I am currently working on a fork of GNOME that replaces the broken UI concepts from Windows with sound UI concepts from the macintosh (e.g. alt command keys instead of ctrl, global menubar, obeyance of fitts law, etc). So far, I've gotten it to the point where recompiling code that uses gnomeui macros produces mac shortcuts, gets rid of those underline accelerators that clutter menus (yes, I'll come up with a better non-mouse solution), and puts the cancel button in dialogs on the left and ok on the right, as is consistent with the western concept of negatives being on the left and positive being on the right. Cancel and OK should be replaced with more descriptive terms, but first things first. As for the problems with consistency between applications, I'm also working on "porting" (for lack of a better term) the code of many GNOME programs to a usable and consistent state. All menus, keyboard shortcuts, dialogs, for similar features will be consistent across apps. I'm screwing with people's code because they didn't screw with it enough.
Why am I doing this? Isn't forking counterproductive? Of course it is. But I am very greatly disturbed by Miguel et. al duplicating many of Microsoft's UI mistakes. When Microsoft designed the windows UI, they made a lot of decisions that were based on being different from(and in some cases, the exact of opposite) apple. The problem with this was that Apple put many years of HCI research into the design of the MacOS UI. By doing the exact opposite of apple of what apple did, Microsoft was going directly against interface designs that were scientifically proven in usability labs to be more effective, efficient, and intuitive. For example, it has been proven that an application menubar at the top edge of the screen can be accessed much faster than a menubar on the window of each application. This is not UI dogma or personal opinion, it's proven fact. But microsoft didn't care about end user's experience, they cared about their legal status. So they ended up throwing efficiency and usability out the door and doing the opposite of what was proven to work. The list goes on an on, ad nauseum. And when GNOME blindly copies the Microsoft (Where do you think the "Exit" menu item came from? The 'Q' in ctrl+q stands for something), they are perpetuating UI mistakes that need to be put to rest. GNOME should be about creating and designing new and improved user interfaces, not perpetuating bad decisions made in the past. I won't be party to putting users through another 10 years of UI misery to keep backwards compatibility with a backwards design.
I'm not saying that my ultimate goal is blindly copying MacOS. The mac interface certainly can be greatly improved upon. But I believe that when a UI builds on someone else's design, that someone else should be someone who knew what the hell he was doing. Microsoft does not fit that description, and Apple fits it better than anyone else who has yet come along.
I know that talk is cheap until you back it up with code, so no one will probably take me seriously. But it won't be too far into the future that I'll post a devel version of the modified gnome-libs and gnome-core on freshmeat. This UI insanity has to be stopped.
In five seconds the PenguinCow revolution will begin...
Many of you might not like Raskin's idea's. But any of you have been using such a beast for years: it's called palm OS. Many of the principles that Raskin describes in his book "The Humane Interface" already exist in Palm OS, any a lot of the guys posting negatives about Jef's remarks certainly have no problem using their Palm. You press the power button, it instantly turns on. You press the phone button, you instantly get addresses. Press the memo button, you instantly get memo's. The storage of the user's data is transparent, so the only navigation a user really has to do is tap an application icon. Okay, maybe there's a few drawbacks to Palm OS (like being limited to 2048 bytes of stack space), but the Palm OS interface allows it's users to use applications in a highly efficient and intuitive manner. Jef sounds crazy when he talks about his humane interface, but then again, people thought he was crazy 20 years ago when we he started the macintosh project.
Yes, I know this is off-topic but you seem to be the type of person I want to pose this question to.
I've been trying to gauge the reaction of fellow mac users to the idea of creating a mac-like fork of GNOME, with all of the mac keyboard shortcuts, menu selections, dialog buttons where they should be, global menubar, etc. A lot of mac users are going to find to GNOME no better than the wintel UI ('cause it basically *is* the Wintel UI), and I think the computing public should not be exposed to another ten years of M$ GUI design mistakes, even if the people repeating them (GNOME) does really have the best of intentions.
If I developed this beast, would you (or anyone else reading this post) be inclined to choose it over the existing GNOME?
Yes, using c for an object oriented coding is kind of f*cked up. That's why (at least I think that's why) the GNOME dudes are developing Inti, a c++ application development platform that has a gtk-based GUI toolkit that's supposed to make writing consistent GUI apps pretty easy. I suspect that the moment it gets real stable, it'll supplant a lot of the Gtk C coding being done right now.
