Federkiel writes: "People working with Apple computers are used to a very consistent user experience. For a large part this stems from the fact that the Lisa type of GUI does not have the fight between MDI and SDI. The question simply never arises, because the Lisa type of GUI does not offer the choice to create either of both; it's something different all along. I usually think of it as 'MDI on steroids unified with a window manager'. It virtually includes all benefits of a SDI and and the benefits of an MDI." Read on for how I feel about this age-old discussion.
Post a Comment
Member since:
2005-07-08
Fans: 0
I am not an Ion/RatPoison user, and I don't know if I *could* become one, because for some reason I prefer to have control over the exact size and position of my windows (this size is seldom "maximized", except on very small screens). But it is interesting and I think relevant that they follow a sort of meta-fitt's law/rule: when using the keyboard (as we usually are), the fastest point to reach on screen still involves your keyboard.
That said, I've observed that conversations of Fitt's law and the archetypal windows/mac interfaces concentrate on the corners and top of the screen. I think this is probably because mac users are more aware of the design philosophy of their platform, or even just more aware of design philosophy itself (as a whole) than windows users. This isn't really a criticism of either camp, and since I don't really know a lot about either anymore, is probably based on inaccurate stereotypes.
Still, the discussion almost always ignores or merely mentions in passing the fastest point to reach on the screen, the fifth point of fitt's law/rule, your pointers current position.
Windows and the applications that run on it (and most Linux applications, too) heavily use this position for contextual menus. In my linux environments of choice, the context of the root window is a list of my programs, which when my pointer is not inside the context of an application further makes use of this point. On interfaces coming from apple, they still use it, but less heavily, and usually less contextually (because they always include these options elsewhere on the interface).
I think that mouse gestures were an attempt at bypassing the limitations of "aiming", but I feel that people with physical handicaps or people who are just plain not practiced with a mouse will probably struggle to do any but the simplest mouse gestures. Does anyone know of any studies done on this that prove or disprove my guess? Have there been any other ideas in user interface history to utilize the fifth point?
Member since:2006-01-02
Fans: 0
Some of us also move our hands when we talk, and move our mouse as we read. One more type of person to go on the list of people that can't get the hang of mouse gestures.
Personally, I think context menus are a great way t use the fifth point, because they allow you to do stuff to what's at that point, which is easy to put into a mental model. I right-click on a nut, I want to see wrench in the list.
IMO, getting farther from the desktop metaphor (like OS X does, XFCE 4, KDE 3, and Gnome 2 allow, E basically forces, and KDE 4 is doing) is more important than worrying about Fitts, except in so far as figuring out exact placement (like the stupidity of leaving a useless 2px border at the bottom edge of the Windows taskbar until Windows XP, or similar borders in KDE, XFCE, and Gnome themes, and other borders that get in the way of otherwise big targets).
Edited 2007-06-25 03:36
Member since:2006-02-04
Fans: 2
Personally I find having all elements of the application (file menu ect.) within the context of that applications window.
I dislike on a Mac having to first bring focus to the appropriate application on the desktop and then having to move to the file menu at the top of the screen to do something with it.
When the file menu is part of that main window, I can bring focus to the application and access the file menu with one action instead of two. Seems to be more efficient to me.
As others have pointed out, there is a matter of training. I'm a long time Windows user. I don't use a Mac as a main machine but it wouldn't surprise me that if I did, the change in interface wouldn't bother me for long. The reason being is that those functions more typically accessible within elements of the document interface or within contextual menus and keyboard shortcuts.
Now as far as "aiming" goes, I suppose mouse gestures could be an attempt bypassing that limitation, but it doesn't have much to do with one design philosophy from another as the forward back keys buttons are within the main application window regardless.
I don't know about any studies, but what might be useful to users who have difficulties with gestures as well as aiming is actual forward back keys. For example Opera. You have the choice between buttons, mouse gestures and the z and x keys to go forward and back between pages.
Member since:
2005-06-29
Fans: 1
Menu bars and such are only a part of the equation. There are many other variables. For example MacOS X and its applications are designed in such a way that it makes using the system very intuitive. When you want to do something on a Mac you just have to do it the most natural way, for example by dragging and dropping something. What I observed in Windows users who switch to a Mac is that they try to do everything in a complicated way, always looking for a workaround when it's not really needed. For some reason Windows leaves an imprint on users' minds that you cannot just do anything easily, there's always some inventive "hard" way. And that's what they try to do and it doesn't always work that's when they start complaining. So actually when you make a switch it's always better to forget what you knew completely and start from scratch. Otherwise however productive and helpful Mac interface can be in theory, a user wouldn't be able to take advantage of it.
