I’ve always found Global menu bars (ala Mac) to be bad usability-wise. They have numerous problems. I’ve always found that having a single menu bar at the top of the screen as Macs are both confusing and inefficient. Menu bars like that are leftovers from when Mac OS was primarily a single process operating system, and it should have been ditched long ago.1) They place unwarranted emphasis on menu bars. Menu bars are a UI element that should not be used frequently. In dealing with basically any application, most of my time is spent manipulating the content, followed by using the tool bars, then lastly menubars. In my word processor most of my time is spent typing, followed by quick formating and saving from the toolbar. In my web browser I frequently interact with the webpage, use forward and back buttons, but how often do I need to use stuff from the menu bars? Its only on the odd occasion that I need to open a website from my harddrive or get help.
Toolbars should hold the most common tasks for quick access, while menus are good because they can hold a lot of commands to all the less used operations. Menus are slow and error prone. It usually takes quite a while to find what you are looking for, and, as with any nested structure, slight movements can put you in the wrong place especially with things as narrow as menus. Menus are an important part of UIs because they allow you to do uncommon tasks, but why make uncommon tasks available at the easiest to reach spot on the screen?
2) A global menu bar is disconnected from the task at hand. It appears to be global, but functions local to what you are working on. If the focus is on the wrong window, for various reasons, it can lead to very confusing and potentially destructive behavior. This is just confusing, toolbars are attached to the application, isn’t it inconsistent to not do the same with menu bars? There is no quick way to tell without thinking which application the menu is functioning for. If menus were connected directly to the task, this problem would not exist. Global menu bars were fine when most people were running one application, but in todays multitasking world it is a poor choice. Global menu bars appear to be a part of the OS by appearing globally, but confusingly function local to the application.
3) Global menu bars don’t work with focus follows mouse.
4) Global menus don’t work well with multiple monitors. For a given application, you have two choices. First of all, you could put the menu always on the first monitor (like OSX). This means if you are using an application on the second monitor, you have to move your mouse to the first screen to use the menu. Confusing and inefficient. The second choice is to put a separate menu on each screen. This ends up being more efficient but just as confusing.
5) This really only applies to GNU/Linux and Unix desktops, but having a global menu bar would be inconsistent. With Macs, Apple has control over the GUI. On GNU/Linux, there are dozens of toolkits floating around, and if one desktop environment switched to having menus at the top of the screen, there is no way that every toolkit would do the same. You would end up with a desktop with some applications having menus at the top of the screen, and others with them on the application.
6) Widgets must be dynamically changed. You essentially have a moving target.
If you would like to see your thoughts or experiences with technology published, please consider writing an article for OSNews.
Global menus are perfectly fit for some applications like photoshop or gimp, which have a lot of narrow windows and many first level menu items. As You can see in the gimp for windows, there is no good window to put a menu in.
Of course these are mostly advanced applications and the best place to put menu would be context menu – but it is imposible for mac apps, which are designed for 1-button mice.
So, for mac it may be the best solution, but only for mac.
For all those talking about Fitt’s law – if You would be serious about it You would want to put up there something frequently used, like “global toolbar”.
And some thoughts about newbie agument – it is true, that newbies mostly work with one app at a time, maximized. But they never “throw a mouse to the top” – they slowly and carefully place mouse cursor _in_the_middle_ of the menu item and then click. So the whole point of Fitt’s law doesn’t in fact apply. So for them difference between global menu and app’s menu is none.
For gimp and the like best solutions for non-global-menu-desktops are in my opinion context menu and photoshop for windows style mdi. Mdi is also nice for separating its windows from the desktop and organizing them depending on semantics of the application and current job.
As someone already mentioned, all these Fitt’s law citers only take into account the time to acquire the target, not the total time to perform their action. Say I have a couple of small app windows near the bottom of the screen. If I need to select a menu option, do something in the window, select a menu option again, etc, I have to travel a lot more up and down with the mouse pointer than with a local menu bar. People who cite Fitt’s law ignore that most of the time I want the pointer back near the active window after selecting the menu item, which usually takes longer with a global menu bar, especially on larger screens.
At work, I currently work on a 2 screen Windows XP PC. To reduce the number of menubar, Windows uses the MDI windows system (application centric) : one app with multiple documents “inside”. This is hell for me. Firstly, I hate this gray background you can’t get rid off. Secondly, when you want to use this app on the two screens (or more), you have to extend it manually. The MDI system prevents you to have any document out of the mother window. If you want to see two documents simultaneously, your mother window must have enough surface to do this !!!! This case is not so uncommon, I often need to work on multiple images in Photoshop.
On mac this is different, MacOS is document centric. It means one document, one window. Very simple. Then I don’t need anything special to work on multiple documents when I have multiple screens. I just drag and drop the right document in its right place. Much better, I can see document from other apps. Because under Windows, you fill desease when your app is not fullscreen (most of the time because it prevents you from accessing the whole menu bar).
Under MacOS the menu bar will EVER be fully accessible. That is very important, and it’s always on the same place. You can drive your mouse there blindly, instead of spending your time to find where is the window you need.
Definitively, I do prefer the MacOS Menubar system. (I work everyday on both systems).
i switched recently to the mac, and in my opinion i disagree with most of what the author says. its SOOO much easier to have a global menu bar, than an individual one for each app. I wonder how many of these people who think otherwise, have really used a mac for more than a couple of hours.
and about the fitt law comments, i dont think its just for newbiews. So for me, yes, its really pratical being able to just slam the mouse to the top of the screen and be able to click and …voilá, menu appears.
and i do use apps with lots of windows(photoshop). the global menu has proved itself one of the best things i get from the mac.
The whole desktop is a part of finder. It’s just like a window except you can’t change the size or move it around.
Yes, I know. My point was that that is not obvious – it’s the freaking desktop! The menu bar at the top of the screen applies to the application in focus, *except* when there isn’t one, then it picks on some seemingly random task to get some menus for.
It seems like a nasty hacky solution to that situation where you don’t have anything to put into the menubar, rather than something put in by choice.
A few points: my ‘workstation’ setup is an nvidia-twinview created 3200×1200 space; I use 4 virtual desktops and I hardly ever keep anything maximized. I find systems that involve a “start” like menu compltetely inoperable; and for similar reasons I dislike global menu bars. Here’s why:
I shift between 3 different operating systems regularly and do not wish to internalize keyboard-shortcuts for any of them; any normal to power user would probably have a similar repitoir of keyboard shortcuts that I ever use (ctrl-c/x/z, alt-f4/f2), and I think the argument of “keyboard shortcuts!” is a poor one to make regarding the placement of menus.
Global menubar’s might be “infinite height” as opposed to my normal windows, but I do not spend my time throwing my mouse around the room to get to menus. If I am at the bottom right and the menu is in the top left, I have to traverse 3200 pixels across and 1200 high; its a special case, but your global menu “usability” issue really becomes moot when you put the menu miles away from my pointer.
Secondly, global menus take your (the users) focus away from the application window where it *usually* is.
Another bad thing is that if I want to access another programs application menu, I have to focus it (which means clicking on it usually, unless I want to alt-tab cycle through ~10 application windows); if my mouse is headed there, why not show the menu bar there?
On top of this, the menu gets drawn and then I have to shift eye focus to the top, re-read all of the options, and pick out which one I want. If the menu was on the window I am clicking on in the first place and happens to be visible (which it very often is), I’d already have had time to look at all of the 1st level menu options and even have started moving my mouse towards the desired one.
Finally, I like having my mouse where my program is so i can manipulate content within it; If this happens to be far away from the top of the screen (or worse, far away to the side of where the menu will be), this makes it a large distance away.
There are some practical reasons for these types of menus, and it seems to fit some peoples way of working, but my way of working is totally different. I keep many small windows around the desktop, most in plain view (not obscuring eachother) as I find its much faster for me to move my eyes than it is to move my mouse or move focus.
Still; people uphold the apple way as a sort of knee jerk reaction recently, especially with regards to usability. I’ve seen a lot of cool ideas come out of the usability camp, but I’m not yet convinced its an exact science; or that what is done with the usability principles ends up benefiting it.
A classic example is the usability crowd’s hatred of right click context menus. People are quick to bring up the corners of the screen as infinite spaces, but are always quick to forget about the fifth infinite space; the current position of the mouse.
