Bug 70506

Summary: In web-browsing mode, the up button should be removed
Product: [Applications] konqueror Reporter: Carlos Woelz <carloswoelz>
Component: generalAssignee: Konqueror Developers <konq-bugs>
Status: RESOLVED INTENTIONAL    
Severity: wishlist CC: esigra, james, michael.jahn, roland
Priority: VLO    
Version: unspecified   
Target Milestone: ---   
Platform: Compiled Sources   
OS: Linux   
Latest Commit: Version Fixed In:
Attachments: Mozilla navigarion bar

Description Carlos Woelz 2003-12-15 16:44:13 UTC
Version:            (using KDE Devel)
Installed from:    Compiled sources
Compiler:          GCC 3.3.1 mdk
OS:          Linux

Small usability bug:

The up button in web-browsing mode is very seldom used, and should be removed, as in many other browsers.
Comment 1 Maksim Orlovich 2003-12-15 16:49:42 UTC
This is not a bug, but a wishlist, and I vote against this
Comment 2 Carlos Woelz 2003-12-15 16:59:20 UTC
Ok. Let's argument a bit:

The up button is usefull for reading docs in html, for browsing directories, etc...) But in websites in general, they are not very usefull.

So for directories, one could use the filemanager mode. For reading dos, well, life is full of choices. I think this button is not used by most joe users in web-browsing mode.Power users can allways add it if they read structured html docs online frequently. Any other reason to keep it?
Comment 3 Thiago Macieira 2003-12-15 17:50:34 UTC
I don't agree either. I find it a very nice feature of Konqueror to allow me to go "up" in the websites, however seldom I use that. Besides, the first six or nine icons in Konqueror are the same in every view mode.

What's more, I don't think there's a way of easily adding/removing that button according to the view mode.
Comment 4 Carlos Woelz 2003-12-15 18:00:09 UTC
Thiago, tudo bem? ;)

Comment #3 From Thiago Macieira  2003-12-15 17:50 
>What's more, I don't think there's a way of easily adding/removing that button
>according to the view mode. 

If this is true, we should close this wish as remind, and I will open a wish for configurable toolbard between views.
Comment 5 Thiago Macieira 2003-12-15 18:15:44 UTC
Konqueror has a <Merge> in the main toolbar, which merges the toolbar buttons offered by the current KPart being displayed. Web browsing uses the KHTMLPart, while the dir listings use KonqDirPart.

So, while it is already possible to change the contents of the toolbar, it's not possible to move items around so as to place the Up button where it is now.

If you want to file a new wishlist asking for that ability for KDE 3.3, you're welcome to do so. (KDE 3.2 is feature-frozen already)
Comment 6 Carlos Woelz 2003-12-15 18:28:55 UTC
So this is impossible right now. Removing it from all views would be a big regression.

I will open a wish for KDE 3.3, and please close or not this bug as you see fit.
Comment 7 Stephan Binner 2003-12-15 22:54:34 UTC
One more vote against removal in any view. :-)
Comment 8 Adrian Head 2003-12-16 18:32:48 UTC
Another vote against removal of the UP button in the HTML view.
Comment 9 Carlos Woelz 2003-12-17 06:33:26 UTC
The up button is confusing for the average user, because in average it don't deliver meaningfull results, in special for more complex websites. This is the rationale for removing. To add a bit of data, I did some homework.

Notice that I think this button is a keeper in filemanagement mode, even when you open web pages. So still can quickly use konqueror for browsing html docs, etc.. in filemanagement mode.

------

Small analysis of the state of hierarchical web sites.

Method: I tried first non geek sites: shopping.yahoo.com, amazon.com and aol.com. Clicked three times in random items. After that, I tried two sites I use a lot: newsforge.com and uol.com. Then I tied three random sites at google.

