Single click vs double-click
Closed, ResolvedPublic

Description

It's clear from the conversation in T7928 that this needs more discussion. Let's do it here:

Advantages of status quo (single-click to open)

  • Makes opening files and folders much easier for users with poor dexterity or mousing skills, laptop trackpads, and touchscreens in particular (where double-clicking is practically impossible).
  • More intuitive; many users on other platforms have never fully understood when they need to double-click, and end up cringe-inducingly double-clicking everything "just to be safe."
  • More consistent with mobile, where double-clicking doesn't exist. Vastly more touch-friendly.

Disadvantages of status quo

  • Unfamiliarity for users migrating from all other desktop platforms, which all select with single click and open with double-click. We discard up to 30+ years of muscle memory.
  • Makes selection much more difficult by reducing the click target to a tiny square in the corner. For touch users, it's literally impossible, since the selection marker only appears on hover, and there is no hover with touch.
  • Makes multi-selection in file open dialogs frustrating, since no selection marker appears and there is no visible method to select multiple files.
  • Makes file save dialogs confusing. Single-clicking on a file suddenly no longer opens it. Why? Not clear. The UX breaks down.

Discussion

Let me mention that I actually really like the single-click setting. I just think it can't work without a lot of engineering effort put into resolving the above issues. Because right now, it's half-baked, halfhearted, and riddled with bugs. If someone proposed it in a patch today, we would block it on the above-mentioned objections, and possibly more.

If we're going to keep the feature used by default, then we need to make it work properly, and resolve the UX issues I detailed above. Mobile OSs that implement single-click-to-open (as well as GNOME apps) typically expose a dedicated "selection mode" that's invoked by pressing a "select" or "edit" button, and while in that mode, buttons for cut copy, paste, delete, and other contextually-appropriate actions become visible. If we're going to stay down this path, we need to be all in; the tiny little green selection button in the corner is not sufficient. We need to do serious usability and engineering work to accommodate common workflows in single-click mode vastly better than we do now.

If we aren't willing to do that or lack the resources, then we should consider it an experimental feature and not have it on by default, and not break people's muscle memory and experience for the sake of exposing them to a buggy, half-baked UX.


Summary of the discussion (including the arguments in from above to have a comprehensive list) - please update, so people don't have to read all comments.

Arguments brought up for single click

  • Single click makes opening items easier, especially for users with poor dexterity or mousing skills or devices that it is hard to double click with (laptop trackpads, touchscreens) (1)
  • Single click avoids confusion about what needs to be double clicked and what needs to be single clicked (leading to some users double clicking everything) (1)
  • Single click is more consistent with mobile touch devices (1) - Argument against it is that mobile devices use a different input method (1)
  • General articles advocating single click (1)

Arguments brougt up for double click

  • People are often used to double click and might have muscle memory (1) - just because they are used to it, doesn't mean that double click is the best (1) - the way how others do things is not considered a valid argument (1) - reference to Jakob's law which says that how others do things should indeed be considered (1)
    • Example list of people complaining (1) - One the other hand examples of Windows users wanting single click (1)
    • Most other systems that use mouse input do rely on single click (including KDE distributions) - Kubuntu changed to double click (1) - an argument against that has been brought up that one should rather look at new users who never used a computer before (1) - comparison with keyboard layout decisions that mostly rely on users who already know the layout (1)
    • When the goal is to get people to migrate, the behavior should be similar to the previous systems, users can then change the setting afterwards (1, 2)
  • Selecting items is harder - one needs to use the small square in the corner (1, 2) - the need of selecting single items in the first place is questioned (1)
    • This is not possible on touch devices (1)
    • Discoverability might be bad (1)
  • Single click is currently not consistently implemented (1) - Things will probably not fixed in the near future, so change to double click until then (1)
    • Qt file save dialogs don't open files on single click but select them (1)
    • GTK file dialogs (like on Firefox) use double click (1) - applications can be told to use Qt dialogs () - that is currently not implemented, though () - that is wrong, it works at least with Firefox in openSUSE (seen on Tumbleweed, likely also on Leap) (1)

General

  • Proposal of having the user choose this in a still to be implemented first start wizard (1)
  • Concern of breaking the workflow for existing users (1) - Important point, but not an argument against a change, since this can be taken care of (1, 2)
  • Idea to have it changed based on the input method used (1)
  • Proposal to include telemetry in this decision (1) - telemety will not tell too much, since it is biased by defaults (1, 2)
    • Link to the current approach to make use of telemetry (1)
There are a very large number of changes, so older changes are hidden. Show Older Changes

I've tortured myself enough. No, double click does not improve usability. My mind cannot override the overwhelming evidence.

Let me suggest a path forward for single-click: Here's an image of GNOME Photos (from the GNOME 3.28 release announcement), which has the kind of dedicated selection mode that I'm talking about:

In that selection mode, the whole icon becomes a click target, and you can see what's selected by the little checkboxes. I think this is a fantastic single-click-centric UI that the GNOME folks should be applauded for. With contextual action buttons that can perform operations on what's selected (cut/copy/paste/rename/delete etc) and some polish work (e.g. select all/de-select all; supporting selection ranges somehow; being able to enter and exit the selection mode with a keyboard shortcut), it would really shine as an actual productivity feature, too. This would IMHO address all usability problems with the current single-click feature and make it 100% touch-ready.

If all of our apps, dialogs, and user interfaces (e.g. Folder View) that supported single-click used a UI like this, I think single-click would be a perfectly acceptable default, and training new users to use it would be not only relatively easy, but also beneficial in the long-run.