Context-menu driven application menus, like XFCE/BlackBox/many *nix WM’s work quite well regardless of resolution as long as you can find some desktop space lying around. A thing that I’ve found working at 3200×1200 is that MANY things in the usability bible from apple simply fall apart at high resolution; assumptions about spatial locality start to fail or become gray, and as we go on and get bigger and badder monitors with better resolutions, I think it’d be smart to re-examine some of these usability decisions from 1984 rather than just hold up apple’s decisions as gospel.
Thank you! Someone who gets it!
All you Fitt’s law fanatics, yes we know about Fitt’s law too, yes it is faster to get at screen edges, no one is even arguing that. How about replying to some of the other points instead of harping on that?
Doesn’t work for FFM, slow for multiple monitors, slow for accessing menu bars of inactive apps, wastes space for “small” apps with only one or two menus, uncommonly used function in easy to access area.
Right now I have my desktop (KDE) set up to have the close button for the maximized window in the top left corner, minimize in the top right, and workspace switcher in the bottom right (mousewheel over it to switch desktops. All the most common things I do with windows are immediately accessible.
What really gets my goat is people claiming users can’t judge usability for themselves. I can and I do. I figure out what tasks I do the most, and I optimize my environment for them.
Look here for a labeled screenshot: http://www.ece.uvic.ca/~lspalteh/usability.png
I use a windows PC all day at work (.net dev) and a mac 80% the time at home with the other 20% being my xp pc so I think im pretty unbiased, and frankly I don’t care either way about the stupid menus. They both have advantages and disadvantages.
The only valid whinges about the global menu IMO are the *current* multi monitor situation.. yeah it sucks, tell apple who knows what they’ll come up with.
And people who cant live without focus follows mouse (freaks!)
and accessibility is what we are talking about here, isn’t it? Of course,w e are talking ’bout usability too, but that’s not the concern here.
I have my grunts with MacOS global menu bar, and I have my grunts with the per window menubar. (admittedly, I as an OS-Developer have already implemented per-windwo-menubars – simply imitating the best-known-behavior-for-me).
The former one is, in my opinion, a bit disturbing, as (for macOS, there is always the Apple-Menue, holding all the system wide settings – and there are all the application specific menues – which switch by changing the focus from one application to the other one. This can confuse the hell out of the average newbie – especially the ones coming from windows.
The per window-menues on the other hand bear the problem with them: there is less space to put stuff into the window canvas. You have to move your mouse around more carefully. They can equally confuse hell outta you – especially if the application programmer has not considered to include keyboard shortcuts (if you work with MAYA 3D, you will love them shortcuts – there are so many and they sometimes switch their contextual meaning).
I’ve found it troublesome for the first time, that – in MacOS, you don’t dispose an application by merely closing the window – to grasp that it is still in memory, waiting for what to do – open a window, loading data etc. But …
YOU GET USED TO IT.
(sorry for the shouting)
anyway: each one is entitled to his opinion and each one is entitled to publish this opinion. We are living in Democracy, and we are all grown ups who know how to talk together without humiliating each other.
Stay safe
PS: Still I like MacOS – the keeping the application loaded bears quite some speed up: you (from the os point of view) don’t need to create process & thread management for that app, don’t need to allocate memory from new on, and you don’t need to murder the entire process structure upon exit – unless the user tells explicitly to “EXIT” the application. Honestly, what’s quicker? allocating space for a window and a few controls or allocating Process structure, virtual memory, paging in memory, acquiring a few files (fd’s) opening a few threads, initialisation etc … (os developers know what I talk about)
Just to comment to a recent post, I bash the menu bar a good bit, and I have been using my mac for a good time, as i expect many others who criticize it. Not everyone caves to all things mac just because they own a mac.
Anyways, like others, people need to get off the Fitts law thing, it’s dated and has little basis in a modern system. But the biggy that drives me nuts when people try to point to it as the best reason for menu bars is when people talk about throwing their mouse to the top of the screen and so forth. Few people do this. I’ve seen some people claim they whip the mouse to the top and even let go of it. Now come on, how many actual humans do you really think do this?
The simple matter is people today keep a hand on their mouse at all times except when entering text where both hands are needed. Just go to any computer lab full of people and you will see this. People use their mice very well, and don’t go slamming them around all over the place. They have very good control, and move at a good clip, but nothing where over shooting something is a problem. Though targeting a corner becomes harder as people get bigger screens which move their focus to the center of the screen and corners/edges of the screen more out of their main field of view. Furthermore do you really think that people think about the top of the screen as something they can over shoot without worry? No they don’t people still try to stop on the mark just the same. So again Fitts law becomes less important since they would have got the mouse there just the same as if the menu bar cut through the center of the screen.
People who Praise Fitts Law are the same who defend 1 button mice to no end. They are the ones who don’t think people learn things over time, even if they aren’t good with it at first. If your trying to get a 5 year old on a computer, sure they might have issue with it, but in time they will be fine, and of course the first question is why are you trying to get a 5 year old to use a computer. Then on the other hand you have older people, well, believe it or not, but they will catch on without much issue to. But at the same time, their will only be first time computer using old people for so long. We are running out of such people. Today its pretty hard to find a person under 50 who hasn’t been using a computer for a long time, and all no new humans are going to grow up in a world without computers (this is not the place for a debate about 3rd world countries and so forth, not relevant to the point in the big picture) .
Any Usability study from pre very late 90s is invalid today. And Studies should be continually done to follow the evolution of people and computers. What is usable to people today will not be the same 5 years from now. This will be from peoples ways changing, and just things dating. Look at OS9, its a confusing land even to people who use OSX (if they only have used OS X). I can’t stand Windows Classic, not that Luna changed windows much, but it changes things enough that going back to it drives me nuts to no end, plus it’s ugly which is another huge aspect to usability. No interface will ever be found usable to the masses if they find it ugly to look at.
Firstly I’m a die hard Mac fan but this doesn’t not mean I think that apple are always correct. For example the single mouse button is just ridicules, particularly now as there is so much more need for a right click. I’m terrible at spelling as you my notice in this post and so I’m using the right mouse button to correct my spelling all the time.
Anyway back to the subject at hand. I do believe that a global menu bar is a good idea on Mac OS X. I cannot see it working with any window manager which has follow mouse focus and and windows for the multiple instance reason.
It does greatly depend on the application. Document based applications such as any Office application work well with a global menu bar which I imagine is exactly the reason that MS have one menu bar for this application. But there are also applications which I use that really do need non-global menus, Matlab for example. When using matlab I don’t want the menu bar of a graph to be in the global menu bar as it would make it even more cluttered. Now before you all say that it wasn’t necessarily designed specifically to be this, way I know but it still happens to be useful.
The argument saying that using a global menu for a small application like a media player or IM app is wasteful is ridicules. These applications spend most of there time, not in focus and so not displaying there menu bar anyway.
Well I hope that makes sense. The bottom line is that it works well in my opinion with OS X but couldn’t work with other OS’s.
I use Ion3 WM and most frequently all my apps in fullscreen mode. So to me, it is like I have a global menubar, no window title bar and no frame around window: it save a lot of screen space!
I have used both a Mac and a windows machine and can say that I prefer the whole GUI implementation on a Mac and in general Macs themselves.
The look and feel is far better and this also applies to the global menu bar.
the Mac implementation tends to focus people on making software that is functional and works as opposed to a snazzy skinned interface as is the thing with most window apps, especially those trying to hide the fact that they do not actually provide any benefit.
I also think you argued the ‘for’ point in your first statement when you stated that you rarely use menu bars, surely not havving them attached to every single window is a benefit to you!
Reading the above, it seems that preference of menu type is linked to the type of work done by the user.
If you rarely have more than a couple of windows open at a time, I think the Global Menu is generally considered to be the right way to go.
OTOH, if you have lots of windows open, the Application menu can be better. In this case, users don’t seem to have a problem with the Global Menu itself, but with the relative difficulty of (a) shifting the focus to a different application before using the menu, and then (b) moving the pointer back to the focused application having used the menu.
I see two possible solutions:
1) Allow the user to select whether or not menus are global or application-specific. I guess the default would be global, as (I suppose) multiple-windows users are more likely to be “power-users”, so are more likely to figure out how to change the setting.
