In the sidebar, various panels (such as Bookmarks and Related Links) are using single-clicks to open/close folders and double-clicks to load in URLs into the content area. This is so wrong that I don't even know where to start... We need to standardize on single-clicking to select something and double-clicking to open/close/load it in.
Passing to German. Just name how is should be done and I will do it.
I think it depends. For example in a what's related panel which is read and click only I would want to a) eliminate folders altogether if possible b)and b make action 'open folder' single click. This is usage based and consistent with what we found Moz Classic usability testing for the bookmarks panel. Double clicks are inherently unaccessible and inconvenient to beginners. On the other hand in mail where selection matters a lot with regards to a folder the action sequence you desribe is the proper one. Ideally the twistie icon "> for closed V for opened" takes care of single-click opening in the sidebar but we not have enough space for it. Maybe a twistie only without a folder for those cases?
For Related Links, while we can get rid of the top level RL folder if we hang all the data off of the root of the tree, you still won't be able to get rid of folders completely as the RL data itself can specify folders internally... things such as the "Site Info for ..." folder. Again, I don't agree that the action of opening folders should be single click. It is very inconsistent to have folders in some parts of the application that require double-clicking while others do not. Usability testing from Moz Classic demonstrated that beginning users could grasp double-clicking... in the worst possible case, they'd try a single click, and if that didn't work, they'd then try double-clicking. The big issue here is that some panels (say the bookmarks panel, for example) are going to REQUIRE double-clicking, as they'll need single-clicks to mean "select". Why? As always, the answer: management. That's why twisties become very useful to have, as they allow the single-click to open behavior. I don't buy into the whole argument of "we don't have enough space for twisties" either, by the way. :^) I'm arguing that we need as much of a consistent behavior across the entire application as possible. Since there will be panels in the sidebar that allow management (whether we do it or not, 3rd parties will), they're going to need single vs double click distinctions. I believe it would be very bad to have one panel where a single-click does only a selection, while in another panel it actually loads something in. That's BAD. (Have I ranted enough?)
I agree with Robert.
Marking beta1, this came up quite a bit at the sidebar bugorama. Slamm needs to be told when you should one click and when you should double click.
Having sat through those usability tests as a specialist, I can tell you that German is right in the findings. Additionally, users had no difficulty with the difference in behavior for mail and the browser as long as the look assisted in determining the behavior. Consistency is not the objective. Understandable, usable and enjoyable is. Users associate different interaction behaviors with different looks. We should take advantage of this and encourage this where it will enhanse the design and usability. We have actual data here that this can work. And of course I can test this again in an up comming study.
I *STRONGLY* suggest that you do some rigorous testing in this area. Inconsistent behaviors don't often lead to a user experience which is either understandable or enjoyable to most users. [German is actually WRONG in regards to Mozilla Classic. (Sorry, German... but you told me this yourself!). Mozilla Classic testing showed that beginning users might not initially understand the difference between single and double-clicking; however, they would naturally try the latter if the former didn't work.]
Oh perhaps I should be more specific. These things need to be taken in context. The SideBar that the users were presented with in this study gave little visual indication to the user how it was suppoded to be used. For that reason we got the response that beginning users had difficulty with the single and double click. For example: things that look like links, or buttons, they are associated with a single click and in mail the folder is associated with a double click. These visulal ques get a far stronger response that consistency across apps. Possibly because they are consistent with use metaphores. The response to a visul cue can be imediate and very strong.
This bug specifically refers to XUL tree listings (such as the Bookmarks or Related Links sidebar panels, as mentioned in the initial posting of this bug), not HTML-based sidebar panels; there are no buttons or items that look like links in the XUL trees in question.
As Lake points all it all depends on context. In the Moz Classic test we did indeed have people see single click and use that once bookmarks were docked to the side of the browsing window. In that context people liked the convenience of single click. We also found that there was -no- harm when single click was used for bookmarks in the sidebar, and at the same time double click was used for selectable, manageable items that could be re-arranged. In the context of the bookmark popout window people did not like single click action when they meant to do selection, and were concerned about 'launching' the bookmark when they did not intend to do so. Also s Robert points out most people double-clicked once they saw that single click did not do anything but selecting the item. I think it all depends on what the search results in the sidebar want to be: here is my opinion: Just becuase technically they are both based on lists things like a mail folder and a search results are not the same and will not be used similarly by end users. Search results do not require any management, that is re-arranging or some such thing. Therefore we should allow the convenience of single click. I believe that we will find once again that that will not conflict with using double click in other contexts such as mail, where managing list items is required. I also do not think it will kill us if we have double click there, but I would like to see how well single click is faring with real users and then make corrections as we need to. In addition I strongly recommend that search items be shown as clickable links because that is what they are.