The result: 

Site: Yahoo Shopping
Works: No
 http://search.shopping.yahoo.com/search/__yltc=s:92253310,d:14489115,sec:search,slk:button?p=.com&did=51

Site: Amazon
Works: No
http://www.amazon.com/gp/browse.html/ref=sc_fe_c_1_0/103-2514953-1839064?%5Fencoding=UTF8&node=1046284&no=1036592&me=ATVPDKIKX0DER

Site: InSyle magazine (found from aol)
Works: No
http://cgi.timeinc.net/cgi-bin/instyle/cg/clogin?zone=IN&cookie=&stat=CHA&url=http%3a//www.instyle.com/instyle/read/transformations/0,12403,557876,00.html

Site: Newsforge
Work: No
http://www.newsforge.com/software/03/12/16/0029234.shtml?tid=130&tid=150&tid=2&tid=82&tid=94

Site Universo on Line: (Brazilian newspapers/magazines site)
Works: yes, for the Folha de S
Comment 10 Thiago Macieira 2003-12-17 17:07:37 UTC
Someone at osnews posted a solution I didn't know of either: even though you can't reorder buttons around, you can completely change the toolbar for different profiles. That is, (and this is what I didn't know) the toolbars are defined per-profile.

In other words, you can already remove the button. If it should be removed by default in the webbrowsing mode, I can't answer.
Comment 11 Sashmit Bhaduri 2003-12-17 19:47:11 UTC
Perhaps it should act like "Rewind" does in Opera 7.x or Snapback does in Safari. Up buttons aren't really meant for http UAs. 
Comment 12 djoham 2003-12-17 19:52:31 UTC
+5 Great Idea! I think that would be an elegant solution to the discussion...
Comment 13 George Staikos 2003-12-17 22:36:55 UTC
Can we make it possible to apply negative votes to a bug?  If you don't like a toolbar entry, edit it out.  Alternatively, it could be made edit-in, but we need it in filebrowser mode anyway (without exception it must be there).  Let's let consistency rule here.  If you don't need the button, don't click on it, or remove it.
Comment 14 Carlos Woelz 2003-12-17 23:01:14 UTC
George,

>Let's let consistency rule here.
Good point. I agree that consistency is important. Also, as I said before, the button in file managing mode is a keeper. So if it means that you have to remove it from file management to have it working in web browsing, then is better forget about it and file a wish for allowing different main toolbars for different view modes. I did that.

http://bugs.kde.org/show_bug.cgi?id=70516

But consistency is not everything: that is why we have different view profiles and even different icons for launching them! We can have different toolbars.

>If you don't need the button, don't click on it, or remove it.
Well, I don't agree in general with this argument. I don't get it, please elaborate. With this argument, you would never remove _any_ button. It is not personal, I like the up button. I use it a lot browsing html docs. I would even add it. But my wife find it very confusing, as it does not give you meaningful results in most sites.

So the question to ask is the opposite: is it really usefull for the average user? Or the fundamental question that helps the answer others is: who do we target as users? I will write something about it and a proposal to kde-usabiliy list.
Comment 15 Leo Savernik 2003-12-18 14:05:38 UTC
I also cast an anti-vote onto this bug.

Generally, confusion is not a good reason for removal. Otherwise, a user could complain to the keyboard manufacturer that he is confused by the Scroll key, and demand its removal.
Comment 16 Carlos Woelz 2003-12-18 23:06:39 UTC
Manuel Amador added an interesting comment on kde-usability list:
Thanks for the study, carlos!

What would happen if the functionality of the Up button was to "Go to
the upper page specified as LINK REL='up' tag"? 

To test this, I hear konqueror has a "navigation toolbar" whose specific
purpose is to have an "Up" button which does this. 
Comment 17 Carlos Woelz 2003-12-18 23:29:21 UTC
As you can see, this did not improve much the odds.
But his check made me think a bit about the up button:

The up button looks similar from the left and right buttons. But they perform very different tasks. The up button navigates in the tree structure, as the left and right follow the trail of visited urls.
In fact, the up button has more in common with LINK REL RIGHT and LEFT (let's call them navigating buttons) than with the left as right trail buttons as we use them (trail buttons).
Imagine that we would like to create in the merge toolbar the up, right and left navigating buttons. It makes sense creating them as we detect the LINK REL on the pages. Then it would be the perfect way to navigate html structured docs. We would want to make the icons look different, otherwise we would confuse them with the trail buttons. And we would end up with two up buttons looking differently and doing the same thing. See?