2) Find a better way of achieving (a) and (b), above Ideas anyone?
Also, multi-monitor support is a must.
All in all, I think that there is still a long way to go when it comes to designing GUIs. Apart from the much-talked-about 3D initiatives, I think it would be great to see (optional) system wide support for a number of features that can only be found in individual applications. Here are some suggestions – most of which have implications for menu management. BTW – appologies to users of OSs that already support some of these:
1) Windows that can stick to each other and the screen edges.
2) Document windows that can be made to “stack”, with tabs for moving through the stack.
3) MDI windows where the children can be dragged out of the parent container, so making them SDI windows.
4) Tear-off menus and tool-bars that can also be made to “stick” to other windows and screen edges.
5) In-built tools for switching between sets of preferences. These should include some useful pre-defined sets, designed for working with (a) one big document, (b) Side-by-side comparison, (c) complex tasks than needs lots of open apps.
That’s the windows way. Have one window with the toolbar on top and all open documents open inside that window. Ugh!
Maybe to you
Some of us prefer that way of doing things– No-one can be truly right on this subject. Certain people will always prefer a certain.
I drive 23 miles to work down a load of manky A roads and through rough areas. Sometimes I like to take the 30 mile route down country roads and through nice areas
Efficiency isn’t always best.
A logical extension of this article is to check out the “Anti-Mac” article from the ACM site:
http://www.acm.org/cacm/AUG96/antimac.htm
The Mac interface, prior to OSX, was locked in a time warp from 15 years earlier. Personally, I have never used a worse interface — but heh I’ve only got 22 years experience in computing.
Ok, Mac bigots, your turn to respond.
I did not even consider MacOS before OS X.
it just burns me when people use old Mac arguments against OS X (the global menu and single mouse button arguments are personal preferences and really there is no inherent problem with them unless you are bound to the other methods via a skin graft or something.)
The thing about the single menubar on the mac that is SO fantastic is that there is NEVER any doubt about what application is in focus. You just look at the upper left hand corner of the screen and you’ll know. You NEVER have to evaluate the contents of the application windows to determine what application is on top.
It also allows you to be incredibly, wonderfully sloppy about mouse placement. You can just slam your mouse to the top of the screen without even looking.
And having the finder come into focus when there are no applications open makes total sense. It’s the most commonly used application.
I use Windows XP 5 days a week, 8 hours a day at work. I use my Mac at home for about an hour a day, perhaps more on weekends. I learned the Mac many years AFTER learning Windows, and I just find the Mac way to flow more naturally.
The new computing is mobile: laptops are quickly becoming desktop-replacements.
And on laptops, with only a trackpad or an IBM clit-mouse around, Fitt’s law is still applicable.
I agree that no one “throws” a desktop mouse, people have just become too used to clicking where they mean to.
But on a laptop? Trackpads and IBM’s nub-mice are just too imprecise for that, notice how no one can play an FPS with a trackpad?
On a laptop, I still “throw” the trackpad, and when I use “thinkpads”, I just can’t get used to the nub-mouse enough to be accurate, so I throw that too.
It still applies, just not everywhere.
Please stop the name calling! Its not needed.
I had so many great thoughts while i was reading through these posts, but there are so many(and im really tired) that i forgot most of it.
With that said, i think Fitts law is right for some instances, but not as much as some people claim. Its extremely true for screen corners. Thats why i have expose mapped to them, and will have spotlight on the 4th when Tiger comes.
Concerning the menu argument: maybe global menus is bad, but maybe this bad thing led to a good thing-pushing people towards keyboard shortcuts. I look at my windows xp machine at work, and none of the shortcuts are labeled(except for copy/paste/save/print) On the mac, almost everything in the menus has a labeled shortcut (at least for most of the apps). And…. I DONT USE the menubar! Honestly i haven’t touched the menubar in forever, except for in photoshop. And the MDI in photoshop for windows is horrible, so i won’t even talk about that. USING THE COMMAND KEY RATHER THAN CNRL/ALT (depending on some unknown rule that i can’t figure out) IS FAR SUPERIOR. sorry for caps.
Oh, the comment about switching to apps that don’t have a window open: If you switch to them in the dock, then it automagically does a command-n for you, so you don’t even have to go to the menubar.
Why is everyone acting like there is no contextual menu on the mac? There is, period.
I agree about multiple monitor problem, only if you don’t use keyboard shortcuts.
In what instance would you want to go to a different app to go immediately to the menu bar? I have never had to do that. Its always either dragging something into an app, or interacting with the main window. I don’t see why people harp on this advantage when there doesn’t seem like a reason to do what you all are describing.
What i find funny is that when i use my mac, i never NEVER have a window maximized all the way (which is where main-menu’s supposedly have the advantage, according to posters). On the other hand, I and all my office pogues, work with windows maximized while at work on the PCs. And i constantly close my stuff on the PC to keep it stable and fast. Weird, huh?
goodnight from korea!
I think that the suggestion that global menu bars derive from mac’s days as a single-process machine is ill-founded because on the Amiga, they too had a global menu bar at the top of the screen and started out as a pre-emptively multitasking operating system with multiple processes and tasks, so the theory doesn’t follow. I always found the global bar to be more useful than contextual menus that you have to hunt around on the screen to activate. The only problem with the Mac menu bar is that the menus handle pretty slowly. They are better with more responsiveness.
On the whole though, I think menu’s of any kind are hard to navigate – you have to develop a skill for keeping your mouse pointer within a thin rectangular area so that you can get to a sub-menu and it often doesn’t work out.
1) They place unwarranted emphasis on menu bars. Menu bars are a UI element that should not be used frequently. In dealing with basically any application, most of my time is spent manipulating the content, followed by using the tool bars, then lastly menubars. In my word processor most of my time is spent typing, followed by quick formating and saving from the toolbar. In my web browser I frequently interact with the webpage, use forward and back buttons, but how often do I need to use stuff from the menu bars? Its only on the odd occasion that I need to open a website from my harddrive or get help.
A menu item takes up no space when not activated, (on a mac anyways) is a quasi-mode which allows for gestures to be developed, and is the perfect place to stick 90% of the features of an app (the stuff that isnt used enough to justify the screen real-estate of a toolbar item.)
Toolbars should hold the most common tasks for quick access, while menus are good because they can hold a lot of commands to all the less used operations. Menus are slow and error prone. It usually takes quite a while to find what you are looking for, and, as with any nested structure, slight movements can put you in the wrong place especially with things as narrow as menus. Menus are an important part of UIs because they allow you to do uncommon tasks, but why make uncommon tasks available at the easiest to reach spot on the screen?
Toolbars are fantastic for quick access, but the more buttons on the bar, the less effective it is (both fittes and hicks laws apply to this). Menus should never be nested more then one level. The algorithm defining the buffer for a nested element is a straight line on every os other then the mac (where it is v shaped), so while a nested element is more error prone on other operating systems, it doesnt have to be. Since menus take up no space, menu entries should be more verbose then they are on non-mac systems, which increases discoverability (it is easy to tell what infrequently used features do simply from the title.)
2) A global menu bar is disconnected from the task at hand. It appears to be global, but functions local to what you are working on. If the focus is on the wrong window, for various reasons, it can lead to very confusing and potentially destructive behavior. This is just confusing, toolbars are attached to the application, isn’t it inconsistent to not do the same with menu bars? There is no quick way to tell without thinking which application the menu is functioning for. If menus were connected directly to the task, this problem would not exist. Global menu bars were fine when most people were running one application, but in todays multitasking world it is a poor choice. Global menu bars appear to be a part of the OS by appearing globally, but confusingly function local to the application.
Actually, in todays multitasking world it makes alot more sense. Each and every problem you mention applies to contextual menus as well (i.e. right click). the only difference is that calling up a contextual menu is explicit, the menubar is implicit (kinda like the dashboard app that nat is working on.). The entire mac environment is built around application transparancy, you work with documents and the “application” environment is built into the os. I agree that without that, a global menu bar is a dumb idea, but i also think that having the os as a bucket for different application environments is a dumb idea, as it highlights the inhearent modality of applications rather then hide it.
3) Global menu bars don’t work with focus follows mouse.
why not?