Target Milestone: M16
Ah ha! Search results can have a *LOT* done to them! They can have context menus brought up on them to do things like filter out bad results, save items into bookmarks, dragging&dropping into bookmarks or a browser window, be re-sorted based upon different criteria such as relevancy or price, and LOTS more. Don't make the mistake of thinking that there is no management to search results. We need to PLAN for both what we want in 5.0 as well as future versions, and build up consistency from the initial get-go, so that the features we add into the product remain understand and usable, instead of us being forced to change behaviors that users will have built up and come to expect.
*** Bug 28103 has been marked as a duplicate of this bug. ***
Users have coped with clicking doing one thing, right-clicking/Ctrl-clicking doing another thing, and dragging doing something else, for several years. They haven't needed to select a search result (before doing something to it) before; they've just brought up the context menu for the link. Surely the same principle could apply to bookmarks and other items which currently need a double-click? Single-click to activate, right-/Ctrl-click to alter, drag to move. (Disclaimer: I have a personal vendetta against the whole idea of double clicks.)
I agree with German and Lake (and the users I've watched); we really need to use single-clicking for most of the default 'load this' actions in sidebar. The poor guy in today's usability study single clicked every single time, then double clicked when nothing happened. Even Lake, when trying to show him how it worked afterward, was single-clicking. In the past six years we have taught users to single-click in the browser, doing so well that it is now an option in the major OS file system GUIs. I understand the argument that you can do other things to many of these items, including manage them. However, the overwhelming thing they are used for is to load them in the browser window. We should optimize for that case by providing and showing all the same affordances as we do for links.
I'm kind of amazed that we are running these tests without knowing ourselves how to use the product. (Single vs. double-clicking)
A different viewpoint: can we write products without knowing how they will really be used? We don't have a captive user base, we can't insist they learn the 'right way'. Lake "knows" how to use the product, that just shows how strong the single-click training is.
Nice try, Peter, but no cookie for you.
I know, and I know it well. That is just the point we are making. Even when people know, they single click on these and double click when it doesn't work. And that is the wording the participants use "When it doesn't work I double click" (this is a direct quote) After the study was done I even asked one: Well, what about sorting and filtering... They said: you don't want to do that here maybe like bookmarks in another window. I have even asked leading questions that should get them to ask for double clicking but no dice. There are still some more studies to go. I bet you want to wait and see.
If we want to wait and see (i.e. wait for the usability study to complete), then great, but why did German just send out a big email about how the search UI is bad with one of the reasons being that everything isn't single-click? Anyway, I'm certainly happy to wait and see the final results. As an aside, I really like hearing direct quotes such as: "When it doesn't work I double click" (this is a direct quote) since it helps add validity to the idea that we can have a consist double- clicking sidebar panel if we choose, and that even beginning users would be able to use it.
I think the more important excerpt from that is "it doesn't work".
Peter, the point is this: If a user tries one thing (single click) and it doesn't work, they do know to try something else (double-clicking) which DOES work. That's a lot different than something which the user can NEVER figure out. Beginning users may not want/need/understand functionality like sorting, selecting multiple items, drag&drop, etc. Intermediate/Advanced users do... and will want this functionality. If we design our product solely for beginning users, we'll be no better than IE's current interface... fine for beginnings, and really annoying for a lot of users who aren't immediate beginners.
These are not immediate beginners. We are testing typical/intermediate users. Even the novice user had some experience with searching and browsing.
I think I understand your point pretty well. Nobody is claiming they'll never figure it out, just that it is not expected, and perceived to 'not work'. We don't design just for beginning or advanced users, we need to design primarily for typical expected users, and the typical cases. So far, data seems to show most people expect to click once on these, and don't expect to sort or manage them much, if at all. This means that we should optimize for this case. Sorting, multiple selection, D&D, etc are all edge cases that should be handled without interfering with the typical expected case. We should certainly optimize for advanced users too, after all we hope that all our users stick with the product long enough to learn it well. Even so, they (we?) are a tiny, although very vocal, minority.
Since we're all picking favorite quotes, here's mine: `After the study was done I even asked one: Well, what about sorting and filtering ... They said: you don't want to do that here maybe like bookmarks in another window.' Given that users are often poor judges of their own past or future behavior, there is no way (unless your usability studies go for weeks on end) that you can get accurate figures on how often users will try to manage bookmarks etc in a sidebar. But I'd agree with whoever was the subject in this test: if I want to do any decent management of my bookmarks, I'll open the bookmarks window, where I have more room than there is in the sidebar. (And, where the bookmarks look like selectable items, rather than like links.) > Beginning users may not want/need/understand functionality like sorting, > selecting multiple items, drag&drop, etc. Intermediate/Advanced users do... > and will want this functionality' ... (a) Intermediate/advanced users are the ones who know how to use context menus, and other forms of interaction which are slightly more complex than a single click. (b) If you implement the standard Shift-click and drag-a-box-round-them methods of selecting items, you haven't taken away the ability to select items in the sidebar. (c) Making items links doesn't take away drag-and-drop ability. (d) If you follow that argument to its logical conclusion, we'll have to make users double-click to follow HTML links as well, because intermediate/advanced users occasionally want to do something else with links besides follow them.
