Pop-ups, why not use them?

Since every thread started by someone asking for coding help with pop-ups ends up being hijacked by the anti-popup crowd I thought I’d give you all the opportunity to vent on this thread. The Usability forum seemed the apprpopriate place for the discussion.

How ironic would it be if someone came on this thread and asked for help coding one?

So, explain the evil of popups here and then next time someone asks for coding help you can refer them to this thread using a link and their own help thread can stay useful and on-topic.

I think that pop-ups used in that scenario should respond to hover/focus rather than a click/keypress. Either that or make their function obvious. Simply because people might be reluctant to click a link (if its function isn’t clear) while completing a form for fear of losing what they’ve typed.

We had/have links in our form that go to whole new pages, or open PDFs (I’ve tried to change these back to HTML pages as much as possible).

This is because people, given the opportunity to read something first, or just start filling in the form, will fill out the form first. Then they get to a question asking about something they have no idea what it is, and need the help text. I am a firm believer in forms only containing form elements, but with our insurance forms, help text was necessary.

It works on focus/hover, but our testers tended to click on them (because that’s what every other web site does), meaning they went to a whole new page without realising they would.

When they hit the back button, their form info might be gone (if browser caching was turned off).

So we added target=“blank” to those links. The ones that only had help-text (no link to other page) went to #void, so users clicking on them did not get a page refresh, brought to teh top, or anything else. I believe one of the usability failures in our forms is, you can’t tell which help texts are links and which aren’t (the “popup” says something like “opens in new window” but then in most browsers it doesn’t anyway, bleh).

We were also forced to do the same for all pdf links. Everyone who tested expected the PDFs to open in new windows or new tabs, so when ours didn’t, they closed the browsers by accident. Every time. Meh.

There was no reason to have the help text in an actual popup, but it made a LOT of sense for any new pages to be separate in some way. This retained unsubmitted form data in all browsers whether the user had caching on or not, and some of our forms are really, really long.

If the “new pages” were short enough, they probably would have been some kind of popup window, but I found it easier to link to the relevant part of the page (if the user needed to know the difference in the types of coverage, I sent them to the “Coverage” page with a hash to specifically the types of coverage).

is the change of focus announced, or assumed by the user if the ARIA role is understood? Or might it cause a problem? Or not?

If the ARIA role(s) are understood, they are supposed to deal with that (the role tells the AT whether it should be in focus mode (role=“application” does this, it’s kinda like forms mode, so users’ keystrokes edit and select stuff rather than navigation/reading) or in a reading mode (role=“dialog” or role=“alertdialog” are supposed to do this I think), whether focus or navigation is allowed outside that area, and supposed to let both the user and the AT know what’s going on.

You would also have a role on whatever the user is clicking on to initiate the dialogue.

Without ARIA, or with sighted keyboarders, I’m still leaning towards JS manually moving focus to the dialogue box and not allowing it out until the user has dealt with that box. Then you’d manually have to put the focus back when they are done, either exactly where they were when they triggered the event or right after it. I actually don’t know what the best way is to do this, and the idea of hijacking the tab key makes me uneasy, but that’s due to my not being an application developer.

Very true. And while lack of side-by-side comparison is an excuse for user-written scripts of various kinds, it’s not an excuse for the site to start making pop-ups.

Should be is nice, but is is what is.

cheers,

gary

I think that pop-ups used in that scenario should respond to hover/focus rather than a click/keypress. Either that or make their function obvious. Simply because people might be reluctant to click a link (if its function isn’t clear) while completing a form for fear of losing what they’ve typed.

Autofocus might be the most stupid HTML5 thing that I’ve yet seen.

Unless you use Javascript to specifically move the focus over to the popup, the keyboard focus is still on the page underneath. I’ve tested the generic Lightbox2 that uses jQuery and while after clicking a thumbnail I do have the ability to use the keyboard to do things like next and prev image and closing the thing, my TAB remains on the page underneath. This breaks the rules of modals: user should have no access to the page underneath until some decision has been made in the modal (such as what happens with alert() and confirm() popups).

I agree about the focus issue and it seems logical to move focus, but it raises questions of its own, like: is the change of focus announced, or assumed by the user if the ARIA role is understood? Or might it cause a problem? Or not?

Talking of such problems (and getting my comment on topic), isn’t there an issue of context with window pop-ups and screen readers? If what I’ve read is correct, that alone is a good reason to very much limit their use.