If we want to double down on single-click, this seems like a sensible direction to me. Thoughts?

This comment was removed by raddison.
This comment was removed by raddison.
This comment was removed by raddison.
This comment was removed by raddison.

That experiment is not really scientific. Of course Windows users will like it better, no need to test that out.

That experiment is not really scientific. Of course some Windows users will like it better, no need to test that out.

Defaults should be about the best setting overall, they are not the instrument of a conversion campaign to bring in Windows users.

Defaults should be about the best setting overall, they are not the instrument of a conversion campaign to bring in Windows users who were never aware they could switch to single even in Windows and got used to a less-than-optimal default.

@ngraham
@michaeltunnell

If double click is the expected and/or optimal default, then why are a lot of Windows users asking about how to set it to single?

https://answers.microsoft.com/en-us/windows/forum/windows_7-desktop/change-mouse-double-click-to-single-click/bdfd7e3f-964a-4d00-bb9c-7203f523d02f
https://www.windows10forums.com/threads/single-click.326/
https://www.sevenforums.com/installation-setup/266206-changing-double-left-click-open-files-single-left-click.html
https://www.tenforums.com/tutorials/40271-open-items-single-click-double-click-windows-10-a.html
https://answers.microsoft.com/en-us/windows/forum/windows_8-desktop/how-do-i-change-mouse-to-single-click-in-windows-8/230820db-ee4a-4250-bf04-d656ec456dc0
http://www.thewindowsclub.com/how-to-change-double-click-to-single-click-to-open-files-folders-in-windows
(just a few links; I can do better)

Let me mention that I actually really like the single-click setting.

@ngraham
Then why are you overriding what you like/know to be the optimal default?
Also, even in Windows, users have adapted to changes much more radical than a click.
I like what you're doing for Kubuntu but have you considered breaking a known default might trigger opposite-to-expected reactions from current users?
Also, isn't Plasma meant to be cross-device? Why should people use different defaults depending on the device? Isn't a unified default (single tap on mobile and single click on desktop) preferable?

ngraham added a comment.EditedMar 20 2018, 5:49 PM

Let me mention that I actually really like the single-click setting.

@ngraham
Then why are you overriding what you like/know to be the optimal default?

I am trying to avoid the mistake of assuming the superiority of my own opinion. Otherwise it's easy to think your favorite settings are the ones that are best for everyone. Sometimes they might be, but sometimes they aren't. Knowing the difference takes empathy and UX design skills.

The problem is that right now it isn't an optimal default--even though it could be with more work. As I've written, I would fully support single-click if we did the requisite UI work to make it a first-class citizen. Right now we have a UI that's half-baked and simply replaces some problems with different ones. If we want to break consistency with the default setting for every other major platform, we need to make sure that we're presenting our users something that's better in every way, because breaking consistency (even consistency with something sub-optimal) is a high bar.

I am very interested in having that conversation and organizing work to improve single-click, as I believe that is the path forward for keeping the current default. In general, when people are arguing over two choices that are imperfect in different ways, the best long-term solution is to commit to one and improve it to solve its problems.

Also, isn't Plasma meant to be cross-device?

Yes, which is why the current single-click default isn't good enough: it makes selection nearly impossible for touch users. It just replaces one problem with another one.

Okay, so you really implemented double click. It's nonsense. So be it.

ngraham closed this task as Sealed.Apr 6 2018, 9:33 PM

What? No we didn't. This particular discussion pertains to KDE. It's only downstream in Kubuntu where we made that change. My personal conclusion from this discussion is not that we should move to double-click here in KDE, but rather that the UI surrounding and supporting apps that implement single-click could stand some improvement. I think I mentioned this about 10 times in the discussion, including many details as well as at least one screenshot of an app that I consider to be a more user-friendly in its single-click UX.

However, that's a long-term goal, and I personally don't have the time to work on it right now, and I'm not aware of anyone else working on it either.

It's nonsense. So be it.

Please keep the discussion technical.

ngraham changed the task status from Sealed to Invalid.Apr 6 2018, 9:34 PM

Whoops, I mis-clicked.

raddison moved this task from Backlog/Planned to Work in Progress on the VDG board.Apr 6 2018, 9:45 PM
raddison reopened this task as Open.

Okay, so you really implemented double click. It's nonsense. So be it.

Okay, so you really implemented double click in Kubuntu. Technically speaking, it's nonsense. But it's not the end of the world. Seemingly, we're humans after all. Driven by emotions.

Why did you re-open this task and put it in the Work In Progress category? Is there something further you'd like to do with it? If you'd like a task tracking the proposed UI improvements to support single-click, I'd prefer to discuss that matter in a new task; this one has too much prior discussion and baggage attached to it, and it will be hard for people to focus on the proposal instead of the controversy.

raddison removed a subscriber: raddison.Apr 6 2018, 10:08 PM
ngraham closed this task as Invalid.Apr 6 2018, 10:13 PM

Closing because the discussion here is probably not going to yield any more fruit. Let's focus on improving the single-click UX over time.

ngraham added a comment.EditedApr 20 2018, 4:31 AM

The GNOME Folks are considering adding a contextual action bar to Nautilus just like I proposed for Dolphin and Gwenview et al:

https://csorianognome.wordpress.com/2018/04/10/nautilus-adding-an-action-info-bar/

Something like this in our apps that support single-click would go a long way towards making Single-Click a first-class citizen by improving the usability for non-open actions.