Member since:2005-06-29
Fans: 20
And that's what they try to do and it doesn't always work that's when they start complaining.
You just don't seem to get it. If a user knows how to do action A on Windows just fine, and then OSX uses a different method (which you deem 'easier', something you cannot prove AT ALL), then it doesn't mean either of the two is "the hard way". It means just that - that both employ a different mean to achieve the same end.
On top of that, dragging and dropping is overrated. It is actually a VERY complicated and muscle-straining way of doing things. Using keyboard shortcuts or context menus to copy/paste things can not only be faster, but also easier on the muscles.
So actually when you make a switch it's always better to forget what you knew completely and start from scratch.
And that's something you cannot do, so this is a completely irrelevant remark.
Otherwise however productive and helpful Mac interface can be in theory, a user wouldn't be able to take advantage of it.
Ok, you REALLY didn't get it. The interface that is the best for user A is the interface that makes them do tasks in a way that is easiest and most familiar for them. If a user has been using Windows for 15 years, then the Mac is simply (probably) not the best way of doing things. It MIGHT become the best way, but that can take years - it might not happen ever.
Experience and training is not something you can just brush aside - something many self-proclaimed "usability experts" seem to do all too easily.
Edited 2007-06-24 15:05
Member since:2006-04-21
Fans: 14
Ok, you REALLY didn't get it. The interface that is the best for user A is the interface that makes them do tasks in a way that is easiest and most familiar for them. If a user has been using Windows for 15 years, then the Mac is simply (probably) not the best way of doing things. It MIGHT become the best way, but that can take years - it might not happen ever.
I think there's something you aren't getting either: Everyone is different. Fred Brooks reports that in the old days, when everyone who had a computer could afford to pay onsite programmers, each company would have its own payroll program, and that payroll program would be written to mimic, as best as possible, the company's paper accounts system. Now that everyone and his dog uses Excel, everyone fits their accounting practices to it (and it would be the same if they were all using Oo.org Calc, VisiCalc or 1-2-3).
In the same way, since not everyone has the time, inclination, or ability to design and write a user interface for their operating system and applications, everyone makes do with the one they're given (whether their system of choice/work is Mac, Linux, or Windows). Even Linux, which in theory allows you to use any number of interface designs, is moving towards "the big two".
But that's far and away a different kettle of fish than if someone did studies (on virgin subjects) to determine which (if any), of a myriad of different interface styles people liked. And the one that grabbed the most amount of people in total is very much more likely to be the one that the greatest amount of people disliked least, than the one they liked best.
Member since:2005-06-29
Fans: 20
I think there's something you aren't getting either: Everyone is different.
Exactly, which is what I said in the article. And because of this, it is hard (impossible) to claim that one design is better than the other - you can, at best, claim that one design is better than the other, for you.
But that's far and away a different kettle of fish than if someone did studies (on virgin subjects) to determine which (if any), of a myriad of different interface styles people liked. And the one that grabbed the most amount of people in total is very much more likely to be the one that the greatest amount of people disliked least, than the one they liked best.
How nice that is in theory, that research would most likely prove to be fairly useless in the western world, since everyone here basically has experience with graphical user interfaces.
Member since:2005-07-07
Fans: 3
Member since:2006-09-01
Fans: 4
That and had the brightest shiny. My roommate's girlfriend had him install Ubuntu on her laptop <s>because of</s> for Beryl after she saw him playing with it.
Not saying that it would be the only reason that people would choose one but attractive packaging does help.
Member since:2005-07-07
Fans: 3
Member since:2005-06-29
Fans: 1
We're talking absolute values here. Okay, somebody might be accustomised to doing everything in Norton/Midnight commander and doesn't want to learn a real CLI or GUI at all. Does it make them more productive? Is it a better way? The answer is "no", just because somebody doesn't want to learn to do things better doesn't mean that Norton Commander is the best interface out there. Same thing for Windows and Macs. It takes a Windows user 15 steps to complete a task while in MacOS it would take only 4 steps, so naturally if the user learns how to do things right it would save him time and make him more productive. The truth is, computer interfaces limit humans in different ways and we're in constant search of new and different ways to make the user more productive. That means learning or re-learning. And honestly I never understood that kind of stubbornes - "I won't learn because I WILL NOT!". One explanation that comes to mind is that people are so afraid of computers that it's a miracle when tasks can be successfully accomplished. Then even if it takes them 30 steps to do something, they'd rather go the longer route because they're afraid that a different method will break everything. It's psychology. It doesn't mean that doing things the way a user doesn't know is inferior. Crying about it doesn't validate it. Consider a photographer who used to do everything on film and is being confronted with the digital workflow. There's no reason to dismiss it and live in the stone age, so he goes and learns Photoshop and Aperture and the color space theory and everything.