4) Global menus don’t work well with multiple monitors. For a given application, you have two choices. First of all, you could put the menu always on the first monitor (like OSX). This means if you are using an application on the second monitor, you have to move your mouse to the first screen to use the menu. Confusing and inefficient. The second choice is to put a separate menu on each screen. This ends up being more efficient but just as confusing.
I dont know about you, but my second monitor alwas holds things that I just glance at now and then (such as bug reports, or specs). I know for artists (the main users of a dual head setup), the second monitor will hold the palletes, the main monitor holds the canvas. I am unaware of anyone who actually uses apps on both monitors simultaniously, but that may just be my own ignorance. as far as i know though, this is a non issue for real-world use.
5) This really only applies to GNU/Linux and Unix desktops, but having a global menu bar would be inconsistent. With Macs, Apple has control over the GUI. On GNU/Linux, there are dozens of toolkits floating around, and if one desktop environment switched to having menus at the top of the screen, there is no way that every toolkit would do the same. You would end up with a desktop with some applications having menus at the top of the screen, and others with them on the application.
Agreed 110%. A great example would be spatial nautilus. Fantastic if you have control over the directory structure (like in home), kinda falls apart when you start from /, and want to end in something 8 directories deep. Then there are things like maximize, which is totally and completely useless in a spatial environment. Its far more important to be consistant, and providing two different paradigms for two different sets of tasks is far worse then using the same one, even if it is less efficient. A global menubar in gnome would be disasterous as soon as you use a non-gnome app.
6) Widgets must be dynamically changed. You essentially have a moving target.
Again, a non-issue for the apple world, as developers follow the HIG like scripture. Menus will be similar, and the differences that do appear are the ones you would expect by shifting context. The big problem is that first application menu shifting sizes, but that can be fixed with fruitmenu.
————————————————
There are a great many reasons that the apple menu bar is a good idea. The reason it isnt used in unix is that microsoft didnt copy it when they did win95. the reason microsoft didnt copy it had nothing to do with if it was a good idea or not, and alot to do with the fact that apple has it patented. Testing has shown that apple users will be 1.5-2x faster with menus then windows users, as fittes law helps ALOT with gesture chunking.
This article and its comments are nothing more than a bunch of “GNU/Linux” users who actually think a floating menubar, despite its slowdown of user speed, is a good idea.
There is absolutely nothing wrong with a global menu bar other than you’re not used to it after years of using Windows and the Linux desktops that have cloned it.
Get over it.
My favorite part of all this, of course, is that if ANYONE defends the global menu bar, you guys say “Well, Mac fanatics always attack me when I give my opinion…”
Please. You offered criticism. People are pointing out what’s wrong with that criticism. Nobody’s a fanatic and nobody’s attacking you.
FACTS:
1.) Usability tests showed users to be faster with a global menu bar.
2.) Fitt’s Law
3.) Floating menu bars are confusing and space-wasting
Deal with it!
Global menubar’s might be “infinite height” as opposed to my normal windows, but I do not spend my time throwing my mouse around the room to get to menus. If I am at the bottom right and the menu is in the top left, I have to traverse 3200 pixels across and 1200 high; its a special case, but your global menu “usability” issue really becomes moot when you put the menu miles away from my pointer.
size is totally irrelivent because of fittes law. heres something i use alot to explain it. stick a launcher on any of the corners of gnome (i generally have a gnome-terminal launcher in the upper right hand corner). now, every time you want to open a terminal, dont take your eyes off what you are doing, and launch it. you will find you can hit it every time. by the end of the week, you wont even think about it anymore. that is because something that is infinate in size in two dimensions is virtually impossible to miss.
Secondly, global menus take your (the users) focus away from the application window where it *usually* is.
if the user shifts his attention to the menubar, or toolbar, or taskbar, or scrollbar, or anything other then the content, then his attention is off the task. it doesnt matter if it is in an application window or not, at least on a mac where the line between os and app is quite blurry.
Another bad thing is that if I want to access another programs application menu, I have to focus it (which means clicking on it usually, unless I want to alt-tab cycle through ~10 application windows); if my mouse is headed there, why not show the menu bar there?
unless you focus by clicking on the actual menubar, then the distance to the target is completely irrelivent, if your mouse is 20 pixels from the menubar, and 100 pixels from the top of the screen, the top of the screen will still be easier to hit. remember, this stuff isnt just people talking out of their ass, there has been a ton of studies on this.
On top of this, the menu gets drawn and then I have to shift eye focus to the top, re-read all of the options, and pick out which one I want. If the menu was on the window I am clicking on in the first place and happens to be visible (which it very often is), I’d already have had time to look at all of the 1st level menu options and even have started moving my mouse towards the desired one.
What your eyes are on, and what your attention is on are two different things. Raskin describes this as a “locus” of attention. If the user has profincy with the app, then he should be able to activate a commonly used menu item while still thinking about his task. That is the whole point of a gui, to allow the user to automate behaviors.
Finally, I like having my mouse where my program is so i can manipulate content within it; If this happens to be far away from the top of the screen (or worse, far away to the side of where the menu will be), this makes it a large distance away.
distance becomes irrelivent when you are talking about screen edges, and the screen area of your content should be way bigger then anything else, which makes it extremely easy to reaquire.
There are some practical reasons for these types of menus, and it seems to fit some peoples way of working, but my way of working is totally different. I keep many small windows around the desktop, most in plain view (not obscuring eachother) as I find its much faster for me to move my eyes than it is to move my mouse or move focus.
actually, that is the best arguement for a mac style window system. the best way to use windows style is to have every application maximized, and use the taskbar to switch between them. that way the menu location stays consistant.
Still; people uphold the apple way as a sort of knee jerk reaction recently, especially with regards to usability. I’ve seen a lot of cool ideas come out of the usability camp, but I’m not yet convinced its an exact science; or that what is done with the usability principles ends up benefiting it.
Usability principals should never be applied blindly, and it isnt some magic pixie dust you can just sprinkle on a project to make it usable. Good usability is a quantifiable science based on principals from cognitive psychology. You build the interface around the way we know the human brain operates when it is working on something. Design choices are made with that in mind, and are tested by scientific studies. Bad usability is done by a comitee of people who have (maybe) read a jakob nielson or don norman book, and think that good design can be done by anyone with a bit of common sense. theres a reason the ipod is a billion times more enjoyable then a zen, and none of that is because of tech specs.
A classic example is the usability crowd’s hatred of right click context menus. People are quick to bring up the corners of the screen as infinite spaces, but are always quick to forget about the fifth infinite space; the current position of the mouse.
that isnt the usability crowd, that is the mac zealot crowd. the easiest cursor to hit is the one your mouse is currently over, its virtual size is infinate (in all directions), and is literally impossible to miss. anyone who disagrees with that really doesnt know what they are talking about.
Context-menu driven application menus, like XFCE/BlackBox/many *nix WM’s work quite well regardless of resolution as long as you can find some desktop space lying around. A thing that I’ve found working at 3200×1200 is that MANY things in the usability bible from apple simply fall apart at high resolution; assumptions about spatial locality start to fail or become gray, and as we go on and get bigger and badder monitors with better resolutions, I think it’d be smart to re-examine some of these usability decisions from 1984 rather than just hold up apple’s decisions as gospel.
I completely and totally agree. On the flipside, a great deal of fundamental usability research was done by apple in 1984, and most usability paradigms stem from microsoft (god awful) ripoff with win95. (and that isnt a troll, they have gotten *much* better over the years, but win95 was a copy without knowing what they were copying). The book shouldnt be held up as gospal, but at the same time shouldnt be thrown out the window. Stuff like reexamining win95 conventions in a non spatial environment would be a very good place to start. for example, why are there desktop objects (like the trash) in a non-spatial environment, where they will be obscured any time real work is done and an application is maximized? When was the last time you actually dragged files to the trash in a non-mac environment? why is the only place i have ever seen this addressed is with ubuntus trash applet?
Your rant about Fitt’s Law makes no sense. Fitt’s law has nothing to do with UI’s or learning. It has to do with hand-eye coordination. As humans have not evolved better hand-eye coordination in the last 20 years, any research done in the 80’s regarding that subject is valid today.
this gets bandied about alot, and very few people know what they are talking about. fittes law is not “A menu at the top of the screen is more efficient then one anywhere else”. Fitts law is
T = a + blog2(D/W+1)
where
* T is the average time taken to complete the movement. (Traditionally, researchers have used the symbol MT for this, to mean movement time.)