Here's the follow-on discussion for how we can improve single-click: T9895: Improving single-click / creation of contextual action toolbars

New user friendliness: double-click must be learned

Yes and they already learned it by using Windows and/or MacOS. (flawed point)

Accessibility for children, the elderly, people with shaky hands, laptop users, and those with poor mouse skills

the people with shaky hands could be a factor but I don't see how the rest is a factor since most people would already be used to the double. The laptop thing is weird because I am not sure how that changes anything but ok.

Consistency with mobile and the web, where there is no double-click

This amuses me because mobile devices don't even use a mouse so the comparison there is rather pointless. (get it . . . no pointer? . . . #dadjokes)

Predictability: no question regarding what needs to be single-clicked and what needs to be double-clicked (we've all known people who double-click everything)

fair point but rarely does it matter if people double click stuff that doesn't need it.


Best of luck for those who insist on single click and hopefully they can find solutions for all the negatives for having it by default. I highly doubt it since it has been over a decade already but :fingers crossed:

clel added a subscriber: clel.Dec 15 2018, 2:28 PM

I came here from https://forum.kde.org/viewtopic.php?f=83&t=135696&p=408438 where I got when searching on previous discussions about making double click the default.

I agree that one should have some good reasons when deciding to behave differently from the rest.
While the discussion here tried to outline some of the reasons to me it still seems not really evaluated. Also I think that it is not enough for single click to simply have the behaviour of multi selections work better as @ngraham suggested, but that this needs to go further like also taking other places into consideration where the behaviour is or can be inconsistent as well as writing some article / wiki / help text to explain the decision to use single click instead of the double click standard.

After a short Google search I came accross two websites that both suggest to get rid of the double click stating that it only was introduced in the first place by Apple to be able to offer a mouse with only one button and naming some negative effects. So there are good arguments for a single click default, but there is still the other side of incompatibility and other things like harder file selection with some of them introduced "by design", so cannot get fixed. The two links:

Maybe this is not the correct place to live up the discussion again but I feel like this should be evaluated from a more objective point to come to a conclusion that can later be used in the proposed article / wiki / help text to explain the way of doing to the users.


Closing because the discussion here is probably not going to yield any more fruit. Let's focus on improving the single-click UX over time.

By all means. T9895: Improving single-click

clel added a subscriber: raddison.Jan 8 2019, 2:55 PM

@raddison If your comment was for me: I did not feel like the linked task represented the topic of my comment, so I went here, although already closed.

@clel No, it was for Mike.

[spam comment removed by sysadmin]

Just my 2 cents: we have telemetry now, so we can actually check whether people prefer single or double click.

clel added a comment.Oct 17 2020, 3:22 PM

That will probably add another piece to the puzzle, although telemetry is also biased and to a certain degree one probably has to go against user preferences if one wants to move forward ;)

apol added a subscriber: apol.Oct 19 2020, 2:50 PM

Just my 2 cents: we have telemetry now, so we can actually check whether people prefer single or double click.

I don't think this would make a big difference. The whole point of changing a default is that people who will benefit from it don't do settings-diving to change such small details. We can lie to ourselves that adding settings makes a difference in user acquisition, but sometimes is just besides the point.

To be honest, single-click default will never be a Plasma selling point and if it's an annoyance for people who come from elsewhere, it sounds like it should be the default. I've been using KDE (back then) and Plasma for too long time to see how it feels for a previous Windows/mac user.

ngraham added a subscriber: davidedmundson.

Looks like it's time to re-open this ancient discussion. :)

It's probably a better place to discuss the matter than an MR anyway (https://invent.kde.org/plasma/plasma-desktop/-/merge_requests/121).

ngraham updated the task description. (Show Details)Oct 19 2020, 3:00 PM
ngraham reopened this task as Open.

Just my 2 cents: we have telemetry now, so we can actually check whether people prefer single or double click.

This is done here: https://invent.kde.org/plasma/plasma-workspace/-/merge_requests/223
It's somewhat stuck.

filipf added a subscriber: filipf.Oct 19 2020, 3:25 PM

Looks like it's time to re-open this ancient discussion. :)

It's probably a better place to discuss the matter than an MR anyway (https://invent.kde.org/plasma/plasma-desktop/-/merge_requests/121).

Do it. Nothing I can say that hasn't already been said, it's all about the weight given to particular arguments. +1

ngraham added a comment.EditedOct 19 2020, 3:48 PM

To help guide the discussion, I think this is basically where we're at:

The best arguments in favor of single-click by default:

  • Single-click is a better behavior even if people have to get used to it
  • Single-click provides a single consistent behavior across desktop, mobile, and desktop-with-a-touchscreen platforms

The best argument in favor of double-click by default:

  • Single-click is not actually better, it's just different; it exchanges one set of annoyances for another
  • Double-click is what everyone is used to from other desktop platforms and we should respect that (Jacob's law)
clel added a comment.Oct 19 2020, 5:20 PM

Despite me never really having used single click, I can see that it might actually be the better concept. However, from the GitLab discussion I have picked up two (I think) good points for double click as default:

  • When you look at user engagement there might likely be a higher conversion rate having double click as default. Still you can make users aware of the nice single click option later (but first you have to make sure they don't leave before that).
  • GTK apps (most prominently Firefox, I guess) use file dialogs with double click that to my understanding cannot easily be changed (and that is a bit of an understatement, I guess). Single click thus introduces inconsistent behavior for similar tasks.

Regarding consistency with touch devices: There has been made the point that mice (or touchpads) make the user aware that interactions are different from using a touch interface. I guess some years in the future the preferred default might acutally change. It is just that currently I don't think enough time has passed that consistency with touch can count as an argument for the majority. To me it looks like KDE was (and still is) ahead of its time. The situation probably will change, but for now I tend to double click as default.