Member since:2005-06-29
Fans: 20
Okay, somebody might be accustomised to doing everything in Norton/Midnight commander and doesn't want to learn a real CLI or GUI at all. Does it make them more productive? Is it a better way? The answer is "no", just because somebody doesn't want to learn to do things better doesn't mean that Norton Commander is the best interface out there.
If that user prefers Commander, then yes, it is the best interface - for him. Just like how you prefer the Finder - making Finder the best interface, for you.
It takes a Windows user 15 steps to complete a task while in MacOS it would take only 4 steps, so naturally if the user learns how to do things right it would save him time and make him more productive.
The world isn't that simple. The tasks a computer has to perform cannot be brought down to "a task" requiring "x number of clicks". Using a computer involves millions of little tasks which all require a certain amount of user input. I'd think you'd have a very hard time trying to prove that the Mac requires fewer click to operate than a Windows box.
And even if you could, it is irrelevant, since a trained Windows user might perform his 15-click task faster than the equiv. 5 click action on the Mac, due to training, and preference.
Member since:2006-02-15
Fans: 0
Glad you brought up the midnight/norton commander.
Something that's great about the midnight/norton commander is the predictability of it all. After building up some muscle memory, I can manipulate files almost faster than my screen can refresh, people watching me work just see the screen flash a bit ;-). Combined with the CLI, it's great. I still use it daily, even though it's a bit dated and clunky in some parts (the editor is only nice because I'm used to it, and very clunky when it comes to copy/pasting for instance)
That being said, even though graphical variants have been tried (a lot) both in windows, linux and other operating systems and desktop environments, somehow they never work out. Someone interested in usability should really look into why it worked in the first place, and why it fails so badly in mouse-based/wimp environments.
I think it worked because of its source/destination system (the 2-panel layout always made clear where your file would end up), the always available command line (it doesn't get in the way of the CLI, on the contrary it augments it), the integrated file search/filtering/comparing, and the integrated editor - and most of all: absolute predictability of the commands.
Maybe I should write my own eventually, not necessarily using the same layout (2-panel + CLI) but copying some of the concepts, and adding some modern features in the mix.
(Something like http://hotwire-shell.org/, the old http://en.wikipedia.org/wiki/XMLTerm, using modern editors/viewers, and try to get away from being "filesystem based", allow different views based on its content/metadata. I still think you need an source/destination thing, so 2 panels might stay ^_^)
Member since:2007-06-24
Fans: 1
Member since:2005-07-06
Fans: 7
"It takes a Windows user 15 steps to complete a task while in MacOS it would take only 4 steps, so naturally if the user learns how to do things right it would save him time and make him more productive."
Show me a task that takes 4x more steps on a Mac. I'll be waiting.
Member since:2005-07-06
Fans: 13
That is not what he/she said.
The OP said it takes Windows users 15 steps to complete a task where a mac user can do it in 4.
I know, exaggeration, but lets see what he replies to this rewording of your question....
Show me a task that OS/X users can do in 4 steps, that would take 15 on Windows ?
Member since:2007-06-25
Fans: 0
he talked about being "more intuitive", not "easier"
maybe. but it's LESS INTUITIVE. do you get it?
why can't you? I did it, and a lot more people out there are doing it... windows -> macos, windows -> linux... and the larger part of them are not complaining at all.
that's YOUR opinion. I switched to Mac two years ago, having used Windows since Win 3.0. It took about *two weeks* to be as productive as in Windows. I'd never go back. so?
Member since:2006-09-01
Fans: 4
Member since:2005-11-12
Fans: 3
Member since:2006-09-01
Fans: 4
It didn't say it was unnatural.
I said it was more complicated, as in more steps are used to accomplish installing an application.