Why choose between a mac interface and linux when you can have the best of both worlds? I'm currently in the process of creating a GNOME fork that is based on the Mac OS UI (as opposed to the current GNOME UI layout that is copied from Microsoft). Due to the open source nature of the GNOME project (thanx for the code, helix, er, ximian) I am able to quickly modify Gtk/GNOME apis so that they will produce mac-like interfaces. And when I recompile code that uses gnome-app-helper macros, voila! Instant mac interface. Cancel/No buttons will be on the left where they belong, as will OK/Yes buttons on the right. The "Exit" menu item will now called "Quit". Simple keyboard shortcuts will use the modifier key by the spacebar (alt), instead of the one in outer mongolia (ctrl). The menubar will be at the top of the screen where it belongs. And list goes on and on. So many GNOME apps on Freshmeat, so little time!
It gets even better. GNUstep is, of course, open source. And what does GNUstep have? Bundles! (just like MacOS). Once gnustep's bundle code is added to the mixture, the stupid hassle of dealing with packaging systems will be less stupid and more robust. Goodbye annoying RPM/deb messages and corrupted binary databases, hello fast, easy installs!
Code is power.
Macs have had this for years
on
MySQL FS
·
· Score: 1
It's called the desktop database. It allows one to do cool things such as change the icon for a single program document (no messy MIME/extensions), store comments on files, and all other sorts of GUI goodness. And it makes searches really fast. Of course, it's probably a little crufty (being over 16 years old. I think it's a flat database), but I think it's definately time linux had a feature such as this.
Several linux distributions (such as Corel, Redmond Linux, MaxOS, Caldera,just to name a few) try to keep their UI's consistent with that of Windows. Even much of GNOME follows much of the windows pattern (same keyboard accelerators and windows, same menu item labels, etc) for the sake of easy transition from Windows. As someone who went from MacOS to Linux, windows UI layouts such as the Ok Button being on the left and selecting "Exit" to quit applications just doesn't quite feel right. Has LinuxPPC ever considered modifying GNOME or KDE to better ease the transition for mac users? Has there been any talk of releasing some sort of "Cupertino Linux"?
Miguel I:"You could say that Evolution is targeted to replace Outlook, but Evolution is just part of the puzzle to provide a complete solution from people migrating from the Windows platform."
If Miguel's goal is to make Evolution very "windowsy", is Miguel going to add in Gnome Basic VBA scripting abilities? It will be very interesting to see just how far Helix takes the whole "let's copy windows" thing.
(You should definately read Jef Raskin's "The Humane Interface". He expands on many of the concepts you mention.)
The add/remove thing in windows is pretty stupid. In the current GUI paradigm, you add something to a to somewhere by dragging it and dropping it, but Microsoft forces the user to go through "Add new software" route, thus adding needless modes and complexity to the process. In windows, there are ordinary folders for documents, and then there are special folders like "My Documents", "Dial-up networking", "Printers", and other such folders with 'special' behaviors that break consistency with the behavior of regular documents and folders. Breaking consistency in a GUI system is A Bad Thing(tm). Microsoft never designed windows to be easy, they designed it to be different than MacOS.
For all of MacOS' technical faults, installation of programs was consistent with the drag and drop nature of the UI. In those heady days of System 7, you installed a program by dragging the folder it was inside from the CD-ROM or floppy and dropping it to wherever the hell you wanted to put it. Everything was either a folder or a file that went into a folder. No folders (in most cases) had any "special" behaviors that were inconsistent with the behavior of your run-of-the-mill user-created folder. MacOS was also built like a tank. Any program could run without the need for a config file. Preference files were seperate files (what I call "protected configuration"), not one big, easily corruptable binary database. A config file got busted, you trashed it and it was rebuilt the next time you ran the program (what I call "regenerative configuration"). Enough metadata was kept in the file system (though the dual fork system) to rebuild the file typing database system if it got munged. It's a system like this that users need. Too many people do too little with their computers because they are so afraid of permanently screwing them up, and we have the folks in Redmond to thank for this. In end-user land, it's not the crash that kills, it's the permanent screw up.