* a and b are empirical constants, and can be determined by fitting a straight line to measured data.
* D is the distance from the starting point to the center of the target. (Traditionally, researchers have used the symbol A for this, to mean the amplitude of the movement.)
* W is the width of the target measured along the axis of motion. W can also be thought of as the allowed error tolerance in the final position, since the final point of the motion must fall within +/- W/2 of the target’s centre.
(http://en.wikipedia.org/wiki/Fitts‘_law)
the best laymans description ive seen is over at asktog, where he says:
Fitts’s Law: The time to acquire a target is a function of the distance to and size of the target.
(http://www.asktog.com/columns/022DesignedToGiveFitts.html)
keep in mind, the tog is a genius, but he also has a massive chip on his shoulder over windows copying many of his ideas (and poorly), and loves to rag on it and microsoft in general. but that page is a great description and explination of what it is, and why its vital to take into account when designing an interface
<- KDE user
Global menu bars are easier to use (though of course that only matters if one uses menus in the first place). Being at the top of the screen makes it a much bigger target.
I would use them in KDE (which someone pointed out is possible), but unfortunately it only works with KDE apps (not even QT apps). QT apps, GTK apps etc don’t use the global menu bar, and the inconsistency is enough to get me not to use it. Until I can be sure I only have KDE apps, I have to pass on the global bar
Naturally keyboard shortcuts are the way to go, but it gets confusing when one switches around the language of the GUI
I have to get used to alt-d, p to save files instead of alt-f, s, and until I do, I’ll be using the menus (and I wish they could be global)
Since the equation isn’t linear, you can’t do the addition trick to count both ways. You have to do each way seperately, and add up the two. For the first trip:
T = A + B log_2(1200/100 + 1) = A + 3.7 x B
For the first return:
T = A + B log_2(1200/100 + 1) = A + 3.7 x B
Total T = 2A + 7.4 x B
For the second trip:
T = A + B log_2(335/20 + 1) = A + 4.15 x B
T = A + B log_2(335/100.0 + 1) = A + 2.12 x B
Total T = 2A + 6.27 x B
Note that the global menu wins the initial motion, and only loses the return motion because we haven’t accounted for mouse acceleration. I’d also point out that your scenario is rather convenient for the local menu. Most window systems cluster windows around the top-left corner, not the right-hand side. Think about your word processor, your web-browser, your text editor. How often do you have them on the right side of the screen? An AIM-type app is rather unique in that it’s usually placed off to the side. I’m typing this on an Ubuntu machine, and just starting up a view random apps shows that metacity clusters them in the top-left corner.
One menu bar on the current window and the other at the top of the screen would confuse any user- particularly Mac users!
One of the Java based programs I used had menus in the Window and I found that really funny and difficult to use on my Mac. It was readily fixed after I wrote to the author of the software.
As you can infer, I prefer global menus, except in the case of multiple monitors (need a choice in this situation as someone pointed).
I guess it is hard to adapt to new concepts when one gets used to doing things in a particular way.
My philosophy is ‘less is more’ or KISS (keep it simple stupid) and that is one of the reasons I like the Mac way-although Mac Os X has evolved to resemble Windows in many ways (to attract potential converts- I presume).
I have to admit that this has been one of the more civil debates (or nicely moderated in any case) I have read on Os News. I like many of the pro-Mac responses-didn’t come across as fanatical.
Cheers!
Really the key thing as to whether the Global menu bar works is down to the size of your screen. Run on a 12″ powerbook, you’re going to end up running most things full screen, so whether you’ve got a Global menu bar or not doesn’t really make much difference. Run it on a larger 19″+ screen and the Global menu bar starts to become a mouse moving pain. Overall I’d vote for ditching it (and while they’re at they should also sort out the stupid inability to resize your window by grabbing anything other than a tiny weeny bottom right hand bit of the window in questions frame).
Yes, users are not the best judges of what is efficient. It might be hard for you to accept, but the fact that customer is not always right is quite well accepted in many fields. Doctors, for example, don’t rely on patient self-diagnosis! The AskTog article has a quote that is spot-on: “(Don’t bother to ask them whether they’ve slowed down. They’ll tell you it sped them up. Only the stopwatch knows for sure.)”
Non global menu bars are ugly-just my opinion.
Just checked my mac here at work (10.3.8).. Guess what? The top left corner doesnt do ANYTHING!
Wow. So much for fitt’s law. They completely threw that out the window. Instead of making the apple menu accessible in the top left, they reserved that space for a rounded corner of the menu bar that does NOTHING.
Yes, apple had some good ideas wrt usability. But you’re kidding yourselves if you think they still religiously follow them today.. THey’ve even given up some of the basics (consistency) in the name of flashy looks.
OS X is nice, but it’s not automatically superior in usability to anything in particular.
Please, criticise my layout in the linked screenshot. I’d love to hear how the extensive control of the interface that I get in KDE has resulted in a sub-par environment.
Actually, contrast that screenshot to OS X. In OS X: top left corner, no action, top right corner, no action, bottom left and right, useless. The only place that OS X has any regard for fitt’s law is the top edge and the dock. My screenshot shows considerably more efficient use of these valuable spaces.
Yes, not every user knows what’s best for them, but that doesn’t mean people can’t optimize for themselves.
1) The top-left corner is one of the most easily accessible places on the system. Putting the close button there is dangerous, because it makes it easy to close the app, and fairly useless, because it’s prime real-estate that could be used for something other than closing the window. Do you really close windows more often then you do anything inside them? The same thing with minimize and maximize.
2) Label-less icons are very hard to hit precisely. Toolbars should be big and have clearly labeled icons.
3) The menu-text is rather tight in the reading axis. Wider fonts (like Verdana), are easier to read.
4) The pager doesn’t show window contents.
First of all, apple didnt come up with fitts law, they were just one of the first computer ui designers to design with it in mind. Secondly, just because apple doesnt use all four corners doesnt make them any less valuable. As you have pointed out, apples design isnt what people should be following.
and osx is probably the worst os apple has ever come out with in regards to usability, as anyone in the field will tell you. the arguement is that the exponential gain in functionality means that the additional complexity is unavoidable, which is of course the exact same thing that jef raskin was told when he first pitched the idea of a computer made for humans. technology wise, osx is a quantum leap forward from os9. usability wise, it is a huge step back.
1) I use close button more often then any single button in my system and it should be easly accesible. I open apps by clicking on documents or opening them from fast start dock, filer or system menu. I put trash applet in bottom-right corner, program menu in bottom-left. I use left border of the screen (except bottom-left corner) for easy access to the desktop (d&d and menu) and for changing virtual desks (with mouse wheel) and right border for scrolbar.
And with maximized windows global or not global doesn’t make any difference (it is less then 10px). I could live with it, except it does not work with ffm, and ffm is just to useful for some tasks to get rid of it.
2) It clearly depends on user, and in most systems text labels are on by default, You turn them off when You decide they take too much space.
3) same as 2.
4) valid point. Basic KDE pager is broken, it does not show contents and does not allow to move windows between virtual desks, gnome’s is better.
Still, mac don’t use most of this important and accesible places on screen, so they are not that important anyway, right? They are only brought in front when mac-users defend global menu.
How can you possibly use the close button m ore often than any other button? That’d mean you spend more time opening and closing apps than doing anything in them.
(2) doesn’t depend on the user. Users *think* they take up too much space, but having text labels makes the targets bigger, which makes it more efficient to click on them.
(3) Same thing with 3. Users think small text is “more efficient”, but studies show that open, wide fonts are much easier and faster to read.
As for your last point, the argument for a global menu has nothing to do with the Mac, except to the extent that the Mac uses it. The fact that the Mac makes other design mistakes is not really relevent here.
1) The top-left corner is one of the most easily accessible places on the system. Putting the close button there is dangerous, because it makes it easy to close the app, and fairly useless, because it’s prime real-estate that could be used for something other than closing the window. Do you really close windows more often then you do anything inside them? The same thing with minimize and maximize.
The whole point of putting it there is to make it easy to close the app. I’m not spastic, so I don’t close it by accident. And yes, I close windows very often.. Most of the apps I use are bound to keyboard shortcuts, so I’ve become a chronic opener and closer of applications because it’s so easy.