I like the entire package design as it is more sreamlined, and it makes backing up my Mac easy. What I don't like is the extras steps needed to get something installed. I'd rather be able to right click on a dmg and select an install option.
I guess I could write a script to do that, but I haven't looked into it.
Member since:2005-11-10
Fans: 14
What, and pressing Ctrl+O is any different on Windows?
That's just one way of opening a file, it's not like you can't double click on a Mac.
I think what you were trying to say is Cmd+Down to open a file, like pressing Return in Windows. In OS X it's Cmd+Down to drill down folder / open files and Cmd+Up to return. It's Return / Backspace to do the same in Windows.
Most Disk Images for Mac Applications now include an alias to the Application folder, so you only need to drag the application an inch to install it; how is that any stupider than going through an MSI in Windows?
Member since:2006-09-01
Fans: 4
Ctrl+O doesn't do anything in Windows.
Most of the time I select the window then type the name of the file to jump to it instead of scrolling through it with a mouse. I don't have to move from home row if the open file hotkey is enter. I do if it is Cmd+O.
I haven't tried Cmd+Down. Is that something specific to Mac's browser view? I'll have to test that when I get back home. That's not any better though. I still have to move off home row.
I use spatial view and Cmd+O is what works for me. I like the view where the folders scrolls inside the window; I can't remember what it's called at the moment though. I like how the structure can be navigated with just the arrow keys then opened with the open hotkey. I just don't have enough screen real estate to really run it the way I want to, 12" PowerBook.
Some of them don't have it. There are still a lot of DMGs that are just the application. If they have it it's nice, if not then I have to open the applications folder first. I'd still rather just right click, select install, and have the application magically appear in my App folder.
I wasn't trying prove Windows was better. I was just trying to point out even the Mac platform has it's quirks that could work better.
Member since:2007-01-30
Fans: 0
Member since:2006-02-04
Fans: 0
well seeing as F2 and enter are the same number of buttons, actually it sorta is AS simple
and seeing as I don't tend to spend much time renaming files by hand in the file browser.
and seeing as it's also available from right clicking.
and seeing as i can also simply click on a file a second time after the double click time window is up.
I'd still say that enter, which in pretty much every application ever means activate/confirm, is the better choice.
Edited 2007-06-24 18:03
Member since:2006-08-08
Fans: 0
Pressing CMD+O isn't easier, it's an extra keystroke and the keys aren't close to each other really. But that's not the point... this isn't about easy.
On a Mac, CMD+O is what you use to open anything, whether a folder or drive in the Finder, or a document in a program. On Windows ENTER only works in Explorer, and then you have to use CTRL+O within an application. Two different key strokes within the OS to do the same thing. Opening something.
I'm not sure if anyone else has had this (I'm sure I'm not alone) but just pressing Enter is a bother when you've selected a bunch of icons and accidentally hit ENTER and they all open! The CMD+O ensures you don't make these mistakes.
It's about the consistency and thought
Member since:2006-06-19
Fans: 0
"For example MacOS X and its applications are designed in such a way that it makes using the system very intuitive. When you want to do something on a Mac you just have to do it the most natural way, for example by dragging and dropping something."
God, I hate the way it is assumed that dragging and dropping makes tasks easier or more intuitive. If anything I find the opposite. As an experienced computer user, I am still never sure when dragging and dropping from one folder to another, if it will copy or move the file in question. The best implementation of I have used is in KDE, where dragging and dropping brings up a context menu offering the option to move, copy or cancel.
As a software developer probably, the most unintuitive load of crap I have to deal with is visual query builders. I have the SQL all worked out in front of me, but trying to get these things to generate the same query is often nearly impossible if they have any form of complexity in them.
Member since:
2006-04-21
Fans: 14
...I'm glad someone else finally admitted that there's no way to measure usability, because I've been saying this for years. Having said that, the each-application-has-its-own-menu approach (which both Linux* and Windows suffer from) really sucks: It really wastes a lot of screen space, and despite Thom's comment about consistent menus, different applications can only have consistent menus up to a point - there's no point in having "ray tracing" options in the Tools menu on a browser, for example, or "Bookmark this page" options in AutoCAD.