I am seriously considering ripping off GNUstep's file bundle code and putting that into some form of Linux distribution. The biggest technical problem that desktop linux faces right now is finding a good way to describe what consistutes an application and the applets, documents, libraries, and other stuff that go along with it. People have tried with RPM and dpkg, but these systems make an OS installation damn easy to destroy, and spread the application installation so far apart it becomes unmanagable. Putting an application installation into a single folder and GUI-wise pretending that it's a single executable is the best way that anyone has ever done it. A HURD bundle service would kick ass (If HURD could just add hardware support for, well, just about everything I'd be set).
Tom Piwowar of the "Computer Guys" show puts this in perspective:
"There is a very important function for linux. Linux dispells ignorance for windows users. All these years, windows users have been mystified at the reaction they get from Mac users when they come over and talk about windows. The mac user looks at you like you just pooped on the floor. Well now, what windows users need to do is go overand look at linux X-Windows GUI, and they've got office applications that run under that, and they need to use it for a while and see how they feel. And what they're going to feel like is that this is a very impoverished user interface, that everything is much slower, that there are more steps, and that it's much more complicated. They're going to feel just like a Macintosh user feels when they see windows!. So, all you windows users, go out there and spend some time with Linux, and then you'll understand what we're talking about"
Yes, I know that sticky keys are a common outcome of typing one-handed (someone would make the joke sooner or later, might as well get it over with).
But seriously--Can the half-keyboard do "sticky keys" (i.e. the ability to hit a modifier button like shift, alt, etc just once, depress the key, and then be able to another key to get a single modified character) in hardware? Or does it rely on the good graces of whatever OS it's running under to support support sticky keys? If someone needs to operate the half-keyboard with just one finger and they had no OS sticky key support, this would be an important feature.
I can buy liquor. I could consume it responsibly, or I might drive drunk and kill somebody, But society recognizes my free will as an individual and lets me take whatever action I deem acceptable, along with the consequences that follow that action. I will go to jail if I drive drunk, but I am allowed to drive drunk if I choose.
I can buy a gun. I could use it to hunt and feed my family or I could use it go on a shooting spree, but society recognizes my free will as an individual and lets me take whatever acton I deem acceptable, along with the consequences that follow that action. I will go to jail and possibly get executed if I shoot somebody, but I am allowed to go on a shooting spree if I choose.
I can't buy a DVD and use DeCSS to decrypt it.
I could play my decrypted copy legally or I could become a pirate and start selling it, but society does not recognize my free will as an individual, and prevents me from taking action they deem unacceptable, Whether I choose to take that action and the consequences that go with it is not a choice I am allowed to make. I will not be allowed to decrypt the DVD.
One day, I won't be able to a book written by written by Karl Marx and read it. I could read it and study ideas of alternative economies, or I could read it an decide to overthrow the government. But society will not recognize my free will as an individual, and will prevent me from taking action they deem unacceptable. Whether I choose to take that action and the consequences that go with it is not a choice I am allowed to make. I will not be able to buy the book written by Karl Marx.
Most problems most people find with today's GUI's are problems because most computers today running GUI's are running windows.
The problems you describe with the icons--Macs don't have those problems. Apple didn't have any dumb DOS filename extensions they could use as a lame argument for not making the icons recognizable, so the file icon *had* to convey what type of file the user was looking at. All you have to do to undertstand what type of file you are dealing with on a mac is LATFI (Look At The Fine Icon). Getting to the argument about the executables and files having the same icons, in most cases, macs have different icons for files and executables. Getting to the whole application integration thing, macs have always been able to do this. A mac file has two basic properties, a creator type and a file type. The creator type says what program the file "belongs" to, and the file type is the file type (jpeg, mp3, etc). When you double click on a file icon (let's say an MP3), the mac will open up the file in MacAmp. Double click a word document, the document
Most of the things you describe in your post are things that macs have been able to do for years. The whole application integration thing has been tackled by applescript, a plain english type of scripting language that has hooks in a great many mac programs. The file/folder system is not a bad system, but it needs to be implemented consistently, like it is on a mac. On a mac, every object is consistently manipulatable. I can choose a different icon for just *one* file of a certain type. If I have a britney spears MP3, I can change the icon from whatever the default mp3 icon is to a pile of dog poo (seems appropriate). I can have the ability to change the icon for a single folder, as well. Every object on a mac (with the exception of trash) is a folderitem--it is either a folder or something that fits in a folder. It can be easily modified, changed, deleted, or moved. Contrast this with windows, where there are regular files/folders, and then there are files/folders that have special "behaviors". Files/Folders like My Computer, Dial Up Networking, Control Panels, Printers, My Documents, etc. These files/folders don't act normally, so they break consistency. Any GUI that breaks consistency with itself is going to be user hostile.
