ciftci
> Subject: Re: [IxDA Discuss] Continuous Scroll
>
>
> I have video of people pounding on their mouse buttons in an attempt to
> get
> to the end of a page that will simply never end.
>
>
> ________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>

31 Jan 2008 - 7:44am
Patricia Garcia
2007

My first reaction when I first heard of this method, oh, maybe a week
ago was I didn't like it - in fact, I would go so far to say, I
hated it. Than I saw this post and checked out the link and again, I
was thinking how frustrating it was.

But than Meredith mentions Google Reader uses this approach. I have
been using Google Reader for a couple months now, it's how I keep up
with my posts here. I miss a day and I have over a 100 posts waiting
for me to read, and with expanded view, it's a long page to scan.
So in these two months, it never occured to me that the infinite
scroll was being implemented, it was invisible to me. I use my
scroll roll on the mouse (what is that called anyway?) so can't say
I paid much attention to the scroll bar length and position. I also
notice that if you auto scroll (press the scroll roll on the mouse)
it will continue to load for you with only a slight delay between
loads for an almost seemless scroll to bottom.

I think it's one of those situations that if you bring it to
someone's attention, they may decide not to like it, but if you
don't say anything, it probably won't interfere with user
experience much.

But of course, it depends. :)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

31 Jan 2008 - 8:42am
Murli Nagasundaram
2007

One of the things not mentioned in this discussion is whether such an
implementation (infinite scroll) is appropriate for all kinds of
information. There is tool/technology/feature, on the one hand, and there
is the content on the other. While there are probably people who read
entire novels online, they likely constitute a minuscule minority. Browser
based interaction is very well suited for relatively small quantities of
content, especially content that is highly hyperlinked (which was the
original purpose of the web, wasn't it?). I don't think the book or
book-like devices are likely to go away because the form is very well suited
to the nature of content and manner in which that content is imbibed.
Likewise, infinite scrolling works very well in Google Reader and other such
applications that contain many pieces of small, independent chunks of
content. I absolutely detest reading looooong articles online (such as
articles from The Atlantic Monthly, New Yorker, Salon). The moment I see
that the article runs into, say, 14 pages, I search for a link that
generates all the content on a single page (if available) and after reading
the first couple of paragraphs, I read the concluding ones and quickly scan
the rest. If it takes too long to load, I bail out.

To return to the main issue, one must evaluate the usefulness or usability
of Infinite Scrolling with respect to the nature of content in question. My
hunch is that if a study were to be conducted, people would rate the U/U of
IS variably according to application/content.

Unfortunately, every time a new technology or paradigm is developed, there
is a rush to try applying it indiscriminately to just every manner of
content on Teh Interwebs.

- murli

31 Jan 2008 - 9:19am
Jeff Seager
2007

Patricia said: "I think it's one of those situations that if you
bring it to someone's attention, they may decide not to like it, but
if you don't say anything, it probably won't interfere with user
experience much."

I'm not so sure about that. It's one of those "features" that
adds nothing significant to the user experience and subtracts some
functionality and choice. How many of us are just like Murli in the
way we scan articles, reading the beginning and end as a way to
decide whether in-depth reading will be worthwhile? I know I do, and
it's something I don't want to change because it saves me time and
adds value to my experience.

Just because it's cool, doesn't mean you should do it.
Well-structured HTML/XHTML loads very quickly in any browser, and is
preferable for many reasons. This is one of those presentational
"enhancements" that will -- in the long run -- encourage people to
disable javascript.

In designing for the Web, I think we should always ask whether a
feature really adds enough genuine value to offset its cost (as
measured in time and comprehension) to the end user.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

31 Jan 2008 - 9:11am
Dan Harrelson
2007

Another good real-world example of this pattern is the Microsoft Live
image search.

http://search.live.com/images/results.aspx?q=ixda

...Dan

31 Jan 2008 - 9:27am
John Gibbard
2008

This paradigm can be considered from a different angle (can paradigms
have angles?) by looking at more 'experiential' sites. See, for
example, the Orange Unlimited site which encourages playability with
the infinite page and initially entices the user to continue their
fruitless scroll. Quite how persistent people are with this remains
to be seen.