2) Label-less icons are very hard to hit precisely. Toolbars should be big and have clearly labeled icons.
They should also get out of the way of my work and take up little screen space. Most toolbar icons have keyboard shortcuts, so I barely use them anyway.
3) The menu-text is rather tight in the reading axis. Wider fonts (like Verdana), are easier to read.
I suppose I could change that.. I’ve never had any problems reading standard times new roman 12 or whatever that is..
4) The pager doesn’t show window contents.
I find that showing window contents doesn’t actually help me. The pager is too small to make a tiny image of the contents useful, and I’m not making it bigger, because that would be a waste of screen space. A tiny screenshot, 20 pixels high, would just deceive me into thinking I could garner useful info by looking at. The basic shapes of the windows let me know that there are windows there, and mostly I remember what they are.
The only thing I miss from OS 9 is the ability to make buttons. I had a folder full of Aliases (shortcuts) in the form of buttons. A kind of launcher for my frequently used apps if you will. I still have a folder for this in OS X but alas, I have to click twice now. …… …… twice…… …… I’m sorry…. … I … I have to go…….<sobs and runs away>.
How can you possibly use the close button m ore often than any other button? That’d mean you spend more time opening and closing apps than doing anything in them.
Not more often than any other button, but given the widgets that I can put up there, its high on the list. My choices are a menu bar, a close/maximize/minimize button, an application menu (the thing that comes up when you right click a window title bar), and the taskbar. I almost never use the menu bar, and everything else is already spread out in easy to access places, so I chose the close button in one corner, and minimize in the other.
I suppose I could change that.. I’ve never had any problems reading standard times new roman 12 or whatever that is…
Times New Roman 12pt is widely acknowledged to be a horrible font. You might not notice it slowing you down, but that’s kind of my entire point. Until you measure it (and people spent a lot of time measuring this sort of thing), you can’t know.
I haven’t used mac’s much, but I’ve tinkered with this emac we have at work that sits in the corner collecting dust. For me it feels like more work bringing focus to a window and then trying to select it’s menu items at the top of the screen. I guess they provide key bindings that could speed this up a bit, but moving the mouse cursor all over a 1280×1024 desktop feels like a big waste when I use the system. Just my opinion on the subject.
The window usually is quite a large target to hit, so it should be quite simple to do that part. To move the mouse over the screen to hit the top menu is very little work as you can have accellerated mouse speed, and you would be at top very quick. The real work you don’t have to do is to exactly position the mouse over the menu. The menubar attached to a window is very small area, and its easy to overshoot. The top window presents an infinit target, you can’t miss it. You could google for Fitts law for more info on this. Or have a look at:
http://www.asktog.com/columns/022DesignedToGiveFitts.html
Until you measure it (and people spent a lot of time measuring this sort of thing), you can’t know.
Which gets me into one of the aspects of these discussions that feel making me bang my head against the wall. The fact that people spent time measuring these things for one population and application domain in the past, does not necessarily mean that will be the best design for a different population and application domain in the future.
There are some major assumptions behind putting pull-down menus at the top of the screen that I have not seen addressed:
1: it assumes that the pull-down menu will the the primary mode through which the user will interact with the application, throughout the user’s relationship with the software.
2: pull-down menus (in general) force the user to learn a complex taxonomy of functions.
3: there are cultural assumptions in regards to written language and directionality of language that are rarely acknowledged with the pull-down menu paradigm.
4: lots of assumptions about the physical abilities of the user.
Times New Roman is a bad font? Sources?
A quick google search turns up this:
http://psychology.wichita.edu/surl/usabilitynews/52/UK_font.htm
With a quote:
“The experimenters found that the participants perceived a difference in legibility between the fonts of which Times New Roman, Verdana and Georgia were most legible.”
And another here:
http://psychology.wichita.edu/surl/usabilitynews/41/onlinetext.htm
Quote from conclusion:
“Several observations can be made regarding the examined font types. First, no significant differences in reading efficiency were detected between the font types at any size. There were, however, significant differences in reading time. Generally, Times and Arial were read faster than Courier, Schoolbook, and Georgia. Fonts at the 12-point size were read faster than fonts at the 10-point size.”
Verdana was the most preferred font, but Times and others were faster to read. And you yourself say that user’s preference should take second place to empirical measurements of efficiency.
How can you possibly use the close button m ore often than any other button? That’d mean you spend more time opening and closing apps than doing anything in them.
That’s easy. I open documents and apps in many different ways. I use them in many different ways. But I close them all the same. With FireFox – tabs, back button and maybe “disable styles” buttons I use more then “close”, but in eclipse I don’t have those, same with “pajączek” or the bat. Close button is universal. It may not be the same for mac, if You have to close apps through menu or key-shortcut (if I remember corectly, I was using macOS around wersion 8 or 7). So, as I said earlier – global menu may be best solution for macOS and still not that good for other systems.
(2) doesn’t depend on the user. Users *think* they take up too much space, but having text labels makes the targets bigger, which makes it more efficient to click on them.
It is like with children – for 3 years old You have this big duplo bilding blocks. But children grow and after some practice it is convenient to use smaller ones. same with this – I have years of practice whith mouse and I can hit 4×4 px square with deadly accuracy. In most cases I use 16×16 icons and it is enaugh _for_me_. The only exception is back button in browser, which should be twice as big – I mostly hit it without looking and sometimes miss, which is annoying – any skin for FF with this one button bigger (longer)?
(3) Same thing with 3. Users think small text is “more efficient”, but studies show that open, wide fonts are much easier and faster to read.
I love big text. I really do. But I only read them first few hundred times. Then big menu text gets in my way when I read other texts (block of program, for example). Thats why I prefer them as small as possible – best solution would be _single_ “menu” button on the toolbar with the rest inside.
It is just logical, if You think about it – big, werbal menus, big, werbal toolbars with tooltips are important for newbies, who know nothing about programs and have problems using a mouse. After some time people prefer less werbal toolbars, key-shortcuts and some strange behaviour like ffm (depends on application they use).
As for your last point, the argument for a global menu has nothing to do with the Mac, except to the extent that the Mac uses it. The fact that the Mac makes other design mistakes is not really relevent here.
ok.
“Fitt’s law has nothing to do with UI’s or learning. It has to do with hand-eye coordination.”
Do you understand how contradictory these sentences are (except GUI of course)? Coordination IS based on training.
It is funny, when someone based on knowledge from wikipedia is correcting everybody. You would be the only one in the world whose motoneuronal coordination can’t be improved by training. My condolences.
Pie menus with keyboard acceleration is the best, IMO.
They are good if you have just a few menu items, but they are not so good if you have many, or need to have submenus. They are also a bit hard to use at the corners of the screen.
First off, an observation from Tiger: Apple has taken Fitt’s law into account and clicking in the upper-left corner now triggers the Apple menu. The Apple menu isn’t drawn all the way in the corner (still has the same amount of whitespace) but the region in which you can click now extends all the way to the corner. The same goes for the new Spotlight widget in the upper-right.
I like the global menubar, but that may be partially because I’ve used Mac OS for year. (And GS/OS on the Apple IIgs before that.)
The top-of-the-screen menu bar does have one very useful aspect: an application may be running while it has no windows running. Yes, Windows users find this confusing (and seem to go to great lengths to kill off open applications) but it can be very useful. Say you’re working with a document and you decide to close it and open a new one.
On the Mac I just hit two keyboard shortcuts: Command-W (to close the current document) and then Command-N (to create the new one). On a typical Windows application, closing the first window would cause the application to quit. I’d have to then go hunt down the application and open it again.
To do it on Windows w/o quitting the app, you have to first open the new window, then switch back to the old window and kill it. (If you use the mouse then you can close the window w/o switching to it, assuming the new window isn’t covering the old window’s close box, but I tend to use keyboard shortcuts for this.) That’s a real pain, as it is not intuitive for me.
Before anyone jumps in and says “well, no one uses applications that way” .. well, I do, and I’ve noticed lots of “normal” users doing so too. It’s a common pattern if you’re working with normal office documents.
Plus there are lots of apps I like to keep running w/o having any windows open. There’s no good way to do this on Windows. (either you end up with a minimized window taking up space in the taskbar, or an icon in the tray, or some nonstandard application-depending thing.)