I fail to see why Miguel is so damned impressed with Microsoft.
>BUT I will keep my faith. People have been >telling me my Macs (I now have 3 powerbooks, 6 >Macs and the ][2) have no future for 15 years. >This is FUD.
Keep your macs. They not only have the future, they are the future. Almost every step of the way, the industry has dragged their heels over implementing lots of new stuff, stuff that would make computers easier and better to use, because they wanted to be conservative and stick with old legacy stuff. The industry didn't kick their mainframe-only addiction till apple pushed the PC. The industry didn't take one step to implement GUI's until after apple did it. The industry didn't make laptops wide enough to rest your hands on until apple did it (with the PowerBook). The industry hardly supported usb and firewire and stuck with serial and parallel--it wasn't until apple burned the bridge and went totally usb/1394 that the suddenly there were all these usb and firewire devices. The industry (with the exceptional oddball like acer and their black machines) didn't dream of making multi-colored computers until apple did it with the imac. In some odd, symbiotic relationship, apple advances the the computer industry in bold, daring moves that the rest of the computer industry is too scared to make, and the industry makes just enough hardware for Apple that Steve Jobs et al can keep the lights on and can plan up the next daring move.
those macintosh people were making usable graphical user interfaces. These programmers, along with the rest of the macintosh community, endured cries of 'WIMP' and 'macintoy' and 'idiot box' and lots of other anti-GUI sentiment. And then those hypocrits from old school unix and DOS turned right around and created X and Windows and, in their arrogance and spite for the macintosh, never tried to learn anything from it. They ignored many intelligent, widely appliable usability principles the mac introduced, or did did the exact opposite of those principles, just so they could be different from apple. Tests showed a menubar at the top of the screen can be accessed faster than one on a window, but it didn't matter to Microsoft. Having the cancel button on the left and OK button on the right conforms better to the left-right nature of English than the OK button left/cancel button right we see in windows (and GNOME) dialogs, but that didn't matter to Microsoft, either. And so windows ended up being a legacy to UI stupidity, and GNOME, through their blind emulation of microsoft, ends up being stupid legacy UI. A lot of gnome people (while well meaning) are a bunch of ex-windows people who conveniantly forget about history. Makes you wonder who really deserves to lead to the linux desktop GUI revolution: the people people who led the first GUI revolution, or the people who fought against it.
I feel that one of the strongest points of Darwin that everyone is really overlooking is the Bundle system, a directory that groups together all the associated files of an application (and which abstracts out this directory *as* the application). This is a system far superior to the way that windows or any linux/RPM deals with the question of "what constitutes a packaged application?". If you have a centralized database such into which app information is installed (such as the RPM database or windows registry) and there is no metadata or anything else from which a new database can be rebuilt, you end up with a techsupport nightmare. The centralized database could (and usually does) get hosed and (re)(de)installing an application is difficult if not impossible. The bundle system presents a far more robust solution, since all files associated with an app are kept together in a directory and not just in a single, fragile, non-rebuildable database. The bundle system could dramatically reduce the TCO that windows incurs through the registry (probably at the cost I/O efficiency), something the corporate world would find attractive (if the dumb bastards actually looked at TCO when making purchasing decisions). Linux should scrap RPM/Deb altogether and simply go with Bundles.
Blame the programmers at Microsoft and in the Windows development community who have such little GUI programming ability and such poor skill at designing usable, user friendly software that they turn the simple act of writing an electronic letter into a shuttle launch. The limited function of internet appliances doesn't make things easier, but rather, give the programmers less opportunity to screw up the interface. If cheap PC's had interfaces written by people who knew what the hell they were doing, I-openers would be unnecessary and the lady in quoted in your would be able to do everything with an E-machine that she did with an I-opener.
And yes, I am programmer, but I actually pay attention to usability issues when I write software.
I did that in grade school--hokey pokey
on
Keyless Keyboard
·
· Score: 1
"You put your right arm, you take your right arm out, you take your right arm in, and you shake it all about, you hit the ho-Key po-Key and you pipe to stdout, that's what it's all about"
Finally, a kindred spirit! I totally agree that the best technology is not always the one who wins (apple vs. M$, BetaMax vs VHS, etc). Magneto optical is truly the best way to go. The 640 MB 3.5 disk holds more than 6 times that of a zip disk and is far more reliable. MO is also far faster than a CD-ROM (esp. CD-RW). And it doesn't have to rewrite the whole damn disk to do updates (again, like CD-RW). I think that one of the best kept secrets of ebay are those cheap 128 MB dynamo drives. For 30 bucks (if you have a fast scsi card) you can get the equivalent storage of a zip drive with greater reliability for a third of the cost.