Double-click is what everyone is used to from other desktop platforms

Just because they're "used to it" doesn't make it better. In fact, double clicking is widely inconsistent, based on whether an item is selectable or not. I've observed people double clicking basically anything that looks like a file even though it's not necessary because they don't know and, frankly, shouldn't know the logic behind what determines that. Whereas our approach is simple: click once, it opens.

If any, we need to improve the multi-selection behavior in e.g. Dolphin by providing a multi-selection mode. For instance, when you select one file, you enter selection mode as I think you're unlikely to want to select one file but then randomly launch another icon.

Also, Gnome sabotaging the user experience by forcing double click on their file dialog is not an argument for switching at all. We should rather make absolutely super duper extra sure a user never ever gets to see this hideous file dialog of GTK, by means of XDG portals and the like, and urge distros to enable for example the portal stuff on Firefox.

ngraham added a comment.EditedOct 19 2020, 6:32 PM

Double-click is what everyone is used to from other desktop platforms

Just because they're "used to it" doesn't make it better. In fact, double clicking is widely inconsistent, based on whether an item is selectable or not. I've observed people double clicking basically anything that looks like a file even though it's not necessary because they don't know and, frankly, shouldn't know the logic behind what determines that. Whereas our approach is simple: click once, it opens.

This is true. I've also seen people double-click on everything "just to be safe" because they didn't grok the logic. This is definitely one of those annoyances and quirks to be learned with double-click.

If any, we need to improve the multi-selection behavior in e.g. Dolphin by providing a multi-selection mode. For instance, when you select one file, you enter selection mode as I think you're unlikely to want to select one file but then randomly launch another icon.

Yes, I'd like that too; see T9895. However that proposal is a lot of work and nobody started on it in the two years since I opened that Phab task. That's what I mean when I say that single-click only really works if you re-engineer the world around it to work with it. In its current state, single-click simply exchanges one set of annoyances and quirks that need to be learned for another set. How you open something becomes consistent and a bit easier, but how you select something becomes less discoverable and more annoying.

Also, Gnome sabotaging the user experience by forcing double click on their file dialog is not an argument for switching at all. We should rather make absolutely super duper extra sure a user never ever gets to see this hideous file dialog of GTK, by means of XDG portals and the like, and urge distros to enable for example the portal stuff on Firefox.

This is another example of needing to re-engineer the world around it to reap the benefits of single-click's promise of consistency. With GTK apps, we add an additional quick onto the pile of things you need to learn: the different behaviors not only in GTK file dialogs (which is solvable through portals), but also GTK apps with icon views (which is not) because they won't respect your single-click preference either. So if you use any of those on your system, you need to learn the difference between those and mentally map two different sets of click logic onto the different apps.

This is why I'm personally in the camp of "not actually better, just different" when it comes to single-click. Note that I'm actually using single-click myself!

clel added a comment.Oct 19 2020, 7:56 PM

Just because they're "used to it" doesn't make it better. In fact, double clicking is widely inconsistent, based on whether an item is selectable or not. I've observed people double clicking basically anything that looks like a file even though it's not necessary because they don't know and, frankly, shouldn't know the logic behind what determines that. Whereas our approach is simple: click once, it opens.

I also have seen it and also thought about it as an argument for single click.

If any, we need to improve the multi-selection behavior in e.g. Dolphin by providing a multi-selection mode. For instance, when you select one file, you enter selection mode as I think you're unlikely to want to select one file but then randomly launch another icon.

I don't feel the timeline to make all of that happen is in the near future. But those paths are still worth pursuing. I think realisticly what can be achieved fast with not much effort is making double click the default. If things are ready for single click (and users are maybe more open to it due to them coming more and more from touch devices for example) that can be reintroduced again (also another option requiring a decent amount of work would be the first lauch wizard where one could select the behavior). Maybe I am wrong and everything will be in place much faster. Then from my view one could overthink matters. Currently however, one has consider those points as downsides of single click on Plasma, imho.

Also, Gnome sabotaging the user experience by forcing double click on their file dialog

I hear you insulting GNOME here. I am rather sure that insult is not backed by facts, but is rather your feeling about it? Maybe we can use less strong words when describing that GNOME has disabled the option of single clicking for those dialogs?

is not an argument for switching at all.

It only is not an argument if there is a good way to circumvent problems resulting from this. Users don't care whose fault it is, they just want things to work.

Some simple statements applicable here and in other discussions:

  • Most of KDE users are former Windows users.
  • Windows users are used to the Windows way of doing things.
  • The older people get the less likely it is that they welcome not doing things the way they're used to.
  • The Linux market is small and there are only so many users KDE is fighting for.
  • KDE's major competitor in the Linux market doesn't necessarily offer people the most familiar/friendliest of environments.
  • There is a gap in the market which KDE can exploit in order to increase its userbase.

And it's important to note that being friendly doesn't necessarily mean copying. The notion of friendliness in particular is a lot more applicable here since single click is a setting that deeply influences the desktop experience, and in such a way that it provides unfamiliarity to new users.

It would be better to approach this pragmatically rather than from a developer point of view.

Why? Because I would be willing to bet the setting chased away more users than it ever won.

(Keep in mind that KDE settings are initially very overwhelming and that some didn't just change them before giving up)

mart added a subscriber: mart.Oct 20 2020, 10:09 AM

first thing: this settings should be published by kuserfeedback and see how many actually go to the lengths of changing the default (and have this data point independent of those from distributions that come with the default changed)

That said, wall of text incoming:
To me, "others do this way" is *never* a valid argument.
Advantages and disadvantages have to be by themselves. Double click is a thing people do struggle with, and i have seen people double opening things *on windows* double clicking taskbar launchers, start menu items and things in the browser.
In fact, places where double clicking is actively used are increasingly smaller and smaller, as they are pretty much stuck where they always were since 30 years: icons in a file manager and icons on desktop. nowhere. else.