Ah. I don’t disagree with that. And the most recent popup thread had plenty of people starting out with “popups eat our children” and then link to some other solution that the OP might not have thought of.

I was referring to advise on how the specific problem is better solved without pop-ups, rather than simply warning against them.

but it still doesn’t invalidate my original statement.

I must be misunderstanding, because I read that you disagreed with my post because “there should be advice about avoiding them”. However, my post has this. The difference is, it doesn’t have it hundreds of times. One post warning of the dangers and misery of popups in a particular thread asking about them should be good enough. If the OP states they’re already aware of the issues, but for whatever reason is going ahead anyway, then the wave of posts stating that they are bad is basically a waste of bytes.

Side-by-side comparison should be natively implemented by the website, though. Why so few sites have this feature is beyond me.

There are a number of ways to show additional info. But the question was about having a rationale for the pop-up. My particular rationale in this example was that being a separate window, you could drag it to the side and continue in the primary page, even popping up additional windows for side by side study if desired. For my own browsing, I have even provided my own “pop-up” when the author didn’t, by purposely opening an info link in a new window, and resizing it for convenience.

As to lightboxes, I have yet to find a rationale, other than hypothesizing a designer’s form of self-abuse.

cheers,

gary

True, that’s a problem, but it still doesn’t invalidate my original statement.

in that advise on how to find alternatives to a pop-up

hundreds of times, ad nauseum?

Once ought to be enough, unless that poster left out one of the many good reasons. Looking at the most recent thread on popups, though, it’s a lot of repetition with vague hopes of converting the non-believer to the cause.

Stephen
True, but that is akin to the user starting his browser, so I wouldn’t really call it a pop-up.

Gary
This could also be implemented as a Javascript-driven fold-out (or lightbox, even though I personally dislike them, and suspect many with visual deficiencies find them hard to use). Implemented gracefully, of course, so that the information is displayed plainly on the page for those who doesn’t have Javascript.

Stomme
I somewhat disagree, in that advise on how to find alternatives to a pop-up in the given situation should very definitely be given.

There is hardly ever a reason to use them, when there are so many good lightbox solutions and the like out there.

Lightboxes are another reason I have Javascript disabled. They’re slow and retarded and eat screen space for nothing but bull and crap. Put those two words together.

Woe be the day when some HTML5 crap starts building those in so I can’t block them, like they’re doing now with autofocus and auto-fill in forms. Hate, hate, hate.

How does a blind person with a screen reader access the info in a popup, for example?

If it’s built by the browser (like you get with actual new windows, alert() popups, etc), focus moves to it and it’s announced. New users are probably confused but, new users are confused by everything anyway.

There’s a problem with HTML/CSS/JS-built modal dialogues though. This includes lightboxes. Unless you use Javascript to specifically move the focus over to the popup, the keyboard focus is still on the page underneath. I’ve tested the generic Lightbox2 that uses jQuery and while after clicking a thumbnail I do have the ability to use the keyboard to do things like next and prev image and closing the thing, my TAB remains on the page underneath. This breaks the rules of modals: user should have no access to the page underneath until some decision has been made in the modal (such as what happens with alert() and confirm() popups).
Some people have been playing around with the new ARIA live regions but I haven’t seen this actually work anywhere yet… application roles stole navigation and reading abilities, alertdialog was buggy everywhere, someone somewhere always broke…

…however I expect these will get ironed out and fixed eventually, and screen reader users will be the easier group to write for… everyone else will be the problem.

Here’s a bad example of a popup: when I tried to sign up for the Mobile Unconference, the sign-in form was in a popup… one that was much, much, much smaller than the form inside, and the morons apparently don’t know how to allow scrolling without implementing 6 more scripts in the popup. Like, hello??? You need Javascript to create a scrollbar??? Since when?? Like many other people, I even missed the fact that the thing could scroll and hit the wrong button trying to submit. They used these stupid little frames to make it cute. After fullscreening the thing you finally realise what the hell they were doing. (Yeah, I was using my slutbrowser, Chrome, to fill this in… it’s my “let everything in” browser where I do stuff where there are just hundreds of retarded scripts doing stuff HTTP should be doing).

I guess it was “let’s try to make things as DIFFICULT and BIZARRE as possible for users” day. Let’s make sure it works so UNLIKE any NORMAL form that people screw it up a few times before they actually get it right.