Slashdot Mirror
I am currently working on a fork of GNOME that replaces the broken UI concepts from Windows with sound UI concepts from the macintosh (e.g. alt command keys instead of ctrl, global menubar, obeyance of fitts law, etc). So far, I've gotten it to the point where recompiling code that uses gnomeui macros produces mac shortcuts, gets rid of those underline accelerators that clutter menus (yes, I'll come up with a better non-mouse solution), and puts the cancel button in dialogs on the left and ok on the right, as is consistent with the western concept of negatives being on the left and positive being on the right. Cancel and OK should be replaced with more descriptive terms, but first things first. As for the problems with consistency between applications, I'm also working on "porting" (for lack of a better term) the code of many GNOME programs to a usable and consistent state. All menus, keyboard shortcuts, dialogs, for similar features will be consistent across apps. I'm screwing with people's code because they didn't screw with it enough.
Why am I doing this? Isn't forking counterproductive? Of course it is. But I am very greatly disturbed by Miguel et. al duplicating many of Microsoft's UI mistakes. When Microsoft designed the windows UI, they made a lot of decisions that were based on being different from(and in some cases, the exact of opposite) apple. The problem with this was that Apple put many years of HCI research into the design of the MacOS UI. By doing the exact opposite of apple of what apple did, Microsoft was going directly against interface designs that were scientifically proven in usability labs to be more effective, efficient, and intuitive. For example, it has been proven that an application menubar at the top edge of the screen can be accessed much faster than a menubar on the window of each application. This is not UI dogma or personal opinion, it's proven fact. But microsoft didn't care about end user's experience, they cared about their legal status. So they ended up throwing efficiency and usability out the door and doing the opposite of what was proven to work. The list goes on an on, ad nauseum. And when GNOME blindly copies the Microsoft (Where do you think the "Exit" menu item came from? The 'Q' in ctrl+q stands for something), they are perpetuating UI mistakes that need to be put to rest. GNOME should be about creating and designing new and improved user interfaces, not perpetuating bad decisions made in the past. I won't be party to putting users through another 10 years of UI misery to keep backwards compatibility with a backwards design.
I'm not saying that my ultimate goal is blindly copying MacOS. The mac interface certainly can be greatly improved upon. But I believe that when a UI builds on someone else's design, that someone else should be someone who knew what the hell he was doing. Microsoft does not fit that description, and Apple fits it better than anyone else who has yet come along.
I know that talk is cheap until you back it up with code, so no one will probably take me seriously. But it won't be too far into the future that I'll post a devel version of the modified gnome-libs and gnome-core on freshmeat. This UI insanity has to be stopped.
In five seconds the PenguinCow revolution will begin...
Many of you might not like Raskin's idea's. But any of you have been using such a beast for years: it's called palm OS. Many of the principles that Raskin describes in his book "The Humane Interface" already exist in Palm OS, any a lot of the guys posting negatives about Jef's remarks certainly have no problem using their Palm. You press the power button, it instantly turns on. You press the phone button, you instantly get addresses. Press the memo button, you instantly get memo's. The storage of the user's data is transparent, so the only navigation a user really has to do is tap an application icon. Okay, maybe there's a few drawbacks to Palm OS (like being limited to 2048 bytes of stack space), but the Palm OS interface allows it's users to use applications in a highly efficient and intuitive manner. Jef sounds crazy when he talks about his humane interface, but then again, people thought he was crazy 20 years ago when we he started the macintosh project.
While Mandrake is pretty pro-kde (not that there's anything wrong with that)the installer and config tools are written in Perl with Gtk bindings.
I've been trying to gauge the reaction of fellow mac users to the idea of creating a mac-like fork of GNOME, with all of the mac keyboard shortcuts, menu selections, dialog buttons where they should be, global menubar, etc. A lot of mac users are going to find to GNOME no better than the wintel UI ('cause it basically *is* the Wintel UI), and I think the computing public should not be exposed to another ten years of M$ GUI design mistakes, even if the people repeating them (GNOME) does really have the best of intentions.