Bookmarks in the sidebar panel currently have icons, toggles, selection, all the goodies of old-fashioned trees; no underlining like links. Earlier comments in this bug note that we all agree that HTML looking links are single-click. No one is arguing that. While you may not, I very much expect to be able to have management in the sidebar panel. I believe we need to strike a careful balance between these camps.
All panels do not have to be the same. Of the 500 panels we ship with beta1, about 490 will be web pages from Netcenter - all of which have single click web links. Most of My Sidebar is on the web. The only way to be 100% consistent is for every panel without exception to be single click. I don't think that is the right thing since there may be cases like Bookmarks where double-click is the correct answer. What needs to be consistent is the way we display a single click (underlined) vs the way we display double clicks, but different panels can have different behaviors. I don't think we have to limit 100% of our panels to just single click. Here is my recommendation. We need to decide for each client panel whether it should be single click or double click. If we all agree that each panel needs to have its own decision, let's then start specific bugs per panel and mark this one as Fixed (as it is). (Marking as Netscape confidential since discussing Netcenter content. After beta1 can make this bug public again.)
spam: changing qa contact on sidebar bugs from paulmac to email@example.com (all 67 of them!)
Whoops, wrote bug. Ignore my attachment.
I'd be interested in hearing how JohnG knows the 'the correct answer' for bookmarks - do you have the teacher's edition John? BTW, I'm making this bug public again based on your b1 comment. FWIW, I've been using Win98 in single-click mode now for a couple of weeks. The main difference for me is having to click a _lot_ less. I haven't haven't had any problem managing files, although some of the techniques are not very discoverable. If only Windows didn't now also assume that I'm too stupid to be shown file sizes...
Ok, so assuming that (1) panels where single-clicking is the opening action will have items which look like links, (2) panels where double-clicking is the opening action will have items which don't look like links, (3) separate bugs will be filed if any particular panel is using the less useful choice of these two behaviors, and (3) Robert in the red corner seems destined never to agree with German, Lake, Peter, and myself in the blue corner ... ... Is this bug just going to ping-pong between Robert and everyone else forever, or are we going to get a decision for slamm to actually implement?
Again, if managing items were the *primary* task of a panel, then link appearance and single click might not be as well suited as double click. I believe the majority will *access* the data in most panels (like bookmarks, whats related, search results or most contetn panels) more than anything most of the time, just like search results, therefore single click and link appearance are the best educated guess we can make. As far as Netscape 6 is concerned I believe the decision has been locked down. I can't speak for the whole mozilla community though :-)
Taking the hint ... Will raise the issue in n.p.m.ui, and reassign to slamm when the group has come to some semblance of a decision. If that's not what you meant, German, then reassign back to yourself.
Assignee: german → mpt
Status: ASSIGNED → NEW
Whiteboard: [PDT-] → [PDT-] -- to be discussed in n.p.m.ui
Since 90% of the tabs that people use will be web pages which are single click, I vote for single click for mozilla. However, for Netscape branded, we will in general use single-click but make exceptions for particular tabs that require double-click for their functionality. German and I agree on that.
M16 has been out for a while now, these bugs target milestones need to be updated.
any update on this bug
Isn't this fixed with bug 41609 ?
Keywords: beta1 → nsbeta1
spam : changing qa to sujay (New Sidebar QA)
QA Contact: shrir → sujay
FWIW, Robert no longer works at Netscape, and may not even be involved in mozilla.org anymore.
Clearing very old milestone (M16) in hope of reevaluation.
Target Milestone: M16 → ---
cc'ing tpringle, sgehani & adding UI keyword. I don't understand how we still require so many double-clicks in sidebar panels, especially for the dominant operations that should be easiest to do. All the UI owners and tested users were in agreement on this at one point; why does the sidebar still have the deprecated behavior?
Whiteboard: [PDT-] -- to be discussed in n.p.m.ui → -- to be discussed in n.p.m.ui
I agree with Peter here. Frankly IE does this the right way in my opinion. It is a single click to open favorites folders/history folders and a single click (shown as a link) to access the favorites/history sites. This is what most average end users expect I believe. I would also guess that many find the "twisty" concept confusing as well as it is not part of the most common OS, Windows. Plus they take up valuable horizontal space where real estate is at a premium. My only other concern is that we are consistent across the application (i.e. Mail, IM) to the extent that consistency makes sense.
This is fixed, now.
Status: NEW → RESOLVED
Last Resolved: 17 years ago
Resolution: --- → FIXED
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.