>From a 'success' point of view it did at least win an IAB award.

URL: http://unlimited.orange.co.uk/flash/go

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

31 Jan 2008 - 11:00am
Robert Hoekman, Jr.
2005

> Another good real-world example of this pattern is the Microsoft Live
> image search.
>
> http://search.live.com/images/results.aspx?q=ixda

This is one of the pages used during the testing I witnessed. Aside from the
problems I mentioned in my previous reply, participants also had a very
difficult time because of the broken Back button. When you click an image to
learn more about it or see a larger view, you leave the page, and when you
hit Back, you have to start at the beginning. It doesn't "remember" where
you left off and return you to the section you last viewed.

This is a problem in many "infinite scrolling" implementations. Reminds me
of the ongoing battle between Flash apps and the Back button.

31 Jan 2008 - 3:29pm
bryan.haggerty ...
2005

> So in these two months, it never occured to me that the infinite
> scroll was being implemented, it was invisible to me.

Part of some of the debate may be due to that the example
implementation provided (http://www.symfony-project.org/demo/
pager.html ) does not have that seamless feeling like Patricia has
experienced with Google Reader. It has a very noticeable stutter
between loads, thus providing a very unnatural scroll feeling.

In regards to the bookmarking issue. It does depend on the context.
For example, bookmarking a page of blog entries is irrelevant as page
3 of blog posts will be different tomorrow than it was at the time of
bookmarking, hence making a bookmark of pagination useless. The same
goes for search results, or any kind of paginated dynamic data.

Bryan Haggerty

1 Feb 2008 - 7:09am
Etkin Ciftci
2007

Since this thread is about a method, I totally agree that we also have to
consider the context its used in.

Some said that it weakens the functionality while some offered new methods
of empowering the ajax pagination. Thanks for those great comments.

Google Reader is a one fine real life example where you don't bookmark
within the browser but star and share individual items within the Google
Reader itself. So, I think there are more solutions out there in order to
empower the method and suit the needs of the content.

Bests.

Etkin Ciftci
Interaction Designer
eciftci[at]bilende.com
Bilende Design & Development
Istanbul-TR

1 Feb 2008 - 10:44am
Jeff Seager
2007

Thanks, John, for posting the link to Ultimate Orange. I think that
kind of ultimately pointless content may be the perfect application
of the continuous scroll, and it's a great proof of concept.

It reminded me of the old story about the guy who admired Britain's
quaint scenery and said to the man beside him on the train, "I guess
all of England could fit into one corner of Nebraska."

The Englishman famously answered, "But to what end, young man?"

Indeed.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

5 Feb 2008 - 7:30pm
Cindy Lu
2006

Hi!

I thought this thread of discussion was interesting so I would like to offer
some comments.

I am working on an application used by investment professionals. Some users
we visited did not like paging. I observed users setting the page limit to
max whenever they started an application and complained why the default was
not the max. When they log in, they want to quickly scan how their clients
are doing, so they scroll down all the clients' records and glance some
key data in columns and rows.

We recommended the continuous/infinit search similar to Microsoft's live
search to our product team. Our product Beta has launched. We receive some
positive reactions. However, we also observed some problems:
1) Performance. As you scroll down the page, there is a delay for the system
to load the data. Filtering and sorting can be slow too.
2) Showing the user the total records and where the user is (see Live
Search). The development team has difficulty in capturing the total records
and showing where the user is because the records are dynamically generated.
3) The jump of the scroll bar. As you move the scroll bar to the bottom, the
scroll will jump back after a short delay and more records are shown.

Currently, the team is in the process of improving the performance and
resolve the total record issue. We are in the process of testing it with
users.

If you have designed such feature for enterprise applications and have
lessons learned for both designs and usabilty testing, I appreciate it if
you could share them.

Thanks!

- Cindy

5 Feb 2008 - 11:42pm
Jeff Seager
2007

Cindy said: "We recommended the continuous/infinit search similar to
Microsoft's live search to our product team."