If I developed this beast, would you (or anyone else reading this post) be inclined to choose it over the existing GNOME?
Yes, using c for an object oriented coding is kind of f*cked up. That's why (at least I think that's why) the GNOME dudes are developing Inti, a c++ application development platform that has a gtk-based GUI toolkit that's supposed to make writing consistent GUI apps pretty easy. I suspect that the moment it gets real stable, it'll supplant a lot of the Gtk C coding being done right now.
Don't f*ck with Pillsbury. They're known for sending 150 ft. demonically possessed doughboys at anyone who opposes them.
(Damn the karma and full speed ahead!)
Why choose between a mac interface and linux when you can have the best of both worlds? I'm currently in the process of creating a GNOME fork that is based on the Mac OS UI (as opposed to the current GNOME UI layout that is copied from Microsoft). Due to the open source nature of the GNOME project (thanx for the code, helix, er, ximian) I am able to quickly modify Gtk/GNOME apis so that they will produce mac-like interfaces. And when I recompile code that uses gnome-app-helper macros, voila! Instant mac interface. Cancel/No buttons will be on the left where they belong, as will OK/Yes buttons on the right. The "Exit" menu item will now called "Quit". Simple keyboard shortcuts will use the modifier key by the spacebar (alt), instead of the one in outer mongolia (ctrl). The menubar will be at the top of the screen where it belongs. And list goes on and on. So many GNOME apps on Freshmeat, so little time!
It gets even better. GNUstep is, of course, open source. And what does GNUstep have? Bundles! (just like MacOS). Once gnustep's bundle code is added to the mixture, the stupid hassle of dealing with packaging systems will be less stupid and more robust. Goodbye annoying RPM/deb messages and corrupted binary databases, hello fast, easy installs!
Code is power.
It's called the desktop database. It allows one to do cool things such as change the icon for a single program document (no messy MIME/extensions), store comments on files, and all other sorts of GUI goodness. And it makes searches really fast. Of course, it's probably a little crufty (being over 16 years old. I think it's a flat database), but I think it's definately time linux had a feature such as this.
Now if we could only put the stuff into a Delorean...
Several linux distributions (such as Corel, Redmond Linux, MaxOS, Caldera,just to name a few) try to keep their UI's consistent with that of Windows. Even much of GNOME follows much of the windows pattern (same keyboard accelerators and windows, same menu item labels, etc) for the sake of easy transition from Windows. As someone who went from MacOS to Linux, windows UI layouts such as the Ok Button being on the left and selecting "Exit" to quit applications just doesn't quite feel right. Has LinuxPPC ever considered modifying GNOME or KDE to better ease the transition for mac users? Has there been any talk of releasing some sort of "Cupertino Linux"?
If Miguel's goal is to make Evolution very "windowsy", is Miguel going to add in Gnome Basic VBA scripting abilities? It will be very interesting to see just how far Helix takes the whole "let's copy windows" thing.
(You should definately read Jef Raskin's "The Humane Interface". He expands on many of the concepts you mention.)
The add/remove thing in windows is pretty stupid. In the current GUI paradigm, you add something to a to somewhere by dragging it and dropping it, but Microsoft forces the user to go through "Add new software" route, thus adding needless modes and complexity to the process. In windows, there are ordinary folders for documents, and then there are special folders like "My Documents", "Dial-up networking", "Printers", and other such folders with 'special' behaviors that break consistency with the behavior of regular documents and folders. Breaking consistency in a GUI system is A Bad Thing(tm). Microsoft never designed windows to be easy, they designed it to be different than MacOS.
For all of MacOS' technical faults, installation of programs was consistent with the drag and drop nature of the UI. In those heady days of System 7, you installed a program by dragging the folder it was inside from the CD-ROM or floppy and dropping it to wherever the hell you wanted to put it. Everything was either a folder or a file that went into a folder. No folders (in most cases) had any "special" behaviors that were inconsistent with the behavior of your run-of-the-mill user-created folder. MacOS was also built like a tank. Any program could run without the need for a config file. Preference files were seperate files (what I call "protected configuration"), not one big, easily corruptable binary database. A config file got busted, you trashed it and it was rebuilt the next time you ran the program (what I call "regenerative configuration"). Enough metadata was kept in the file system (though the dual fork system) to rebuild the file typing database system if it got munged. It's a system like this that users need. Too many people do too little with their computers because they are so afraid of permanently screwing them up, and we have the folks in Redmond to thank for this. In end-user land, it's not the crash that kills, it's the permanent screw up.
