If you die while surfing the net in Nepal, you can be buried in your own case. How cool is that?
Tooltips are good, but...
on
GNU Emacs 21
·
· Score: 2
I saw the screenshots a poster had linked to and I noticed the toolbar buttons weren't labelled. Tooltips are nice, but they are no substitute for labelling buttons. The label decreases access time with a mouse because it makes the toolbar a button bigger target and thus easier to hit via Fitt's law. Labelling the button also immediately tells the user what the button does, and they don't have to wait with the mouse hovering over the button for several seconds. That xemacs on my machine has labelled toolbar buttons and the one in the screenshot didn't is something I consider to be a step backwards. It's another case in the linux community where the "let's make it perty" crowd won out over the "let's make it usable" folks.
Emacs is a macho editor
on
GNU Emacs 21
·
· Score: 2
Real men speak with a lisp.
Are you being served + Omron Tech support
on
Robot Cat 'NeCoRo'
·
· Score: 2
"Hello, Omron customer support. How can I help you?"
"I went all the way to Tokyo and paid $1500 for some hot pussy and she gave out after the 5th day"
"Sir, put your hand inside the pussy and and press that little thing to see if it's turned on"
"Whoops, sorry. I didn't push hard enough. She's turned on now." Click.
I read this article yesterday and I totally agree. It's written by Jakob Nielsen, one of the most respected members of the interface design community. It basically sums up what the parent post said and adds some serious street cred.
My favorite quote from the article:
"For example they're so proud once they've ported [sic] PowerPoint. But that doesn't give us a new way of doing presentations."
P.S. mod parent up. He knows what he's talking about. Kinda rare in slashdot discussions about interface design
I agree that the aesthetic quality of OSX is superior to that of the GNOME/KDE environments. However, graphical environments are like love: "lookers" can be nice for a couple of weeks but if you have to live with them for any extended period of time, you'd rather have them treat you well than look pretty. Good design and usability engineering based on cognitive psychology is far more important for a graphical environment than aesthetics. Not that I'm criticizing OSX specifically, rather I'm criticizing aesthetics as main criteria for judging the usability of a graphical environment.
I agree with your comments about the installers. In fact, many of the installers are still very hard to use, but now confusing text-only parameters are replaced by confusing widget layouts. Virtually all the people who say that these installers are ready for prime time are the geeks and engineers who can use their prior linux expertise to get around the most confusing points of the linux graphical installers. I actually talked to one of the people who worked on the Red Hat installer and I mentioned some of the usability problems that made the installer difficult to use and difficult to navigate. He couldn't understand why these designs were problematic and thought that what I didn't like about the installer was that it "wasn't pretty enough". Which sort of goes back to issue 1.
Labels and tooltips are two different things. Labels are static text that always appears on or around a widget such as a button. Tooltips are text that pop up around the widget when the user keeps the pointer over the object for a second or two. The first interface device, the label, immediately adds greater clarity to the buttons function the second a user looks at the button and the label makes it faster to access because it makes the button bigger. The second UI device, the tooltip, is good when it is used to complement the label and used to give a more detailed description of the buttons function, but used in place of a label forces a user to wait for an unreasonably long amount of time to get the most rudimentry information about the object and it does not improve the access time of the control by making it larger.
As for the point about turning stuff on/off--The more steps you require users to go through to configure something and the more hierarchical menu levels they have to dig through and options they have to browse through the fewer the users will who actually change something, and the more geeky those users will tend to be. The most commonly used options for the greatest percentage of users will be the defaults. It makes more sense to make options which are most usable the default and let it to those who will have more patience and more enthusiam for customization (i.e. geeks) the option to search for how to choose the less usable options.
As for just about about everyone else who replied to my post by the nature of your responses you have proven my point in my first post far better than I ever could.When someone brings up usability problems in the world of Desktop Linux they are met with denial and flames. Bill Gates doesn't have to worry about trying to kill desktop linux; so many people with such attitudes are already doing his work for him.
Sure they've added rendering this, and office app that, but they still blow it on the most important usability stuff. They still don't label toolbar buttons. They still have the same ridiculously small buttons that have ridiculously slow access times and icons that are so small they don't mean a damn thing.