I said that I liked this button, and only wanted to change it because of the usability for the average user. I changed my mind. I want to change it because of that _and_ because it does not make logical sense to mix navigating and trail buttons. And more: the up button should be graphically different from the left and right buttons to state that it has a different function. This way you could implement the merge navigation toolbar if you wanted. (And it is a nice addition).
Comment 18 Frans Englich 2004-01-06 14:52:12 UTC
Subject: Re: simpler UI for konqy

On Tuesday 06 January 2004 14:23, Carlos Leonhard Woelz wrote:
> On Tuesday 06 January 2004 00:34, Aaron J. Seigo wrote:
> > On Monday 05 January 2004 07:52, Gustavo Sverzut Barbieri wrote:
> > > As discussed here in kde-usability, maybe the "Up" button should also

My little thoughts on the subject:

The button has severe usability problems: For anyone who doesn't have a 
technical understanding of the media(ie. html) the behavior is inconsistent -  
sometimes it works, sometimes not. And that is pretty bad. To put in a 
practical situation, Ofcourse a lot of developers like it because we know 
when to use, hence we don't find it confusing. But a typical user don't.
Having the button enabled by default follows with the view that the /majority/ 
of KDE's userbase(and future's) grasp the button and how use it(and ofcourse 
will). And that means <strike>Desktop for the masses</strike>.

	
			Frans

Comment 19 Tom Chance 2004-01-06 16:26:41 UTC
I agree with Frans. In fact, it isn't even a matter of majority here, since even if it were 60/40 in favour of keeping it, those 60 percent would be more than capable of adding the button, whilst many of the 40 wouldn't know to take it off.

As for the consistency argument, I don't think most users think of browsing the web and browsing for files as the same kind of operation. In my experience, a lot of people are surprised even to find out that a web site is a collection of files on a remote server! Konqueror ought to adapt itself to be most suitable for the operations that the view profile demands, and so we shouldn't have to worry about a few toolbar buttons disappearing between file and web browsing.

If you compare the current (3.2) default Konqueror toolbar to Mozilla, Firebird, Safari, Internet Explorer, Opera and just about any other browse, Konq features FAR more buttons. Are all the other browsers wrong, or is Konq? I think Konq is, and it really needs fixing. I've seen both of my parents and my girlfriend be confused by the up button. I've now taken to completely customising Konq's toolbars on every KDE set-up for average users (i.e. not techies) to make it usable.

I use the up button myself, but I customise my toolbars as soon as I start using the app, so it'd hardly be a problem for me if it weren't there by default.
Comment 20 Roland Seuhs 2004-01-06 19:53:46 UTC
I also vote against removal.

Several reasons:

1) Some webservers operate on Apache's default navigation style, which tries to mimic a filemanager.
2) To go to the homepage, the "up" button works, and works *always*. As such it is very useful, for example on bugs.kde.org I can go to the homepage from anywhere, not matter how far I've scrolled down. This is useful on a lot of webpages - of course the use is apparent only after using Konqueror for a while. And that's exactly why it should stay there - To make this useful feature known and not forgotten.
3) Constantly changing icons is a bad idea anyway

and now the MOST IMPORTANT reason:

4) To attract users, KDE has to do things BETTER than the rest. To remove functionality just because nobody else has it, is a stupid idea, especially when it's used so often (as seen by the number of votes against removal). IIRC, Google's very popular bar for IE also includes an "up" button, btw.

"I agree with Frans. In fact, it isn't even a matter of majority here, since even if it were 60/40 in favour of keeping it, those 60 percent would be more than capable of adding the button, whilst many of the 40 wouldn't know to take it off."

That's just plain wrong. Removing an icon is easy and fast - takes less than a minute. To *add* an icon, you have to do:

- Know that the feature/icon exists. So you either have to manually try out all icons or get the tip from some 3rd party.
- Search the icon in the toolbar-editor. Since there are a lot of icons there, and because icons can't be sorted in a meaningful way (the alphabetical sorting is often not useful because the user doesn't know the action's name) it's time-confusing and error-prone.
- Then think about where to put it and move it up/down, apply settings to see it in Konqui, maybe fine-tune afterwards.

Removal on the other hand is pretty straigtforward: No additional information is needed, no searching is involved and no fine-tuning neccessary.

Comment 21 Tom Chance 2004-01-06 20:09:21 UTC
> "I agree with Frans. In fact, it isn't even a matter of majority here, since > even if it were 60/40 in favour of keeping it, those 60 percent would be 
> more than capable of adding the button, whilst many of the 40 wouldn't know 
> to take it off." 
 