*KDE mitigates it somewhat because you have the option of having the menus at the top of the screen, but this only works for KDE apps (surprise, and I'm not suggesting that there's anything they can do about it). And no matter how much I try to stick to KDE-only apps, sometimes I have to use something like Firefox or (gasp) pidgin.
my £0.02
Member since:2005-06-29
Fans: 20
Member since:2005-07-11
Fans: 2
and despite Thom's comment about consistent menus, different applications can only have consistent menus up to a point
I believe Thom was talking about the names and placement of the menus, which should be consistent across apps, and not their contents, which are (obviously) application-dependent.
The menu bar should look the same across apps, and for the most part, they do:
KDE:
File Edit View ... Tools Settings Window Help
WinXP:
File Edit View ... Tools Window Help
Don't have access to a Mac, so can't comment on the order of the menus there.
Member since:
2006-09-01
Fans: 4
Gui design is entirely subjective. There is no right interface just like there is no right color. Red is not better then yellow, but I do prefer red cars.
People will adapt to whatever the designers decide. I was looking at a mockup for KDE with the KDE menu in the middle part of the bottom of the screen. That flies in the face of the "the menu should be in the corner" wisdom. Will it work? Yeah, people will get used to it after feeling odd for about a week. If someone really wanted to make a useful UI, it would tailor itself to the user, or they should get rid of most of it altogether. (<--Subjective Opinion so totally invalid)
Microsoft's surface interface looks really interesting from a UI point of view. I'd like to see them adapt it to a regular mouse and keyboard desktop though.
As far as Win versus Mac goes. I personally like to have things in neat little packets, it's my nature, so the application centric interface feels best for me. It also helps me switch between tasks, it's like context switching for my brain.
My main thing against document specific interface, or at least Apple's version, is that the menu bar changes. I like things to stay the same, and I learn hotkeys for any frequent task, which cuts down on my usage of it. I'd rather use the space for something more productive.
Lastly, I think you should check out Office 2007's new ribbon interface. It really is pretty cool, and much more useful then the File menu plus toolbars approach.
Member since:2005-06-29
Fans: 20
Lastly, I think you should check out Office 2007's new ribbon interface. It really is pretty cool, and much more useful then the File menu plus toolbars approach.
I'm a total addict of the ribbon interface element as used in Office 2007. Great stride forward compared to 2003's interface.
Member since:2005-11-10
Fans: 14
Even as a Mac user, I've followed the development of the Ribbon interface extremely closely. It is one of Microsoft's best interfaces and even gives me some Windows-envy. The ribbon is scheduled to appear in Office:Mac 2008, albeit in a slightly more traditional way.
Coming up with a way to get rid of 100's of menu and toolbar buttons, and end up with something that's actually more productive than before is as big a UI innovation to me as the migration from DOS to GUI. It makes the GUI usable like it was supposed to be in the first place
Member since:2005-07-06
Fans: 0
I used to know where things were on the menus, and some of those things aren't even present on the ribbons. Secondly, Microsoft really screwed over OLE integration with the ribbons. Used to be that there was menu merging when doing embedding. Now Microsoft doesn't support it--it has the ribbons instead. Plus the ribbon takes up more screen real estate.
Member since:2005-11-10
Fans: 14
The ribbon _DOES NOT_ take up more screen real estate. In fact, it takes up less space for the same amount of total UI (i.e. turn on all the toolbars in Office2K3, the ribbon never gets taller).
Read this and get edumacated:
http://blogs.msdn.com/jensenh/archive/2006/04/17/577485.aspx
Member since:2005-07-22
Fans: 0
>Gui design is entirely subjective.
Let's lay this myth to rest once and for all.
Imagine a GUI that is controlled by a mouse, but the mouse pointer moves in a random direction every time you move the mouse. Clearly, this would be bad for *everyone*.
A mouse-driven GUI where the pointer always moves in a consistent direction is, therefore, a demonstrably better design than a similar GUI where the pointer moves in a random direction.
Therefore, GUI design is not *entirely* subjective. There are at least *some* absolutes.
Of course, this does not mean that GUI design is not *partially* subjective.
Member since:
2005-11-12
Fans: 3
This really makes sense since most people read top left across and down. Microsoft made the menu in the left bottom corner popular but to me thats just screwy.
I like how OS X put the window functions on the left hand side, this makes sense since most of what you do is on the left hand side anyway.
Member since:2006-07-13
Fans: 0
Member since:
2006-02-04
Fans: 0
From the article: "But the human visual sensory equipment has limits. There is a limit of how much information you can get on a area of a certain size. And there is a limit to the area the human eye can usefully overview. And while there are people that are working with two, three or more screens, this is only the exception."