You try to tell these people about something like Fitts Law but they really don't want to hear that. I'm sure some people will argue with me about this. Some people, linux developers especially, do not want to pay attention to 20 years of HCI research. And its precisely these people who should be held up as the shining example of why linux in its current form will never reach the desktop.
For years, people in the mac gaming community have raved about Marathon, citing its entertaining plot and playability. And they were told by the PC gaming community to shut up about some game made in 1994 and to stop screaming about innovation and creativity and get a "real gaming machine" that would be taken seriously at a gaming party. Now all the same people who said this are deluged with an armful of FPS's that are basically doom with better graphics and its *them* who are complaining about a lack on innovation. The macintosh is the future of gaming--it's users expect things from their games today that PC users learn the hard way to expect 5-6 years from now.
If microsoft wants to avoid having employees come down with anthrax, they should try to get the terrorists to send the letters to departments at microsoft which don't exist and therefore the mail would be lost in the mail system and never delivered and opened. Names such as the "Microsoft Usability Testing Group" and the "Microsoft Security Testing Group" would be excellent choices.
I've heard that Nvidia cards do something very similar. I read awhile back on a games developer mailing list that NVidia's scheme has game video data compressed and sent in compressed form to the card (which does the decompression nice and fast in hardware). Supposedly this saves AGP bus bandwidth. I'm not sure if this is true, but if it is, it would explain why Nvidia refuses to open source their drivers for linux--they're trying to protect their compression algorithms needed to communicate with the card.
This feature is really cool. I ask my palm "would you add this phone number to my list of contacts?" and it asks me "would you like a piece of toast?"
Check out Humane Interface
on
Talking Palm
·
· Score: 2
If you want to understand what the guy you responded to is saying (which, generally speaking, is correct) you should check out Jef Raskin's Human Interface. He explains this problem of modes very well.
When a Jedi comes knocking door at my at 11AM to give me a pamphlet and ask me if I've ever used the force, I think I'm going to take them a lot more seriously. That lightsaber really stings!
Remember the java applet that emulated the PDP-something-or-other so you could play the original Spacewars? How cool would it be 20 years from now to have a Java that emulates a dreamcast machine? Processors at that point in time should just barely be able to run the thing JDK6.5 with decent framerates.
You know, if you overclock the all-metal PowerBook G4 to a fast enough speed, put a piece of cheese in between two pieces of bread, and slam it shut between the lid and the handrests for about 5 minutes, you could make a really awesome grilled cheese sandwich.
Trojan. Rockets are simply the best marketing vehicle for phallic shaped products. Now if you could just find a way to slide a 100ft rubber over the next cargo shipment to the ISS...
Slashdot Mirror
It's only a model.
If you die while surfing the net in Nepal, you can be buried in your own case. How cool is that?
I saw the screenshots a poster had linked to and I noticed the toolbar buttons weren't labelled. Tooltips are nice, but they are no substitute for labelling buttons. The label decreases access time with a mouse because it makes the toolbar a button bigger target and thus easier to hit via Fitt's law. Labelling the button also immediately tells the user what the button does, and they don't have to wait with the mouse hovering over the button for several seconds. That xemacs on my machine has labelled toolbar buttons and the one in the screenshot didn't is something I consider to be a step backwards. It's another case in the linux community where the "let's make it perty" crowd won out over the "let's make it usable" folks.
Real men speak with a lisp.
"Hello, Omron customer support. How can I help you?"
"I went all the way to Tokyo and paid $1500 for some hot pussy and she gave out after the 5th day"
"Sir, put your hand inside the pussy and and press that little thing to see if it's turned on"
"Whoops, sorry. I didn't push hard enough. She's turned on now." Click.
I'd get rid PalmOS too.
And I will counter your arguement.:)
Labels and tooltips are two different things. Labels are static text that always appears on or around a widget such as a button. Tooltips are text that pop up around the widget when the user keeps the pointer over the object for a second or two. The first interface device, the label, immediately adds greater clarity to the buttons function the second a user looks at the button and the label makes it faster to access because it makes the button bigger. The second UI device, the tooltip, is good when it is used to complement the label and used to give a more detailed description of the buttons function, but used in place of a label forces a user to wait for an unreasonably long amount of time to get the most rudimentry information about the object and it does not improve the access time of the control by making it larger.