Almost no new interaction devised since then uses that: is pretty much not existent in anything web based, is pretty much not existent on mobile, and used but very rarely in complex creation applications for not scrictly needed secondary actions, where is usually associated to an action of selection, rather than triggering (select whole word in text editor, select whole solid in cad apps and stuff like that)

Avoiding double click as much as possible, is a better design pattern by itself (Bruce Tognazzini also acknowledged it was not a right choice back then designing the first Mac due mostly by the artificially imposed limitation of wanting a mouse with a single button), modulo the argument of users being used to other things...
I actually can see double click if not disappearing being dialed even further down on other platforms. it's understood that's it's mostly horrible legacy, and young people aren't really "used" to it (it's what that cumbersome platform uses that they have to grudgingly use when they have to do something actually productive), i do expect new generations of designers on those platforms, even internal of ms/apple coming up with things to make it less necessary

to conclude, some reads:)
https://www.nngroup.com/articles/when-to-open-web-based-applications-in-a-new-window/
https://www.joelonsoftware.com/2000/04/27/designing-for-people-who-have-better-things-to-do-with-their-lives-part-two/
https://medium.com/@aureliensalomon/lets-get-rid-of-double-clicking-apple-microsoft-please-do-something-328f77c9be31

Those are good arguments.

michaeltunnell added a comment.EditedOct 20 2020, 2:52 PM

KDE frustrates me. KDE makes the most powerful and technologically valuable DE on Linux yet it's not used by the vast majority of Linux users. Why is that? Well to me the reason is obvious, KDE makes a three-year goal to improve onboarding and ignores all of the problems with onboarding of Plasma. The process to switch to Plasma should be painless and frictionless but instead it is notoriously known as being overwhelming and confusing.

To quote someone who tried Plasma and left within minutes: "why would I use this, they can't even get the mouse to work right?"

Windows is 80% of the market = double click, Mac is 18% of the market = double click, Linux is 2% of the market . . . 70% of that 2% marketshare is likely Ubuntu using GNOME. Then you have PopOS with GNOME, Manjaro with Xfce, Fedora with GNOME, openSUSE with GNOME, Linux Mint with Cinnamon, Red Hat with GNOME, etc . . . all of these are Double Click by default. So I think its fair to say roughly 99% of the computing market using Double Click but that's still not enough somehow.

99% of computing is double click and yet I have had this conversation with dozens of KDE people over the years with same result = play the greatest hits of arguments and then refuse to change it.

Greatest Hits =

  1. "Mobile doesn't double click"
    • true, also mobile doesn't use a mouse AT ALL so that's an irrelevant point.
  2. "young people just use phones these days, they don't know double click"
    • let's say that's true, how does that matter? Microsoft can't get them to use Windows what makes you think they will use Linux much less a KDE flavored Linux (remember not the default recommendation)? In my opinion, that argument is irrelevant on face value.
    • now, this is not true because look at the insane amount of popularity there is for Game Streaming and PC Gaming . . . these people are using Windows 99% of the time, they know how double click works. This is irrelevant on face value because the logic doesnt even work and it's also not accurate anyway.
  3. "web browsers don't use double click so its inconsistent"
    • if you double click on something that doesnt respond to double click what happens? NOTHING. It might slightly delay the response from clicking the link but to the average user, it makes no difference. Web Browsers do NOT respond to double click at all and thus this point is irrelevant.
    • With single click on by default, If you double click, which is the expected behavior by MILLIONS of people, then you accidentally launch things you didn't intend to and you may launch the same thing multiple times. This is a clear usability problem.
  4. "Single click is superior method"
    • No, it isn't. If it were then when Windows tried about 20 years ago they would've kept it. They had a ton of backlash and switched it back to double click.
    • It can not be superior when you have to train someone how to do something in contrast to what they expect.
    • It is true that you have to train people double click as well but there has been 35+ years of people already trained for this behavior. The superiority argument is irrelevant when you are battling a fundamental usability structure with decades of training.
  5. "really old people don't understand double click"
    • I'd say this is probably true that most don't however I'd also like to point that its also probably true that most don't use a computer or have any intention of doing so. Just like many people turn their nose at TikTok these days for "being for teenagers" . . . a similar mentality that older people had for computers when they arrived. For example, how many people here reading this have EVER installed and tried TikTok? I don't expect the number to be high.

I think you get the point . . . the arguments are not about usability or onboarding but always the arguments are "this is superior" ignoring what make superior usability design is . . . removing burden from users.

KDE chose to use Single Click by default in the early days of KDE 2.0 and alright fine, you tried something. It's been 20 years, it has been a burden on so many people who consider KDE (then and now) to be confusing and overwhelming. I found posts online dating back to 2004 infuriated by this default setup to the point some made articles about how to "fix it". The experiment has had 20 years of testing and it has not changed the user behavior paradigm at all. In fact, it never had a chance to change it in the first place.

It's time to go back to double-click.


all of the above comments are not related to any particular comment on this thread because they are very commonly made arguments but the following is a response to a specific comment.

@mart

Avoiding double click as much as possible, is a better design pattern by itself (Bruce Tognazzini also acknowledged it was not a right choice back then designing the first Mac due mostly by the artificially imposed limitation of wanting a mouse with a single button), modulo the argument of users being used to other things...