> That's just plain wrong. Removing an icon is easy and fast - takes less than 
> a minute. To *add* an icon, you have to do: 
 
> - Know that the feature/icon exists. So you either have to manually try out 
> all icons or get the tip from some 3rd party. 
> - Search the icon in the toolbar-editor. Since there are a lot of icons 
> there, and because icons can't be sorted in a meaningful way (the 
> alphabetical sorting is often not useful because the user doesn't know the action's name) it's time-confusing and error-prone. 
> - Then think about where to put it and move it up/down, apply settings to 
> see it in Konqui, maybe fine-tune afterwards. 
 
> Removal on the other hand is pretty straigtforward: No additional
>information is needed, no searching is involved and no fine-tuning neccessary. 

What you say is true, but not the whole story. It seems that the general opinion is that most techies and KDE developers like the up button, but that most "average" computer users interested in simply browsing the web in the style they are accustomed to find it needlessly confusing.

Really, just seeing a toolbar with different icons and new buttons can *completely* put off a new user, so it's not just a matter of them being slightly confused, as some have suggested. Techies, OTOH, are unlikely to be confused and put off by a lack of a button they might find useful.

Assuming they get over their confusion, few average users would think to right click on a toolbar, click on "configure toolbars" and use the unavoidably difficult-to-use interface to set the toolbar to their liking. Techies, OTOH, would be far more likely to do this, and even though it takes slightly more clicks, would be both capable and likely if they knew it was there. If you're going to argue that they might otherwise miss the up button feature, we might as well put every button on the toolbar just in case.

So all decisions in usability seem quite clear to me. The default interface MUST be acceptable for average and illiterate newcomers without making it unusable for the techies, and the apps must provide the configurability to suit the techies. That surely means removing the "up" button in the web browsing profile, and forcing everyone who voted "no" to this idea to simply do what they find easy and obvious, that is customise their toolbars.
Comment 22 Carlos Woelz 2004-01-06 21:24:59 UTC
This reasoning opposing users and developers is not convincing, as you have to back observations like "the user wants this or that" with hard data, not with what you think a user want. The observation that different icons and new buttons can *completely* put off a new user is a bit of an exaggeration too. Otherwise you would never innovate.

So Roland is also stating what *he* thinks is better for the user. Stating otherwise won't make the discussion improve.

But Roland, you said that the up button always work.
I and Amador Aguiar found evidence on the contrary (comments 16 and 17). Also, I claim that putting together "navigation" buttons and "trail" buttons is confusing (see comment 9). Do you contest my findings?

Cheers,

Carlos Woelz
Comment 23 Carlos Woelz 2004-01-06 21:32:55 UTC
I said:

So Roland is also stating what *he* thinks is better for the user. Stating otherwise won't make the discussion improve.

I wanted to say:

Roland is stating what *he deduce* is better for the user. You have to fight his arguments, not state that he does not represent the average user.

Comment 24 Roland Seuhs 2004-01-07 11:38:06 UTC
> It seems that the general opinion is that most techies and KDE developers like the up button, but that most "average" computer users interested in simply browsing the web in the style they are accustomed to find it needlessly confusing. 

So that's the "general opinion"? Did we have a vote about that?

It's like with multiple tabs: If they don't know them, they don't miss them, but once they know it, most generally love it (at least that's my experience and I showed it to non-techies, too.) If we reduce Konqueror to IE-standards, it's just another browser that runs on an OS that can't run any Win32-games, that isn't preinstalled and unsupported by most software. In that case running it doesn't make the slightest sense and you could as well stay with Windows that was preinstalled anyway.

Let's face it: To gain users from Windows KDE has to be *BETTER* than Windows. That means we have to have *MORE* features and we have to show off that features.

In my opinion, on a purely technical level KDE is already a lot better than Windows, unfortunately that's mostly irrelevant to most users because most users either don't know KDE or want Win32-only software or both. Comparing user numbers that way is unfair.

What *IS* fair is comparing KDE-user numbers to other desktop environments and windowmanagers on Linux/Unix. And since according to all polls I saw, KDE runs more desktops than all other Linux solutions combined, KDE must be doing something right. If KDE is going the Gnome-route and wants to become yet another desktop environment without any obvious distinguishable features, it will most likely lose it's position on top and the userbase will decline to numbers similar to Gnome, Blackbox etc. Since KDE will still not be able to run Win32-games, don't expect any new Windows users.