Doesn't this actually argue for keeping the menu WITH the window.
Also "Microsoft may have recognised the folly of their action back in the ‘80ies; Microsoft's Office 2007 suite behaves very much like a Lisa-type user interface when in full screen. But only if in full screen. And they are constantly trying new user interface concepts. The new Internet Explorer hides the menu bar in the default configuration, don't be surprised if it re-appears on the top of the screen in Version 8. And the new Microsoft Office 2007 plays with a interesting new concept; Ribbons. Maybe they'll come up with something new altogether, who knows."
This has nothing to do with one placement of the menu being better than another and everything to do with menu's sucking balls regardless.
Member since:
2006-02-05
Fans: 13
First off, it is more then personal taste, some interfaces ARE better then others. A better designed interface takes less effort (i.e. the brain consumes less energy while using it)
As for the article, I think pretty much everyone abandoned the idea of SDI awhile ago, in favor of permanent maximized document window switchable with tabs. This is because the SDI was a prime example of microsofts attempts to copy something they didnt understand. The spatial design metaphor worked because the entire operating system functioned under it. When you start mixing spatial stuff with other stuff, the whole thing falls apart. The whole reason the spatial design works is that it is consistant, and you can count on thinking of objects as objects. This is why the spatial explorer was horrible in win95, and why spatial nautilus was only slightly less horrible in gnome.
Last point is ribbon, which IMHO is one of the best UI innovations to come out of Microsoft since the tool tip. Menus on windows suck, and microsoft should be lauded for their attempts to get rid of them. My only criticism is that it seems like every vista app takes a different approach. You have the ribbon in word, tabs and button menus in IE, a contextual toolbar in explorer, mode switching/menu buttons in WMP, and an almost webpage style UI for windows defender. While ribbon is, IMHO, the best of the bunch, MS really needs to sit down and decide on some standards, because vista looks like it was conceived by a designer with ADHD.
Member since:2005-07-06
Fans: 7
"My only criticism is that it seems like every vista app takes a different approach. You have the ribbon in word, tabs and button menus in IE, a contextual toolbar in explorer, mode switching/menu buttons in WMP, and an almost webpage style UI for windows defender. While ribbon is, IMHO, the best of the bunch, MS really needs to sit down and decide on some standards,"
The Office team has always tried new GUI stuff first...if it works (which in the case of the ribbon, it definitely seems to be), the rest of the OS will soon follow. It eases the burden of transition a bit financially. I expect ribbon menus to appear in Vista in SP1 or SP2 given how well it's doing in Office.
Member since:
2006-10-11
Fans: 2
What I don't get is the obsession with short ways and speed, that's a manufacturing belt mentality and reduces the human being to a time factor. The value of my work, computer work BTW, is not measured in units but in quality. My employer couldn't care less if I performed a repetitive mundane computer task with a 100 or a 1000 clicks a day as long as my code is A-OK.
That being said I love my setup -- menus pop up under my mouse pointer if I press hot-keys on the keyboard -- I'd hardly say it is the fastest way of doing things, but it works for me.
Member since:2006-03-08
Fans: 0
It's not so much about being in a hurry as it is about getting rid of unnecessary work.
I have to use my computer every day, and pretty much all day. Two extra clicks for doing a task starts to get annoying pretty fast if there is no good reason for the extra clicks to be there. Especially for people used to the expressiveness of the commandline.
Regarding your employer, if it takes you twice as long to do your work because you spend most of your time doing 1000 clicks when you could be doing 100, I'm pretty sure he'd start caring about why half your salary is spent clicking around when you could be doing actual work.
Member since:
2006-03-15
Fans: 6
All of this stuff that you guys are commenting on, back and forth, is something I've been wrestling with for a while now - why it seems that everyone claims to be an expert on what's "intuitive" (FWIW, I consider myself a student of usability, but no expert but there are seemingly endless debates on what's right.
Usability is both an art and a science -- both objective and subjective.
Some components of usability can be easily verified scientifically, such as Fitts' Law, the time savings in using keyboard shortcuts, or why it's so time-consuming to switch between the mouse and keyboard. Other aspects aren't so simple, like what constitutes a logical order in window control placement for a particular task or what would be a suitable toolbar icon for a "Compile" icon in an IDE.
Case in point: the menu bar placement. This is an age old debate -- each has its pros and cons.