As for the point about turning stuff on/off--The more steps you require users to go through to configure something and the more hierarchical menu levels they have to dig through and options they have to browse through the fewer the users will who actually change something, and the more geeky those users will tend to be. The most commonly used options for the greatest percentage of users will be the defaults. It makes more sense to make options which are most usable the default and let it to those who will have more patience and more enthusiam for customization (i.e. geeks) the option to search for how to choose the less usable options.
As for just about about everyone else who replied to my post by the nature of your responses you have proven my point in my first post far better than I ever could.When someone brings up usability problems in the world of Desktop Linux they are met with denial and flames. Bill Gates doesn't have to worry about trying to kill desktop linux; so many people with such attitudes are already doing his work for him.
Sure they've added rendering this, and office app that, but they still blow it on the most important usability stuff. They still don't label toolbar buttons. They still have the same ridiculously small buttons that have ridiculously slow access times and icons that are so small they don't mean a damn thing. You try to tell these people about something like Fitts Law but they really don't want to hear that. I'm sure some people will argue with me about this. Some people, linux developers especially, do not want to pay attention to 20 years of HCI research. And its precisely these people who should be held up as the shining example of why linux in its current form will never reach the desktop.
The desktop end-user *is* a real-time application
For years, people in the mac gaming community have raved about Marathon, citing its entertaining plot and playability. And they were told by the PC gaming community to shut up about some game made in 1994 and to stop screaming about innovation and creativity and get a "real gaming machine" that would be taken seriously at a gaming party. Now all the same people who said this are deluged with an armful of FPS's that are basically doom with better graphics and its *them* who are complaining about a lack on innovation. The macintosh is the future of gaming--it's users expect things from their games today that PC users learn the hard way to expect 5-6 years from now.
If microsoft wants to avoid having employees come down with anthrax, they should try to get the terrorists to send the letters to departments at microsoft which don't exist and therefore the mail would be lost in the mail system and never delivered and opened. Names such as the "Microsoft Usability Testing Group" and the "Microsoft Security Testing Group" would be excellent choices.
I've heard that Nvidia cards do something very similar. I read awhile back on a games developer mailing list that NVidia's scheme has game video data compressed and sent in compressed form to the card (which does the decompression nice and fast in hardware). Supposedly this saves AGP bus bandwidth. I'm not sure if this is true, but if it is, it would explain why Nvidia refuses to open source their drivers for linux--they're trying to protect their compression algorithms needed to communicate with the card.
This feature is really cool. I ask my palm "would you add this phone number to my list of contacts?" and it asks me "would you like a piece of toast?"
If you want to understand what the guy you responded to is saying (which, generally speaking, is correct) you should check out Jef Raskin's Human Interface . He explains this problem of modes very well.
I always wanted a watch that reset itself to January 1, 1970.
I can just imagine astronauts desperate for fun being forced to cannabalize the thing.
Pinky: Linus, what are we going to do tonight?
Linus: the same thing we do every night, Pinky, try to take over the world!
Almost everyone knows who Bert is. What have those Al-Qaeda folks been doing, living in a cave? On second thought...
When a Jedi comes knocking door at my at 11AM to give me a pamphlet and ask me if I've ever used the force, I think I'm going to take them a lot more seriously. That lightsaber really stings!
Remember the java applet that emulated the PDP-something-or-other so you could play the original Spacewars? How cool would it be 20 years from now to have a Java that emulates a dreamcast machine? Processors at that point in time should just barely be able to run the thing JDK6.5 with decent framerates.
You know, if you overclock the all-metal PowerBook G4 to a fast enough speed, put a piece of cheese in between two pieces of bread, and slam it shut between the lid and the handrests for about 5 minutes, you could make a really awesome grilled cheese sandwich.
Trojan. Rockets are simply the best marketing vehicle for phallic shaped products. Now if you could just find a way to slide a 100ft rubber over the next cargo shipment to the ISS...
Unix + "Crisp, sharp, graphics" = Crispix.