If KDE wants to challenge Windows, it first has to get rid of the barriers, the most important being Win32-incompatibility. Integrating Wine into KDE would open KDE to the masses.
Another area KDE lacks is marketing. Essentially the same people who gave KDE a bad name because it's "too much like Windows" a couple of years ago are now giving it a bad name because it's "not enough like Windows".

If you don't know what some icon does, either you find out (by trying) or you just ignore it. No "confusion" whatsoever.
In general there must be a general delusion that only "techies" are intelligent beings and "non-techies" are complete morons bordering somewhere between humanity and animal life to say that they are "confused" by an icon they don't know in such a bad way that they can't use the application. Anybody who thinks that "average" users are so stupid please explain how those average users are able to navigate through millions of webpages with millions of different and inconsistent designs.

Also you didn't reply to any of my arguments, I repeat:

1) Default Apache generated indexes mimics a filemanager
2) The up icon is great for going to the homepage
3) Appearing/disappearing of icons is inconsistent and bad
4) More features is never wrong, having less features is something that should worry.

> I and Amador Aguiar found evidence on the contrary (comments 16 and 17). 

Let's say you proofed that randomly pushing buttons often doesn't make sense.

However, if you want to go to the homepage, the up-button always works (possibly with the pop-up list that appears when holding the up-button) with rare exceptions only on pages that use different subdomains. Of course this makes only sense if you want to go to the homepage just like the back-button only makes sense when you want to go back.

Essentially webpages offer a way to go back to the homepage which is very inconsistent, sometimes not very accessible and often hidden. If Konqueror offers a way that works on every page equally I see this as a strength not a weakness.

> Also, I claim that putting together "navigation" buttons and "trail" buttons is confusing (see comment 9). Do you contest my findings?

Well, in my opinion the logical way would be to use the "undo"/"redo" icon-pair for "back"/"forward" (or maybe have a new design for either the up-button or the back/forward pair), I agree with that.
Comment 25 Tom Chance 2004-01-07 12:03:53 UTC
> 1) Default Apache generated indexes mimics a filemanager 

And they have an "up directory" icon at the top. Will your average user care that much that they have to use this convenient link, and don't have an "Up" button?

> 2) The up icon is great for going to the homepage 

I'm not sure I nor Carlos understand you here. Can you give an example page? Whenever I try it, it only goes up one "directory", which therefore only works if the web site uses directories or mod_rewrite in a particular way to organise its content.


> 3) Appearing/disappearing of icons is inconsistent and bad 

Sometimes, yes, but then we also have to be context sensitive. People don't want to see "Burn ISO to CD" when they right click on a text file, for example. Likewise, Konqueror already removes several redundant buttons when switching between the default html and file view profiles. I'm suggesting that "Up" should be another such button.


> 4) More features is never wrong, having less features is something that should worry.

You're talking from the point of view of technology. Fine. But in design, you couldn't be more wrong; there's a saying along the lines of: "A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away". More is NOT inherently good in UI, and makes the desktop extremely confusing to new users because the user has to take in a large number of elements.

Now I disagree with GNOME's approach, which is to simplify everything as much as possible, but KDE's approach shouldn't be to keep things complicated because *we* like it that way. We should be loooking to present a VERY complex environment in an accessible way, whilst making the huge potential accessible even though it is often at first hidden.

To compete with other DEs, KDE shouldn't take the attitude that more is better, and that the more cool features it can sport, the better it will do, because I am quite certain in my opinion that KDE may as well say goodbye to huge chunks of the average user market unless distributors pick up KDE and do a lot of work with a very different attitude.

I am simply saying that in my experience, no average users I know find the idea of an up button intuitive in the context of the web, and they have all found it confusing to see it there, adding to the confusion they already feel when presented with different icons for back, forward etc. They are also unlikely to, without help, remove it. Those who do like the button are quite capable of adding it themselves, and if they're not likely to find it, then KDE has a big problem with presenting this customisability.
Comment 26 Roland Seuhs 2004-01-07 13:49:15 UTC
> And they have an "up directory" icon at the top. Will your average user care that much that they have to use this convenient link, and don't have an "Up" button?

The link is *on* the page which means it isn't visible when you have a larger directory and you have scrolled down.

> I'm not sure I nor Carlos understand you here. Can you give an example page? Whenever I try it, it only goes up one "directory", which therefore only works if the web site uses directories or mod_rewrite in a particular way to organise its content.