Oh, I know designers love trying goofy new things, but this is why they should all get chained down to a chair and forced to learn basic. freaking. usability. before being allowed to touch a keyboard. Oh wait, I mean mouse, these people only know how to use mice. ARG.

And on that note, I agree 100% with JJ that it’s kinda getting retarded whenever someone asks “how do I do popup type X?” that more than one response has to be mentioning how horrible the things are. One is probably enough, and if the OP already says “yeah I know people hate them” then I don’t see where the benefit is of ranting further about them in that thread. (though, the latest thread, where the OP kept insisting that his popups were special and adorable and not annoying at all in any way, that just made me laugh really hard. Yeah right, bro)

When someone says “I want to cut my own head off” and one person say “that will be painful and cause your death” and the OP says “Yeah but I really wanna, it’s a science experiment” then we’ve done our job and now anyone who wants to explain to the poor sod how to cut his damn head off, do so. We’re not required to spend 16 out of the 20 responses on how much it really sucks to have your head cut off, are we?

If users hate popups, they’ll do what the rest of us do: leave. They are usually on commercial and personal sites, not sites we’re forced to use like our utilities, banks, and government sites (if they are, then we do get to spend 19 out of the 20 responses complaining about how much popups suck balls and bring Satan back to Earth and cause wee orphan children to contract cancerous forms of herpes, of course), so we generally have the option to leave.

Authors are allowed to destroy their own sites, especially if they live in countries without any demands that they actually build accessibly and well.

BTW: next site I get a retarded popup in, I’m actually going to spend time writing them an angry letter just so the site owners can enjoy my colourful language. After all, they asked for it, knowingly.

*edit oh, I need to add a word to my post to make it complete:

hitler

Funny how this topic is in the Accessibility forum and no one wants to discuss this (although I guess usability is getting a run to some extent).

I agree. In fact, I’ve become so quick to disregard anything in a popup since its usually a garbage ad that one could contain the winning lottery numbers and I wouldn’t even notice.

I have found one class that I appreciate when done well. When viewing a catalog page, and I desire to see a larger view of the object, or a set of specs, a somewhat smaller than the browser pop-up is handy. I say smaller because I want to recognize its function. I want a graceful failover to a normal hyperlink.

I’m trying to remember when my default browser wasn’t opened first thing, or at least immediately after a terminal and Emacs.

cheers,

gary

The only situation I can think of where popping up a new browser window is actually required is where there is a web page link in an email and there is no web browser currently open. Then you need to pop up a web browser window in order to be able to view that web page.

Poor accessibility and usability. They’ve always been bad because they broke the regular browsing paradigm, but with the current crop of tabbed browsers and mobile browsers, they’re oh so much worse. Even assuming the pop-up isn’t blocked (which is like assuming it will be sunny in England in the summer), it’s likely to appear as a regular tab (ie, full height and width of the window) rather than at the dimensions specified. That makes it even more likely that it will confuse people who don’t understand why the back button isn’t taking them back, or how to get rid of it. For people using assistive technology, the nature of the problem is the same but the outcome is many times worse.

I would include unrequested lightboxes in the same category. When I arrive at a site and the first thing that greets me is a lightbox asking if I want to fill in a survey, or inviting me to take part in some free prize draw, my impression of the site plummets like a lemming going over Niagara Falls. With lead boots on. Anything that interrupts my reading and use of the page – whether it’s an unrequested pop-up, an unrequested lightbox or a badly placed on-page advert – is an irritation that degrades my experience of the site, because it requires extra effort to do what I went onto the site to do.

To get the attention of the visitors who CAN see it. Browsers can block images, Jave Script, Direct X, cookies, and a whole bunch of other stuff so you can never be sure that all your visitors are seeing the same thing anyway. Why then should you regard popups any differently from all the other content that browsers can block?

Pop-up blockers are much more common than users blocking images or scripts, and are more likely to be switched on by default. But that’s beside the point – nothing critical should ever rely on images, Javascript, DirectX, cookies, Flash, the latest version of IE or any other random thing that people might or might not have. You should always ensure that your criticial content can be accessed as far as possible by a basic, bare bones browser reading the HTML and nothing else.

I’ve clicked on pop-ups, lightboxes and other interferences by accident plenty of times. And you can guess what that does to my impression of the host website.

Stevie, you are ever so right.

And now, for the ultimate question: Can anyone, anywhere, present a realistic use case, just one, in which a pop-up is the best solution to a given problem? I maintain that the answer is no.