Ahh. Now that’s a good point in favor of the global menu.. Good to hear that the corner business has been fixed in Tiger too.
I started using a PC in the DOS times and I hated Windows the first time it came out. I remember cursing: “Why do it take me so long to move all the *.dat files from directory A to B with all these ‘intuitive’ drag-and-drops?”. Then I thought the global menu bar was the worst idea ever in a GUI and the Macs were so behind. But when I bought my first PowerBook I realized I liked this global menu bar; I find it much more convinient than any other interface I had come accross before
. Times change…
So is the law of gravity if you are trying to fly.
Unfortunatly that doesn’t affect its well tested
validity.
The top-of-the-screen menu bar does have one very useful aspect: an application may be running while it has no windows running. Yes, Windows users find this confusing (and seem to go to great lengths to kill off open applications) but it can be very useful.
Because it is confusing – You have app running without any visible sign of it running. And it takes memory, in some cases a lot of it.
Say you’re working with a document and you decide to close it and open a new one.
In windows? try ctrl+w and ctrl+n
to quit app try ctrl+q
It works with most apps (photoshop, word, opera, FF, gimp, etc).
(correction: ctrl+q don’t work with FF, some kind of bug, I suppose).
On a typical Windows application, closing the first window would cause the application to quit. I’d have to then go hunt down the application and open it again.
Try it again (ctrl+w + ctrl+n) – it works.
“Fitt’s law has nothing to do with UI’s or learning. It has to do with hand-eye coordination.”
Do you understand how contradictory these sentences are (except GUI of course)? Coordination IS based on training.
Slightly motorically challanged people, may have problems even in spite of training. There many things that affects our motorical abilities e.g. age, illness,…
Slightly challenged people will do better after training. That goes with any of your “exceptions”.
It does not change anything. This is not a comparison of two subjects, this is about one subject before anf after training, actuall reading of Fitt’s would help
While I’m a big Apple fan, I find a single menu bar that works well for the user can be as much a case of learned behaviour as much as it’s a case of better design. Look Fitt’s Law is great and all, but it cannot be the single source for the design and placement of interactive elements (whether it’s software or hardware).
One good argument, I think, for window-located menu bars is basic functional locality; the expectation that elements relevant to the action/work being performed be in the same functional location of that action/work. Now, I don’t disagree that Fitt’s law says it’s more difficult to locate the menu in the window than globally, except that we’re not just locating things to click on here. We’re also talking about behaviour and expectations of behaviour. When a menu item is selected the behaviour is expected to impact whatever is being worked on. From a functional locality standpoint, window-located-menus, maintains the simple paradigm that actions within the window affect it and actions outside of the window don’t (yes, the system tray can violate functional locality).
Global menus require that you leave the window, which brings with it the inevitable need to determine what application the menu will affect, verify that it is focused (if there’s a window at all), then make the menu selection. Yes, it may be easy to locate via Fitt’s law, but Fitt’s law is not the be all and end all of Interactive Design. There are many other things to consider (like functional locality and others) and tradeoffs to make. Apple erred on the side of being able to locate the menu bar. Others have made other tradeoffs (errors).
PS. Many people can learn to be productive with bad designs. doesn’t mean the design is neccesarily the best one.
In Tiger, You’ll have expose desktop, expose all windows, Dashboard, and spotlight in the corners!
You can’t possibly have slicker corner use than what’s offered in Tiger, sorry!
Umm..if I didn’t save it, I would be stupid! It is not destruvtive at all anyways. Programs have an “are you sure you want to close it”, remember?
Well, everyone knows that global menubars have downsides, like everything! The challenge is to weight up the pros and cons and then use what works best for you. Or at least that’s how it would be if we’d have the choice.
It’s important to realize that this could be implemented in a totally optional way, by using a similar approach to the freedesktop notification area spec. Toolkits or apps would be able to query for the existance of something that swallows the menu (it could be anything!) and if it’s unavailable, they would simply manage it themselves. Enabling or disabling this feature would become as easy as to add or remove the appropriate applet.
Regarding point one, this is actually one of the reasons why I tend to be in favor of a global menubar. To me, it clearly de-emphasises the menus, because they are away from the area where I actually do the work.
Point two is not an issue for me, because as you said, menus should be avoided for regular tasks. IMO the importance of the menu is mostly to list available functions and to serve as a shortcut reference.
Point three is refuted by the fact that it would be optional. None of the popular desktops use focus follows mouse by default, so this should hardly be a huge problem for you (and those which do wouldn’t have to provide a global menubar).
Point four doesn’t sound unsolvable to me (put a menubar on each screen!).
Point five would be semi-solved by a freedesktop specification. Of course there could always be old or rare applications which don’t support it, but the situation would not be worse than on OS X (in case both Gtk and Qt offer support for it).
Point six wouldn’t be more of an issue than with the taskbar.
There are downsides to a global menubar and I’m not totally convinced that it’s the best approach (right now my main reason to be in favor of it is, that it simply looks a lot better), but in any case I’d love to have a GNOME applet available for it. So let’s stop arguing and start working, damn it.
how can you access the menu bar using a keyboard shortcut in OSX ?
in Panther:
System Preferences, Keyboard & Mouse pane, Keyboard Shortcuts tab, then check the “Turn full keyboard access on or off” checkbox in the list box. Or press control-F1 to toggle it on/off instead. Afterwards, control-F2 opens a menu and as with practically everything else, you can start typing in the name of the menu item you want and the selection jumps to it. Only problem… control-F2 is hard to reach for, no? Thankfully you can customize all the keyboard shortcuts you want in that same preference pane. Change it to alt-m, control-j, whatever you want…
Similar to what someone mentioned before, when I first use a new Macintosh application, I systematically look through all of its menus to what its capabilities are. It doesn’t take too long to do for most apps. I can’t do that on Windows though, I have to actually try all the menu items and then right-click on every element to see if there’s something hidden away in the contextual menu.
Personally I prefer not having a menu bar on my second monitor. I do my main stuff on my main monitor (naturally), and the second monitor holds logs or a Safari page or a live preview of what I’m writing… for example, I’m typing this in an external editor on my main monitor as I read through osnews on my second monitor.
About the single-button Mac mouse… sheesh, haven’t enough people said to go out and spend $15 for a cheap USB mouse if you really mind that much? I started with a single-button mouse way back when, then I used a 2-button mouse with scroll wheel for about a year, and now I’m back to a one-button mouse. The Mac OS makes it so you don’t NEED a multi-button mouse. It’s completely supported but completely optional. My scrollwheel is space and shift-space, scroll wheel button is command-click, and right-click is control-click. When I use my computer, my left hand never leaves the keyboard anyway.
“The only thing I miss from OS 9 is the ability to make buttons”
Um… the Dock? Single-click is all it takes. That’s what the Dock is: “a kind of launcher for my frequently used apps.
“Try it again (ctrl+w + ctrl+n) – it works.”
I believed you for a moment there… until I tried it: I have nothing open, I click IE, I hit ctrl-w then ctrl-n, but nothing happens. I tried it in Windows Explorer too. Same thing. Please explain.
whether or not humans evolved since the 80s is debatable, their has been a full generation since, so really, yes their has been some evolution going on.
But my point was as a technology is around people longer they get used to it, it becomes more natural and the ease/speed at which they interact with it improves greatly. A person who has never used a computer will have a rough go at it the first few days/weeks, but in a few years, they will be very efficient with it. This becomes more true with people who have only grown up in a world of computers. Their physical/mental abilities at using a computer is far greater then someone who hasn’t used one. And people in general aren’t going to go back in the opposite direction since computers are here to stay.
A law such as fitts law will not hold forever, it will only work for a while. Tweaks to it may keep it going longer, but in the end it just won’t fit the way things are anymore. Same goes for laws like Moores law. They aren’t laws, they are just patterns that fit for that bit of time, but as the world moves forwards they fall apart.
The point of Fitt’s law is not really about HCI interfaces, but about human perception, which is relatively conservative on regards to change. Bascially, what it says simply is that bigger targets are easier to hit with a pointing device. Certainly, the speed at which a person hits a target might improve, but the overall rule suggests that even expert users will hit a bigger target quicker than a smaller target.
While this suggests that the screen borders are ideal places to put an interface element. (Because they are easy to hit.) It does not say anything about WHICH interface elements should go where.