For example on this page "http://bugs.kde.org/show_bug.cgi?id=70506" the "up" menu has the entries:

- http://bugs.kde.org/show_bug.cgi
- http://bugs.kde.org/

Going to "http://bugs.kde.org/" via up-button is much faster than scrolling to the top and selecting "Bug Tracking Home"

It's actually a very good example. Just like on this page "Bug Tracking Home" isn't found instantly and isn't very accessible, it is similar on a lot of pages.

> Now I disagree with GNOME's approach, which is to simplify everything as much as possible, but KDE's approach shouldn't be to keep things complicated because *we* like it that way.

Well, first you make an assumtion here - that it is (too) complicated, which isn't proofen.
But aside from that: If you define "we" as "KDE-users", yes exactly we should keep it the way we like it.

Alienating the current userbase to optimize for some non-exiting "average user" that might or might not like the optimizations isn't a very good idea in my opinion.

You *assume* that the "average user" is confused by the up-button. You *assume* that the "average user" doesn't like the up-button once he gets to know it.

However, many KDE-users actively use the up-button, so many that some even found this wishlist-item and spoke against it.
Those are real people and that they use the up-button is a fact, not an assumtion.

So if I have the choice in pleasing real people that use the feature for a fact or pleasing some "average user" which is supposed to jump to KDE because somebody assumes he has very different preferences as the current userbase and assumes he likes the new way - well yes, I vote for the real people.

> To compete with other DEs, KDE shouldn't take the attitude that more is better, and that the more cool features it can sport, the better it will do, because I am quite certain in my opinion that KDE may as well say goodbye to huge chunks of the average user market unless distributors pick up KDE and do a lot of work with a very different attitude.

Fact remains that KDE is being used much more often than all other comparable (Only DEs on Unix/Linux are comparable) DEs combined.

I think that pretty well supports my belief that currently KDE is the best choice for the "average user". BECAUSE OF THAT I THINK THAT THE FUNDAMENTAL PHILOSOPHY AND POLICIES BEHIND KDE ARE CORRECT AND SHOULD NOT BE CHANGED. "Never change a winning team" if you will.

There are DEs "simpler" than KDE like Gnome and there are DEs more complicated than KDE like most classic windowmanagers.

Why did most users choose KDE?

BTW: 5) If the up-button disappears from time to time, the back/forward buttons will move and it is harder to build up motor-memory (or how it was called).

> I am simply saying that in my experience, no average users I know find the idea of an up button intuitive in the context of the web, and they have all found it confusing

Again, a lot of assumtions:

You assume that all those who use the up-button (for example those who posted on this bug) are no "average users".
You assume that those "average users" will never find new features useful.
You assume that those "average users" already have the perfect user interface and every change is only "confusing" and should be avoided. However at the same time you assume that those users will jump to KDE as soon as it is made similar to their current UI.
You assume that the needs of those users are more important than the needs of the current userbase, you would like KDE to risk their current userbase (which is the more than all other DEs combined!) to optimize for users which may or may not become KDE users.

None of those assumtions have been proved.

For all we know, the desktop needs of the "average" Linux user and the "average" Windows user could be exactly the same and the only reason why most people are running Windows is becaue their games/apps run on it.

If KPersonalizer cripples KDE when set to "Windows-mode", fine. But please don't change KDE's general look&feel just because of assumtions about non-KDE-users.
Comment 27 Tom Chance 2004-01-07 14:08:11 UTC
Roland, people posting their disapproval of the idea on bugs.kde.org is in no way an accurate representation of the KDE userbase; it reflects the thoughts of the technical elite, of developers and enthusiasts. When I talk about the average, I talk from experience helping friends and family, and working in an IT training center with hundreds of people from those who are taking their first steps with Windows to those who want to get an MSCE.

I suppose at the end of the day, the only way this discussion could be resolved is if a usability study were done with a random sample of users from a variety of backgrounds, since our own interpretations of the average and the userbase aren't going to be very reliable.
Comment 28 Carlos Woelz 2004-01-07 14:44:07 UTC
I agree with Roland in general with:

3) Appearing/disappearing of icons is inconsistent and bad 

and

5) If the up-button disappears from time to time, the back/forward buttons will move and it is harder to build up motor-memory (or how it was called). 

You are right. The user should not have a "dancing" toolbar, as in the windows explorer. The "changin icons" should be on the left.