I am seriously considering ripping off GNUstep's file bundle code and putting that into some form of Linux distribution. The biggest technical problem that desktop linux faces right now is finding a good way to describe what consistutes an application and the applets, documents, libraries, and other stuff that go along with it. People have tried with RPM and dpkg, but these systems make an OS installation damn easy to destroy, and spread the application installation so far apart it becomes unmanagable. Putting an application installation into a single folder and GUI-wise pretending that it's a single executable is the best way that anyone has ever done it. A HURD bundle service would kick ass (If HURD could just add hardware support for, well, just about everything I'd be set).
Keep up the good work with FLTKTom Piwowar of the "Computer Guys" show puts this in perspective:
Yes, I know that sticky keys are a common outcome of typing one-handed (someone would make the joke sooner or later, might as well get it over with).
But seriously--Can the half-keyboard do "sticky keys" (i.e. the ability to hit a modifier button like shift, alt, etc just once, depress the key, and then be able to another key to get a single modified character) in hardware? Or does it rely on the good graces of whatever OS it's running under to support support sticky keys? If someone needs to operate the half-keyboard with just one finger and they had no OS sticky key support, this would be an important feature.
I can buy liquor. I could consume it responsibly, or I might drive drunk and kill somebody, But society recognizes my free will as an individual and lets me take whatever action I deem acceptable, along with the consequences that follow that action. I will go to jail if I drive drunk, but I am allowed to drive drunk if I choose. I can buy a gun. I could use it to hunt and feed my family or I could use it go on a shooting spree, but society recognizes my free will as an individual and lets me take whatever acton I deem acceptable, along with the consequences that follow that action. I will go to jail and possibly get executed if I shoot somebody, but I am allowed to go on a shooting spree if I choose. I can't buy a DVD and use DeCSS to decrypt it. I could play my decrypted copy legally or I could become a pirate and start selling it, but society does not recognize my free will as an individual, and prevents me from taking action they deem unacceptable, Whether I choose to take that action and the consequences that go with it is not a choice I am allowed to make. I will not be allowed to decrypt the DVD. One day, I won't be able to a book written by written by Karl Marx and read it. I could read it and study ideas of alternative economies, or I could read it an decide to overthrow the government. But society will not recognize my free will as an individual, and will prevent me from taking action they deem unacceptable. Whether I choose to take that action and the consequences that go with it is not a choice I am allowed to make. I will not be able to buy the book written by Karl Marx.
I think a $1000 toilet seat with buttons and USB support would kick ass.
Most problems most people find with today's GUI's are problems because most computers today running GUI's are running windows.
The problems you describe with the icons--Macs don't have those problems. Apple didn't have any dumb DOS filename extensions they could use as a lame argument for not making the icons recognizable, so the file icon *had* to convey what type of file the user was looking at. All you have to do to undertstand what type of file you are dealing with on a mac is LATFI (Look At The Fine Icon). Getting to the argument about the executables and files having the same icons, in most cases, macs have different icons for files and executables. Getting to the whole application integration thing, macs have always been able to do this. A mac file has two basic properties, a creator type and a file type. The creator type says what program the file "belongs" to, and the file type is the file type (jpeg, mp3, etc). When you double click on a file icon (let's say an MP3), the mac will open up the file in MacAmp. Double click a word document, the document
Most of the things you describe in your post are things that macs have been able to do for years. The whole application integration thing has been tackled by applescript, a plain english type of scripting language that has hooks in a great many mac programs. The file/folder system is not a bad system, but it needs to be implemented consistently, like it is on a mac. On a mac, every object is consistently manipulatable. I can choose a different icon for just *one* file of a certain type. If I have a britney spears MP3, I can change the icon from whatever the default mp3 icon is to a pile of dog poo (seems appropriate). I can have the ability to change the icon for a single folder, as well. Every object on a mac (with the exception of trash) is a folderitem--it is either a folder or something that fits in a folder. It can be easily modified, changed, deleted, or moved. Contrast this with windows, where there are regular files/folders, and then there are files/folders that have special "behaviors". Files/Folders like My Computer, Dial Up Networking, Control Panels, Printers, My Documents, etc. These files/folders don't act normally, so they break consistency. Any GUI that breaks consistency with itself is going to be user hostile.