A global menu bar (like MacOS) consumes less screen space and, according to Fitts' Law, is easier to click on because the user need only concern himself with horizontal aiming. Having a menu bar in each window has a natural mapping of tools to the document, as you have said.
Both cases deal with mode -- the current app/document -- in different ways. The global menu bar provides a consistent place for the user to look for what app he is dealing with but at the cost of a visual disconnect between the menu and the document window. Window menu bars utilize nearness to connect tools with a document but at the cost of being slower to click on and more screen space. These are the empirical parts of menu bar usability. Which one is easier for a user is partly the human factor, but also it's a matter of training / habit.
We all have some low-level training, such as using the left mouse button for regular clicking. The art of usability is really a balancing act between habit, training, scientific facts, and the ever-so-unpredictable human factor. It's this fuzzy, subjective stuff that causes all the bikeshed debates, or at least as far as my experience has been.
Member since:
2005-09-21
Fans: 5
Everyone loves to talk about Fitt's Law, because it's quite well established, and very simple.
Unfortunately, like Thom said, it doesn't take training into account. It also doesn't take inexperience into account. I have never seen an inexperienced computer user take advantage of Fitt's law. Even if the start menu is in the bottom right corner, or the close button in the top right, they still carefully move the mouse cursor on top of the button and click. Fitt's law, while a valuable measure, is not as important as everyone seems to assume.
As for the global application bar, I'm against it, mostly because I do take advantage of Fitt's law, but I very rarely need to access the menus in an application, and would much rather have other functions in the top corners of the screen (minimize and close). If I need to use a function often, it's in a toolbar or a keyboard shortcut. I only very rarely use menus at all.
Thanks Thom for a great article. People need to stop thinking of usability as a black and white issue. I resent desktops which force one interaction model on me and then claim that it is optimal. I decide what is optimal, not them.
Member since:2006-07-30
Fans: 2
Oh, it does take training into account but not explicitly.
The general formula "T=a+b*log2(D/W +1)" is the same for trained and untrained but the constant "b" would be a lot lower if you're trained to using this interface.
Let me explain what this means by a simple example:
Let's say we have a 1cm target (W=1) 30cms away from your pointer (D=30). This yields:
T=a+b*log2(31)=a+b*5
"b" could be 1 for a very inexperienced user and 0.1 for a trained user. "a" is probably mainly the reaction time + time needed to decide which button to hit, somewhere between 0.2 and 1s.
Let's say it's 0.3 for the pro and 1 for the noob.
noob: T=1+1*5=6s
pro: T=0.3+0.1*5=0.8s
So our pro could still be 7-8 times faster, all within the mechanics of Fitt's law.
Now let's say the target was 2cm and 10cm away:
noob: T=1+1*log2(10/2 + 1)=3.6s
pro: T=0.3+0.1*log2(10/2 +1)=0.6s
While the pro would still be 6 times faster, he would also be 33% faster than with the smaller target at a bigger distance.
Of course, the noob would be 1.7 times faster so any change in UI layout has the biggest effects on noobs.
So while the influence of Fitt's law is most easily seen for noobs, it is still valid for a trained person.
Edited 2007-06-24 18:30 UTC
Member since:
2006-08-23
Fans: 0
As I commented to TFA, I always experience mingled amusement/annoyance at reading these methodical argued "analyses" of GUIs where it seems clear that the author has only used 2: Windows and the Mac. There are many other ways; NextStep's cascading context menus from a fixed position, for instance.
But one I think should always be considered, as it shows a genuinely different model, is Acorn's RISC OS. No fixed menus at all; as the 1st commenter says, the easiest menu to hit is the one which requires you not to move the pointer /at all./ Other things RISC OS pioneered 20yr ago are OS-integrated global font anti-aliasing, solid window drag, the global taskbar - all now widely used - and drag-and-drop file open and close instead of file selectors with a mini-browser, which, alas, no-one else has adopted.
You cannot judge these things on a sample set of 2. Especially 2 which are actually quite similar.
If you want to judge GUIs, go out and master half a dozen different ones first. *Then* and only then can you make informed, educated comment.
The same goes for most things, of course, from word processors or operating systems to types of cheese or varieties of beer. Who would take a wine critic seriously who had only ever tasted 2 wines?