However, the current technology implemented on KDE CVS is a unique "per .desktop file" XML ui. This open the possibility of having a more simple, usable aplication, with only icons, menus and hopefully configuration options related to the web browsing task available to the user. This way we could easily have a Firebird/Mozilla killer.

Obviously, all current funcionality would be kept by firing up konqueror in "normal" mode. And I don't think users would be confused by the two modes, as it  fells very natural since they are different tasks.

Summing up: if you want to browse the web, manage files, view images, all at the same time, open konq in normal mode.

If you want a konq version specialized in web browsing, open the web browser.

It is a problem to have different toolbars? I don't think so.
it is a problem to have a "dancing toolbar"? Yes, i agree with you.
Comment 29 Carlos Woelz 2004-01-07 15:05:22 UTC
Created attachment 4021 [details]
Mozilla navigarion bar

Mozilla does not have the up button, but it has a navigarion toolbar, that
reflects the rel up, right, left, docbook documents, etc... It can be
configured to be always on or to appear as needed.
This bar could provide the up funcionality when in web browsing mode and
browsing files using the file manager profile, and is IMHO, the ideal solution
to the dilemma.
Comment 30 Sashmit Bhaduri 2004-01-08 00:33:26 UTC
A expanded navigation toolbar is a good idea, but it __definatly__ (and the up button) should not appear as *default*.
Comment 31 Adrian Dziubek 2004-01-27 16:50:32 UTC
I think it's a wonderful button, which no other browser has. The only thing I would change about it, would be some more intelligent behaviour on php/cgi/... arguments, but it's one of the best things in konqueror.

Is there any discussion about toolbar default content anywhere here? I would change a bit in that matter. I find most of those displayed not useful, I changed it myself, why should one do it once again ;)
Comment 32 Gunter Ohrner 2004-02-04 09:53:47 UTC
Just wanted to say, I love this button in http-mode, too. :-)
Comment 33 Olivier LAHAYE 2004-02-04 10:58:08 UTC
1st of all, I'd say that if this button is removed, then I'll drop konqueror for another browser that has this feature.

But more generaly, it's a pity to see people discussing about usability when they clearly don't know any other product on the market. After ready this post, it's clear that many people think that konqueror is the only browser that has this feature...PLEASE LOOK outside.

- On windows, the "average user" is using a Wonderfull browser called "Avant broser" (http://www.avantbrowser.com/) that has many "confusing buttons". The up button is realy excellent on this browser BTW (Konqueror developpers SHOULD use it once as there are many excellent ideas in it).
- On windows, the "average user" can also use "MyIE2" another extremely cool browser (http://www.myie2.com/html_en/home.htm) that has also tons of buttons and execellent features.
- ON windows, if you're to dumb, then you can still use IE.

- On Linux, a very cool and promizing browser (Konqueror) has many cool features, but <troll>GNOME guys that have removed everything in their tools and are now trying to remove things in KDE ;-P</troll>
- On Linux, If you too "average" then use Mozilla.

To people thinking that buttons (back, forward and up) are not concistants and would be confusing for many people especialy windows user, they should log in once on a Windows XP system. Those 3 buttons are all over the place in explorer windows, and to my knowledge, nobody is confused.
On this point, I have to report that many of my users (I'm sysadmin) moved to KDE because of the file dialog box that has those navigation buttons (history navigation and physical navigation).

<troll on>
So If you want to remove an "UP" button, then remove it from a non KDE app.
<troll off>

NOTE: I'm Linux user using KDE since V0.4 Konqueror is my default browser since V2.(but I'm aware of windows softwares though).

Regards.
Comment 34 Andreas Daab 2004-03-15 21:32:31 UTC
We should not remove this icon in the default konqueror profile, but perhaps the ordering of the up, back and forward icon should be replace with back, forward and up.
Comment 35 james 2004-03-18 08:46:10 UTC
As a user (admittedly advanced) & having watched people (family) use this feature after I have pointed out what it does, I highly suggest that it be kept. 

1) If it isn't we make it inconsistant (bad)
2) They wouldn't use the feature if they had to configure it, but they use it now. 
3) It's darn useful, when sites move stuff, and don't provide a link, you can get the homepage easily, and find it if it still exists.
Comment 36 Ralph 2004-03-29 16:08:34 UTC
Please don't remove the up button.  I find it extremely useful, and miss it when using other browsers on someone else's computer.  And if the definition of a power user is knowing how to add it in after you remove it, I'm definitely not a power user.  