Why? What advantage does this offer over raw and semantically correct
XHTML or XML, which loads quickly, can natively incorporate tabular
data and everything else you need, is fully searchable from the
start, easier to maintain ... ?

Seriously, why add complexity when it flies in the face of user
expectations and wishes? Why wait for a study or an "expert" to
tell you what you've already been told?

Granted, my perspective is skewed a bit by designing mostly for the
Web and being held to strict standards of accessibility. But again
and again, here's what users tell me: "Deliver the goods, and
don't be too cute about it." That's especially true for financial
data and any other information they consider critical to their daily
decision-making. Nielsen and others have confirmed this time and
again with careful research.

In general (and there may be some really good exceptions, especially
in a closed system, but I'm not recalling any right now), I'll say
"yes" to server-side scripts because when they're well implemented
they can deliver the goods faster or more elegantly -- "includes"
can actually _reduce_ the code burden -- and I say "maybe" to
client-side scripts because they can impede or completely block some
or all users.

Not everyone agrees with me, but I'm for standards-based progressive
enhancement. It's straightforward and logical, and that usually
serves everyone better if the data is also straightforward and
logical. It's not that I'm a Luddite; I just don't want to use a
cannon when a water pistol will do the job.

The immediate impression is not the lasting impression, and the
"wow" factor can wear thin very quickly.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=25287

6 Feb 2008 - 12:02pm
yannis vasmatzidis
2006

Cindy,

you mention that

> "I observed users
> setting the page limit to
> max whenever they started an application and
> complained why the default was
> not the max. When they log in, they want to quickly
> scan how their clients
> are doing, so they scroll down all the clients'
> records and glance some
> key data in columns and rows.

Perhaps this behavior is because of a poor design of
the current system that does not reflect an
understanding of end-users? What are the "key data"
you are talking about? With the solution you
described, you still are leaving it up to the users to
scan the data and identify the key information - a
potentially frustrating experience with financial
professionals. If you understand what the key data
are, would it be difficult to display these first,
thus mininizing end-user effort?

--- Cindy Lu <cindylu01 at gmail.com> wrote:

> Hi!
>
> I thought this thread of discussion was interesting
> so I would like to offer
> some comments.
>
> I am working on an application used by investment
> professionals. Some users
> we visited did not like paging. I observed users
> setting the page limit to
> max whenever they started an application and
> complained why the default was
> not the max. When they log in, they want to quickly
> scan how their clients
> are doing, so they scroll down all the clients'
> records and glance some
> key data in columns and rows.
>
> We recommended the continuous/infinit search similar
> to Microsoft's live
> search to our product team. Our product Beta has
> launched. We receive some
> positive reactions. However, we also observed some
> problems:
> 1) Performance. As you scroll down the page, there
> is a delay for the system
> to load the data. Filtering and sorting can be slow
> too.
> 2) Showing the user the total records and where the
> user is (see Live
> Search). The development team has difficulty in
> capturing the total records
> and showing where the user is because the records
> are dynamically generated.
> 3) The jump of the scroll bar. As you move the
> scroll bar to the bottom, the
> scroll will jump back after a short delay and more
> records are shown.
>
> Currently, the team is in the process of improving
> the performance and
> resolve the total record issue. We are in the
> process of testing it with
> users.
>
> If you have designed such feature for enterprise
> applications and have
> lessons learned for both designs and usabilty
> testing, I appreciate it if
> you could share them.
>
> Thanks!
>
> - Cindy
>
________________________________________________________________
> *Come to IxDA Interaction08 | Savannah*
> February 8-10, 2008 in Savannah, GA, USA
> Register today: http://interaction08.ixda.org/
>
>
________________________________________________________________
> Welcome to the Interaction Design Association
> (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................
> http://www.ixda.org/unsubscribe
> List Guidelines ............
> http://www.ixda.org/guidelines
> List Help ..................
> http://www.ixda.org/help
>

____________________________________________________________________________________
Be a better friend, newshound, and
know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

Syndicate content Get the feed