“Certainly, the speed at which a person hits a target might improve, but the overall rule suggests that even expert users will hit a bigger target quicker than a smaller target.”
where did you get this impression? Definitely not from his (Fitt) work. There is clear assumption in Fitt’s two articles. His law applies to untrained subject, unless you are willing to correct him or you have some experimental evidences you are simply wrong.
training has very little to do with it. a mac user only has to “aim” the mouse in two directions, while a user of any other os needs to “aim” in four directions. with equal amounts of training, it would still be faster to do the one that requires less work.
“training has very little to do with it. a mac user only has to “aim” the mouse in two directions, while a user of any other os needs to “aim” in four directions. with equal amounts of training, it would still be faster to do the one that requires less work”
Training does have something to do with it. A user trained in mouse use is more capable of acquiring arbitrary targets than an untrained user. Not surprising, really: an untrained typists is not as proficient as a trained typist, correct? All human motor-mechanical operations are learned and certainly those requiring fine motor skills. Fitt knew this.
Besides, aim alone is not everything. A large target that is far away can be just as difficult to use as a close-by small target. Returning to the original point of interest is MORE difficult the further you are also by Fitt’s “law” (its not a law, of course, we just call it that out of laziness). So in general, Fitts law tells us a small something about user positioning ability and virtually nothing about user interaction other than inferred relative metrics (which are always in respect to the system design and capabilities)
For example, most people seem to neglect the fact that on a typical global bar (or whatever gizmo is put there such as an icon bar) on a multi-monitor setup, the bar exists on only one window. So if your monitors are side-by-side (which is terribly common) you lose at least one infinite plane (typically the right one). For most right handed mousers, it is already easier to move left-up then right-up, so this is significant in terms of interacting with the global bar. Again, there are many variables that can affect performance on a modern system.
The point: global menu bars have a very specific sort of system design in mind. They work best for single, full-screen apps on constrained screen sizes (and there, they are better than local bars by far — this is, after all, what all the testing has shown). Otherwise, why are toolbars not put in place of the global bar (or another infinite surface) instead of below the application title bar? Surely, toolbars are used more often and so should benefit more from an infinite target, correct? It should be pretty obvious that the extra mouse miles required to acquire an infinite surface on a large desktop destroys this notion. Save for one: the current mouse pointer is the best (and unfortunately underutilized) of all the infinite surfaces.
Another point: no one seems to mention that people generally do not have any trouble acquiring the application title bar to move an open window. It would be unreasonable to assume that acquiring the menu should be very much more difficult (the 4 directions is a misnomer — if you are working in an app then you at most must move two directions to the local menubar, same as for a global bar but closer.)
Here are some ideas: instead of being fixed in old paradigms of design, wouldn’t it be cool if we could harness some of the new features of modern systems to enhance our interaction with it in ways not previously possible? So that depending on the type of gesture (direction, speed, landing “zone”) the mouse either knew to “jump” (and stick) to the right place? Or perhaps have various objects magnify as they are approached to make their target momemtarily larger (like the Apple dock, but generally more useful)
Here are some tips for folks who posted in this thread:
— bind a window close key (whatever appropriate for your OS) to a seldom used button on your modern mouse. I like putting it on the thumb button. I do find that users not trained for that will often accidently close windows — but usually not more than once or twice; I do find it most useful on windows (shudder) with IE (shiver) where you may open 6 or 7 windows rapidly and close them just as rapidly. With a tabbed browser (again, no one seems to complain about acquiring a tab location!) middle clicking the tab to close it is generally a better option, IMO. Best of all, of course, is to learn the keyboard equivalents.
— in windows (shudder) when you close an app window, you really do close the app (except for those strange close to tray apps). So if you really want to close your current work and start a new window, the usual thing to do is use File/Open (Alt-F O). That is, you simply “re-use” the window.
I’m all for utilizing all available infinite planes, but I feel that the screen edges belong to the maximized app that extends to those areas. Generally, this means your desktop so things that control your workspace are best placed there — which is why launchers are there. The reason? You generally don’t want to return to your last place of focus when manipulating your workspace so this one-way interaction takes the positive side of Fitt’s law (acquiring the remote object).
Further, it is more intuitive — every container is uniquely delineated and unchanging. For this reason, I tend to think the desktop should always own the screen edges. After all, modern systems are built on desktop metaphors so a user is most likely to associate the “monitor” with the “desktop”. Not doing this leads to this typical windows behaviour: if you put an auto-hide bar on the left or right side of your desktop you will find that a maximized window occludes that edge–you can’t acquire the destop edge to activate your launcher unless you un-maximize your application. I think that is terribly counter intuitive. Of course, if applications decided to use autohide bars it would probably make sense to have the windows like behaviour but of course that is a usability nightmare in anything but full-screen.
You need to read Fitt’s papers. Everybody is tossing around “Fitt’s Law”, evidently without understanding the basics. This law is not about computers obviously, computers have not existed in the times when he published his ideas. Unless you do really understand the basics of neurophysiology, this is only funny. Coordination is the result of training. I really don’t care about computers, but most of the post here suggest a lot of ignorance, because people are disscussing a theory without even reading it. Wikipedia is not enough.
Fitt published his ideas in the Journal of Experimental Psychology. If you want to apply it to the computers that is fine, the Law will apply here too, but citing equations suggest only how far from this Law you are.
Your notion that it has nothing to do with training is only one more ignorrant statement. The fact that you are knowledgeable in GUI design does not mean that you automatically get Fitt’s Law.
putting the application menu first in OSX is a mistake in my opinion. It used to be that every program had File, Edit, View, and then some other random ones. That kept a large portion of the global menu fixed in the same place. Now, the app menu pushes the rest of the menus around depending on how long the apps’ name is. I think a good way to fix this is either put the app menu in the middle of the global menubar, on the right side, or give it some fixed horizontal space that allows it to display about 12 characters at normal size, and then would scale down smoothly to fix this area if the name were larger.
Personally i think putting it in the middle would look cool, and be pretty functional. Can anyone write a haxie?
People complaining that MacOSX should give developers the option to switch between global and local menus should just stick with linux and stop pining. Thats one reason why linux is good- It provides lots of choices to satisfy people with custom needs and/or preferences.
Mac behavior is different than Windows or Linux.
All these folks who are used to Windows and Linux want the Mac to morph into Windows and Linux-like OS.
I suggest that you folk stick to what works for you and not mess around with Mac. What would then differentiate a Mac from Windows or Linux? Next you will be asking for the same viruses and trojans!
Although, I would agree that the Application menu can be shifted back to the right corner as it was in Classic. It still does not work for me in its current position. And ‘quit’ under file menu worked fine for me even though it meant quitting the application and not the file.
Global menu it is for me.
Personally, I thought the idea one developer had where you can have global menus as an OPTIONAL feature by placing them on a panel was pretty sweet, basically in gnome you’d have it as just another applet on your panel, that when used would turn off the menus in the individual windows. this gives best of both worlds, mac people get their way, and the rest of us do too.
Why do I always find my self longing for them when I’m on a PC running either Linux or Windows?
I like having the menubar show what application is in focus.. and I always know where to find it. Applications have a tendency to take over the entire screen.
I like how KDE gives you the option to put the menu bar where ever you want it.
-Dennis
I’m driving my Mac for 10 years and the presence of the menubar is natural to me. It’s not in the way, you cannot overshoot it, there are no n amount of menubars [where n is the amount of open windows].
I’m also using a nice Dell POS Win XP at work and I’m always trying to see the whole menu [I don’t use that switch context with the mouse thing, it’d drive me crazy].
When I maximise the window, the menubar is on top, just as with the Mac.
Actually: here the Mac shows its finesse again. The Windows menubar is not set to infinite height. You actually see the title of your window above the menu so you can overshoot the menu to end up in the title bar. Not possible with the Mac.
I never have to worry what app is in focus: the top left corner lists the name of the active app [OS X of course]. The menu is your steady compass, you always know where to go to. Apple did a great job with that menu. Nobody has done anything significantly better.
That’s not to say that when you like the Windows way better somehow you’re a dimwit. Not at all. Live like you want to live, baby. I’m not judging you.
And to the power user with the Terminal window: Command n opens up a new document/window since forever. You should know that by now. Even Windows does that.