As for a "newbie", the first time my brother used Konqueror, he was using the up arrow (for navigating web sites), even before he discovered the tabbed browsing feature.

It turns into a tremendous chore to edit the url instead of using an up arrow to get the same result.  Just about any news site is a great example of the usefulness of an up arrow for navigation.  Doing it once or twice isn't a big deal.  Navigating dozens of news sites, and hundreds of urls, and it is a big deal. 

I suggest a new feature, to the bugs.kde.org site.  Negative voting.  So that we can each deduct 20 points from killing the up arrow removal.  

Adding it in myself wouldn't work for me, as I stated above.  If it had to be added in later, I would never have discovered it, just as I don't know most of the shortcut keys, simply because they aren't in front of me.  I know ctrl-s, and a few others that are shown in the menu, but I'm sure there are many more that I'll probably never use because they aren't in front of me. I sorely miss having the "shred" choice in the right click context menu, even though it is in the edit menu.  Can it be added to the context menu?  I'm sure it could.  Can I figure out how?  Not unless someone tells me, and not unless I know enough to ask.  The same is true with adding an application to the K (start) menu.  Is there a way?  I'm sure there is.  Do I know how? No. Will I ever use all the applications in my distro? Unlikely, even though I've explored most of the applications in the KDE start menu.  Because there are applications missing from the KDE start menu, I'm much more likely to not find out about them, until I learn much more of Linux, and until I find it absolutely necessary to locate an application with the functionality I need.
Comment 37 klee 2004-04-20 21:03:39 UTC
There were some comments about this on Bug #8333.  That BR is getting long, so I will reproduce the main points in my argument from that BR for convenience:

* Nearly every website on the Internet is at least 1-level hierarchical, in that http://hostname/ will be the home page, and http://hostname/some/path/ will be a subpage of that page. Even naive users understand this convention.
 
* Having the "up" button disappear/reappear will shift the spatial position of *all* the other buttons on the toolbar. A user who is accustomed to clicking "back" and "forward" based on muscle memory --- which even naive users do --- will curse the Konqueror designers every time they click "forward" when they meant to click "back". For this principal reason, I think the "up" button should be left as is.
Comment 38 esigra 2004-04-21 01:14:54 UTC
I would like to add my minus 20 votes to this wish. I also find the UP button very useful in all modes and I too miss it in other browsers I have to use.
Comment 39 Guillaume Pratte 2004-06-14 07:50:09 UTC
There really should be a way to vote against bugs, I would give -20 to this one right away. 
Comment 40 James Richard Tyrer 2004-07-11 01:21:58 UTC
Re: Comment #38, 39.

YES!  I vote NO.

I also miss the Up button when I have to use FireFox to view a broken website.

And removing it only for web browsing has serious usability issues.

I would close this as WONTFIX, but I would probably get in trouble.

--
JRT
Comment 41 Luke-Jr 2004-07-14 00:44:11 UTC
I use the Up button with webpages nearly every day. Another vote against its removal..
Comment 42 Jo Ă˜iongen 2004-07-16 12:55:02 UTC
I'm woting NO (to removing)! Using browser without the UP button is a PIA. Please close as WONTFIX.

Cheers Jo
Comment 43 David Walser 2004-08-13 22:14:57 UTC
I agree, please close this as INVALID, I vote against it.  As has been stated, it's useful with broken websites, and this is true whether you're a novice or advanced user.
Comment 44 Michael Jahn 2004-09-15 23:23:00 UTC
Ok, I am the one to mark this WONTFIX. Judging from the discussion I don't think that the up button will be removed and I don't want to have another bug hanging around which will probably be marked wontfix in 5 years. */me puts on his flame-proof asbesto suit and hopes for the best* :-)

To make konqueror easily store toolbar settings for different profiles vote here: bug 56730.
Comment 45 Carlos Woelz 2004-09-16 01:50:58 UTC
I am the reporter, and I don't even agree with my wish anymore, but it is still usefull to direct people to the discussion before bugging us at kde-usability. If it is closed, people will create another bug report or just bring the discussion again (and again, and again...).
So i vote for keeping it open.
Comment 46 Siegbert Baude 2004-11-13 01:43:12 UTC
I would like to invite all visitors of this bug to comment on my proposal in
http://bugs.kde.org/show_bug.cgi?id=93192