Member since:2005-07-06
Fans: 11
I can do pretty much all of that on my NeXT from 1991 too 
It has fixed menus on the top, left hand side of the screen AND you can activate the same menus from anywhere on the screen with the right mouse button. It also does AA, full window dragging, etc etc... And its really fast and stable on a 33mhz 040 with 128mb of ram 
Heres a screen shot of my NeXT... WHich I'm probably selling, sadly :'(
http://helf.freeshell.org/NeXT-9-4-06.jpg
Member since:2005-07-06
Fans: 11
I also forgot to mention that the GUI on NeXT machines and on risc OS macihnes ( I used to have a RiscPC 700, loved it...) do what a GUI should do. They are intuitive, fast to get around in, and best of all, STAY OUT OF HTE FARKING WAY. I don't want shiny shit bouncing around everywhere. I want to work and not be annoyed
great, now I'm talking myself out of selling my NeXT. poo 
I need that money
Member since:2006-08-23
Fans: 0
Nice!
But for me, the idea of a context menu that you have to move over to seemed to be a bit... well, daft, frankly. If I summon a menu, I want it where I am now! If I have to go to a specific place, then why not just leave the menu always there, in a single cascading tree, like Nokia's Series 90 (Symbian) or Hildon (Maemo Linux) GUIs?
Alas, I have /very/ little NeXT experience. I've always wanted one, though. I'd happily buy yours, except that I am in the UK and you're probably not, and I can't afford to spend any more money on old computers while I have so many...
Member since:2005-07-06
Fans: 11
Yeah, I'm in the USA. Shipping it to the UK would be around ~80GBP probably.
I'm not quite sure if I get your meaning on the menus without seeing that in action... But on the NeXT you can 'tear' off menus and make them stay on the screen if you are going to use their fucntions a lot. Comes in quite handy. you can also rearrange most of the screen.
Do you have any pictures showing what you were talking about with the menus?
You can summon menus anywhere your pointer is on the NeXT. using the Right mouse button. so you have a context menu that stays on the screen in the top left and one you can summon whereever your mouse point is...
Member since:2006-10-08
Fans: 6
"I'm not quite sure if I get your meaning on the menus without seeing that in action... But on the NeXT you can 'tear' off menus and make them stay on the screen if you are going to use their fucntions a lot. Comes in quite handy. you can also rearrange most of the screen."
As far as I know, Gtk has the "tear off" feature, too, very useful for Gimp.
WindowMaker allows having menus on the screen, they get an "X" button next to the title then. This gives a little feeling about how NeXT looks and feels like.
Member since:
2005-07-08
Fans: 3
I haven't seen much mention made of mouse buttons 1 v 2,3 or more and multiple clicks and mouse use v keyboard use.
As one get older, one's methods may need to change due to the sheer pain of RSI and even during an extended session the mode of use will change from use of multi clicks and mouse use to keyboard use and single clicks to avoid pain. It no longer makes sense to assume the fastest sequence will be the best way even over an hour since we need to change the exercise the digits get. Might as well build in 2 or more ways to do the same thing just to avoid RSI.
In some applications I may have to use 5-10 mouse actions over and over, if the developer had known what I was doing, they might have realized a single command to do the same thing would really speed things up. In the past I might have used a macro tool to repeat these sequences but they never really worked well.
Another issue that bothers me is the the notion of counting and sorting file names that have partial numerical values, say x1,x2,..x10,x11. It is pretty common to see files get sorted and placed in a folder in the wrong place because the sort is just plain wrong. Sorts done by character position of digits rather than by the digit decimal significance is just inexcusable.
The issue of menubar at top v each window pales in significance when these other factors are considered.
Member since:2005-07-06
Fans: 1
Agreed.
As with the other issue, I would add:
- consistency: for example tab handling is vastly different from one app to another which is annoying.
On Unix, command-line due to history bagage is too different from the GUI: rm should move a file into the trash as the GUI does, there should be a single shortcut sequence for Copy be it in the GUI or in the console, etc.
- Undo-ability: every action should be undoable, now it's nice that we can reopen a closed tab on many application, but we should be able to do the same thing for a closed window.
- Data persistence (related), as much data as possible should be kept all the time. So that in case of failure/error, it is possible to go back quickly to the previous state.
- Responsiveness: applications should be at least as responsive as BeOS apps were.
And let's kill the 'start-up screens', and use instead a start-up window instead which you can iconify, kill (if you started the wrong application by mistake), put behind other window..
Member since:2007-02-17
Fans: 1
Well first of all in *nix, each desktop environment handles its trash folder ina dif