It might be a better design pattern than double click but KDE is not going to convince the planet to change. KDE doesn't have any level of power to initiate change among the general populous of computer users. KDE doesn't even have a level of power to change the usability of Linux users. KDE has been trying this for 20 years and still . . . 95% of Linux users use double-click. I hope one day KDE can get influential enough that maybe then this "better design pattern" can be implemented but that day, is not today or any day preceding it.

I actually can see double click if not disappearing being dialed even further down on other platforms. it's understood that's it's mostly horrible legacy, and young people aren't really "used" to it (it's what that cumbersome platform uses that they have to grudgingly use when they have to do something actually productive)

On other platforms? who will be doing that dialing in?

"young people arent really used to it" based on what metric is that stated? Windows users and Mac users are all presented with Double Click by default . . . where do you even get that claim from?

i do expect new generations of designers on those platforms, even internal of ms/apple coming up with things to make it less necessary.

Why do you expect this? Microsoft and Apple have 30 years of usability and Microsoft is very much locked to backward compatibility so to change a fundamental user experience navigation behavior . . . very unlikely.

In my opinion, Microsoft and Apple have zero incentive to change this and millions of reasons not to. (you know the millions of people who will be mad because they would be changing a fundamental thing of computing)


I know nothing will come from this discussion but I wish that weren't the case but after years of discussing this with KDE, I've accepted that growing the userbase is not important, so it will likely always be the case that Plasma is the most powerful, featureful and useful DE that is used by a small percentage of a small ecosystem.

I have been using KDE Plasma exclusively for 6 years now and I am a big fan but I also recognize where it falls short. This is one of those things and the mentality of superiority over user expectation applies to more than just this topic. This is the approach used by KDE often. This is why KDE (/Plasma) has been around as long as GNOME yet GNOME has default choice on all the major distros (well the distros that make a choice).

GNOME does battle user expectations but the difference is that GNOME doesn't battle users on fundamental pieces like how a mouse should work.

clel added a comment.EditedOct 20 2020, 2:53 PM
In T8187#243395, @mart wrote:

To me, "others do this way" is *never* a valid argument.

The underlying argument of "others do [it] this way" is "users are used to it" and "there might be a good reason for it". For the second one I would agree with you, one can hopefully directly state those good reasons. But for "users are used to it": You are basically saying Jakob's Law (https://lawsofux.com/jakobs-law) cannot be used as a argument ever. I don't think this is true at all. Let me give a more drastic example to better demonstrate this: Think about keyboard layouts and the history involved with them. One now could say that technically there are obviously better layouts than the QWERTY layout that users are used to and discuss to offer those as default for KDE Slimbook, Plasma mobile etc. One can also use the argument there that people who never used a computer before will probably benefit of it and that the historical reasons for it are not justifying the old way to be still used (both arguments are brought up against double click here). (Technically with a different keyboard layout there is only a downside of shortcuts not working out that well, but I guess that can easily be overcome with changing keyboard shortcuts for KDE as well to match that layout).

gikari added a subscriber: gikari.Oct 20 2020, 3:13 PM
In T8187#243302, @clel wrote:

Regarding consistency with touch devices: There has been made the point that mice (or touchpads) make the user aware that interactions are different from using a touch interface. I guess some years in the future the preferred default might acutally change. It is just that currently I don't think enough time has passed that consistency with touch can count as an argument for the majority. To me it looks like KDE was (and still is) ahead of its time. The situation probably will change, but for now I tend to double click as default.

For me that is an argument to not switch because in $timeframe you likely want to switch the default again. Also debatable when the time will have come or maybe if it even has already come.

Michael's made a good point about the longevity of this default setting. If it's been set to single click for 20 years and the majority (?) of the users still use double-click then it's ultimately a failure, regardless of whether or not the engineer arguments are valid.

Regarding the (?), I've made a Reddit poll just for exploratory purposes. The sample will not be representative, but telemetry won't fare better in this regard anyway.

No, it isn't. If it were then when Windows tried about 20 years ago they would've kept it. They had a ton of backlash and switched it back to double click.

This does not follow. They had backslash because they switched something not because double click is objectively better.

  • It can not be superior when you have to train someone how to do something in contrast to what they expect.
  • It is true that you have to train people double click as well but there has been 35+ years of people already trained for this behavior. The superiority argument is irrelevant when you are battling a fundamental usability structure with decades of training.

As you said people also have to train double clicking. And double clicking is even more complicated because some things respond to double clicking but other only to single click. So you need to know where to apply which technique.

From the second part of the argument I guess we can never improve fundamental user interactions/differ from Windows in your opinion?

Why do you expect this? Microsoft and Apple have 30 years of usability and Microsoft is very much locked to backward compatibility so to change a fundamental user experience navigation behavior . . . very unlikely.

In my opinion, Microsoft and Apple have zero incentive to change this and millions of reasons not to. (you know the millions of people who will be mad because they would be changing a fundamental thing of computing)

So the only reason they are not changing it because of legacy and not because it is better, then we should not blindly follow them. Also just because they did not change in 30 years doesn't mean that what they are doing is the best usability one can achieve (certainly the case when looking at Windows).
Also what about our 20 years of legacy? Should we make our users angry just for the potential that some new users don't have to change the setting or get used to single click (which I believe one notices after the first few clicks).

I know nothing will come from this discussion but I wish that weren't the case but after years of discussing this with KDE, I've accepted that growing the userbase is not important, so it will likely always be the case that Plasma is the most powerful, featureful and useful DE that is used by a small percentage of a small ecosystem.

Don't you think that's a bit hyperbolic? Just because people think there are good arguments for single click doesn't mean they do not care about increasing the number of users, I even think that almost everybody is happy about each additional user of Plasma.

I have been using KDE Plasma exclusively for 6 years now and I am a big fan but I also recognize where it falls short. This is one of those things and the mentality of superiority over user expectation applies to more than just this topic. This is the approach used by KDE often. This is why KDE (/Plasma) has been around as long as GNOME yet GNOME has default choice on all the major distros (well the distros that make a choice).

In fact even longer ;)

to @clel

Your point about the QWERTY keyboard is a great example of usability through user expectation.


@filipf

If it's been set to single click for 20 years and the majority (?) of the users still use double-click then it's ultimately a failure, regardless of whether or not the engineer arguments are valid.

This is the point exactly yes. KDE has insisted on the superiority approach and it hasn't worked yet.


@davidre

For me that is an argument to not switch because in $timeframe you likely want to switch the default again. Also debatable when the time will have come or maybe if it even has already come.

Don't switch it to what users expect because someday their expectations might change? User expectations should be met or exceeded not unsuccessfully challenged for decades.

This does not follow. They had backslash because they switched something not because double click is objectively better.

this is the problem, KDE often argues about which click method is "objectively better" rather than what should be debated. "Which is objectively better: fighting user expectations or making enjoyable to use the software".

This is where the disconnect happens. The default response of KDE has always been "single click is superior" but the people advocating for double click are almost never, in my case absolutely never, arguing for it due to "superiority". The argument for double click is to eliminate burden of users especially average/beginner users. KDE Plasma is often the first time people even experience single click for navigation so in addition to switching away from their comfort zone of Windows or Mac, they also have to deal with losing the comfort of knowing how a mouse works. This is the problem.

As you said people also have to train double clicking. And double clicking is even more complicated because some things respond to double clicking but other only to single click. So you need to know where to apply which technique.

I said they have to be trained AND that they have been training on this method for 35+ years. Its not about what is better, its about meeting and exceeding user expectation where as now KDE is battling user expectation.

From the second part of the argument I guess we can never improve fundamental user interactions/differ from Windows in your opinion?

This is a strawman. I never said that things cant be improved on but there are some things that "improvement" is actually counter-intuitively detrimental. Trying to change the way a mouse functions with it likely being the first time they experience this setting will not change their minds but create a barrier for people to look no further than that.

clel added a comment.Oct 20 2020, 3:47 PM
In T8187#243302, @clel wrote:

Regarding consistency with touch devices: There has been made the point that mice (or touchpads) make the user aware that interactions are different from using a touch interface. I guess some years in the future the preferred default might acutally change. It is just that currently I don't think enough time has passed that consistency with touch can count as an argument for the majority. To me it looks like KDE was (and still is) ahead of its time. The situation probably will change, but for now I tend to double click as default.

For me that is an argument to not switch because in $timeframe you likely want to switch the default again. Also debatable when the time will have come or maybe if it even has already come.

It is a bit unfortunate, I agree and one could argue that users might be confused by switching it back and forth. However, I think currently even with it being likely to be changed later again (but probably not in the near future, I fear) currently one has to focus more on the status quo (and that is likely that users are driven away by this default and how it currently works). I am very open to change it back properly, when the time is "right", though. And I also agree that it is hard to know when that time will be and views will be different on that as well. I don't see that as an argument that double click shouldn't be made default, though.

No, it isn't. If it were then when Windows tried about 20 years ago they would've kept it. They had a ton of backlash and switched it back to double click.

This does not follow. They had backslash because they switched something not because double click is objectively better.

I am not into the facts of this, but I don't think it really matters what objectively is better, but what is better subjectively from the user's perspective (to a certain degree). See my keyboard layout example. I think that is a pretty good one showing that objectively a different layout might be better, but the users prefer QWERTY. (Or one could say that you use "objectively" in a different way and ideally when assessing which things are "better" one obviously has to take into account the opinion and subjective feelings of users that are in contact with it.

Finally:

If we believe single click is better than we should keep it as default. It makes no sense to have the better option off and the worse option on by default. Otherwise only people who already used single click because it was the default in the past will enable the thing we believe to better and everybody else will stay on what we think is the inferior setting.

ognarb added a subscriber: ognarb.Oct 20 2020, 3:48 PM

I use single click and I like it, but I wouldn't mind changing the default to double click if and only if we don't change the behavior for existing users. People currently using single click shouldn't need to change their behavior again. This was something that I didn't like with the move from Alt to Meta shortcut for Kwin. Let's not break the workflow from our existing users.

clel added a comment.Oct 20 2020, 3:48 PM

Finally:

If we believe single click is better than we should keep it as default. It makes no sense to have the better option off and the worse option on by default. Otherwise only people who already used single click because it was the default in the past will enable the thing we believe to better and everybody else will stay on what we think is the inferior setting.

Please take a look at the keyboard layout example. This shows pretty well (I think) why it does not count what "we" believe is better, but also what the users think.

@ognarb I think it is reasonable to make it so people who currently like single can keep it and that is should be how it is for them.

There is one extra note though that the people who like single click understand there is a difference and can change it vs the beginner user never seeing single click before having no idea what is happening and are just confused. So if KDE Plasma can not keep the setting for single click for those who want it for any technical reaso then I still think that it should be changed for double because the experienced or intermediate user having the burden of change is much more reasonable than the burden being on the beginner.

clel added a comment.Oct 20 2020, 3:57 PM

I use single click and I like it, but I wouldn't mind changing the default to double click if and only if we don't change the behavior for existing users. People currently using single click shouldn't need to change their behavior again. This was something that I didn't like with the move from Alt to Meta shortcut for Kwin. Let's not break the workflow from our existing users.

This requirement has at least been mentioned by @ngraham on the MR not taking care of that, so I think it will be made sure that this won't be changed for existing users.

@ognarb I think it is reasonable to make it so people who currently like single can keep it and that is should be how it is for them.

There is one extra note though that the people who like single click understand there is a difference and can change it vs the beginner user never seeing single click before having no idea what is happening and are just confused. So if KDE Plasma can not keep the setting for single click for those who want it for any technical reaso then I still think that it should be changed for double because the experienced or intermediate user having the burden of change is much more reasonable than the burden being on the beginner.

Please watch https://www.youtube.com/watch?v=u8SI6AEwMnQ&t=8886s. Breaking the workflow of existing users is not ok, even if those are long time plasma users and more advanced (whatever this means).

Please watch https://www.youtube.com/watch?v=u8SI6AEwMnQ&t=8886s. Breaking the workflow of existing users is not ok, even if those are long time plasma users and more advanced (whatever this means).

I think there is a bit of a misunderstanding here. I wasnt saying this would be okay but that it would be the lesser of the 2 negatives. The double clicking action, specifically on desktop & file manager actions (not everywhere), is something that is expected by millions of people. If this change created an irritation to the experienced users on a temporary basis, this would be a better result than the usage barrier that beginners unfamiliar with single-click paradigm experience in the current setup.

I wasn't saying that its ok to break paradigms and that I completely agree that if at all possible to reserve that configuration for existing users then it should be. The point I was making that irritation for experience users who are familiar with the option of single-click paradigm should not supersede the user experience of beginners who are presented with something unfamiliar and arguably frustrating for them.

It's a moot point as if we change the default, we can easily keep the current/old behavior for existing users and only apply the new defaults for new installs.

It's a moot point as if we change the default, we can easily keep the current/old behavior for existing users and only apply the new defaults for new installs.

that's awesome!

clel updated the task description. (Show Details)Nov 18 2020, 3:23 PM
clel added a comment.Nov 18 2020, 3:29 PM

I finally got around adding a summary of most arguments brought up in the discussion now. Please update it when things change, in case I have forgotten things or when I misrepresented comments. That way people can easily see the current status and ideally don't bring up similar arguments again in the discussion which only makes it harder to follow.

The summary includes the arguments brought up in the description itself, since some didn't appear in the comments and I wanted to have an exhaustive summary of as many points as possible.


Now that we have the summary, I still don't really see progress with this. I seriously wonder how that could be achieved.

In T8187#244981, @clel wrote:

Now that we have the summary, I still don't really see progress with this. I seriously wonder how that could be achieved.

Send a MR ;)

That one has been closed due to lack of prior discussion, so now that the discussion has been had, someone could submit a new one (that one can't be re-opened since @The-Feren-OS-Dev has since deleted the source branch in his fork).

clel added a comment.Nov 20 2020, 11:39 PM

Hm. I think that currently we are at a state where probably most arguments have been brought up, but still the views (most likely the weighting of the individual arguments) are still different on this whole topic (meaning there is no consensus currently).

I can try to prepare a new MR, though, I guess @davidedmundson gave a hint for where to look (https://invent.kde.org/plasma/plasma-desktop/-/merge_requests/121#note_92482).

clel added a comment.Nov 21 2020, 1:20 AM

I created https://invent.kde.org/plasma/plasma-desktop/-/merge_requests/236 now (currently making the identical change as @The-Feren-OS-Dev). So this probably needs some work, help is very welcome as after reading some documentation I still don't know how to achieve this :)

Just as a data point, my 8 year-old son was annoyed by having to double-click to open something on Windows yesterday and asked "why do I have to do it twice?"

I explained that it was to make selection easier, and he gave me a sarcastic look. :)

It almost does seem like we'd optimally have different paths and settings for people who migrate from another OS (double-click) and people who are using Plasma as their first experience (single-click).

clel added a comment.Nov 23 2020, 4:03 PM

:)

I guess that integration into a yet to come first run wizard will be the best in the future.

clel added a comment.Jan 18 2021, 7:25 PM

Just adding two things I noticed some time ago, but did not get around to adding here:

Makes file save dialogs confusing. Single-clicking on a file suddenly no longer opens it. Why? Not clear. The UX breaks down.

This is true, but it should be noted that single click still opens folders (just adding for clarification, as I think I expected a different behavior).

More importantly, I noticed that openSUSE (Tumbleweed) already has support for KDE file dialogs in Firefox. Not sure how feasible this is for other distributions but at least it is available somewhere in production, which I was unaware of before.

clel updated the task description. (Show Details)Jan 18 2021, 7:28 PM
clel updated the task description. (Show Details)
ngompa added a subscriber: ngompa.Apr 10 2021, 7:42 PM

openSUSE's support for KDE file dialogs is based on a patch that has been rejected by Mozilla for its entire existence. Unless someone wants to get that patch upstreamed, it's never going to mean anything for anyone else.

openSUSE's support for KDE file dialogs is based on a patch that has been rejected by Mozilla for its entire existence. Unless someone wants to get that patch upstreamed, it's never going to mean anything for anyone else.

As of Firefox 64 you can use KDE dialogs via xdg-desktop-portal without having to patch Firefox.

ngraham closed this task as Resolved.Oct 11 2023, 9:12 PM

We've switched to double-click by default in Plasma 6; closing.