I fail to see why Miguel is so damned impressed with Microsoft.
C:\ONGRTLNS.W2K
>BUT I will keep my faith. People have been >telling me my Macs (I now have 3 powerbooks, 6 >Macs and the ][2) have no future for 15 years. >This is FUD. Keep your macs. They not only have the future, they are the future. Almost every step of the way, the industry has dragged their heels over implementing lots of new stuff, stuff that would make computers easier and better to use, because they wanted to be conservative and stick with old legacy stuff. The industry didn't kick their mainframe-only addiction till apple pushed the PC. The industry didn't take one step to implement GUI's until after apple did it. The industry didn't make laptops wide enough to rest your hands on until apple did it (with the PowerBook). The industry hardly supported usb and firewire and stuck with serial and parallel--it wasn't until apple burned the bridge and went totally usb/1394 that the suddenly there were all these usb and firewire devices. The industry (with the exceptional oddball like acer and their black machines) didn't dream of making multi-colored computers until apple did it with the imac. In some odd, symbiotic relationship, apple advances the the computer industry in bold, daring moves that the rest of the computer industry is too scared to make, and the industry makes just enough hardware for Apple that Steve Jobs et al can keep the lights on and can plan up the next daring move.
those macintosh people were making usable graphical user interfaces. These programmers, along with the rest of the macintosh community, endured cries of 'WIMP' and 'macintoy' and 'idiot box' and lots of other anti-GUI sentiment. And then those hypocrits from old school unix and DOS turned right around and created X and Windows and, in their arrogance and spite for the macintosh, never tried to learn anything from it. They ignored many intelligent, widely appliable usability principles the mac introduced, or did did the exact opposite of those principles, just so they could be different from apple. Tests showed a menubar at the top of the screen can be accessed faster than one on a window, but it didn't matter to Microsoft. Having the cancel button on the left and OK button on the right conforms better to the left-right nature of English than the OK button left/cancel button right we see in windows (and GNOME) dialogs, but that didn't matter to Microsoft, either. And so windows ended up being a legacy to UI stupidity, and GNOME, through their blind emulation of microsoft, ends up being stupid legacy UI. A lot of gnome people (while well meaning) are a bunch of ex-windows people who conveniantly forget about history. Makes you wonder who really deserves to lead to the linux desktop GUI revolution: the people people who led the first GUI revolution, or the people who fought against it.
I feel that one of the strongest points of Darwin that everyone is really overlooking is the Bundle system, a directory that groups together all the associated files of an application (and which abstracts out this directory *as* the application). This is a system far superior to the way that windows or any linux/RPM deals with the question of "what constitutes a packaged application?". If you have a centralized database such into which app information is installed (such as the RPM database or windows registry) and there is no metadata or anything else from which a new database can be rebuilt, you end up with a techsupport nightmare. The centralized database could (and usually does) get hosed and (re)(de)installing an application is difficult if not impossible. The bundle system presents a far more robust solution, since all files associated with an app are kept together in a directory and not just in a single, fragile, non-rebuildable database. The bundle system could dramatically reduce the TCO that windows incurs through the registry (probably at the cost I/O efficiency), something the corporate world would find attractive (if the dumb bastards actually looked at TCO when making purchasing decisions). Linux should scrap RPM/Deb altogether and simply go with Bundles.
Blame the programmers at Microsoft and in the Windows development community who have such little GUI programming ability and such poor skill at designing usable, user friendly software that they turn the simple act of writing an electronic letter into a shuttle launch. The limited function of internet appliances doesn't make things easier, but rather, give the programmers less opportunity to screw up the interface. If cheap PC's had interfaces written by people who knew what the hell they were doing, I-openers would be unnecessary and the lady in quoted in your would be able to do everything with an E-machine that she did with an I-opener. And yes, I am programmer, but I actually pay attention to usability issues when I write software.
"You put your right arm, you take your right arm out, you take your right arm in, and you shake it all about, you hit the ho-Key po-Key and you pipe to stdout, that's what it's all about"
Finally, a kindred spirit! I totally agree that the best technology is not always the one who wins (apple vs. M$, BetaMax vs VHS, etc). Magneto optical is truly the best way to go. The 640 MB 3.5 disk holds more than 6 times that of a zip disk and is far more reliable. MO is also far faster than a CD-ROM (esp. CD-RW). And it doesn't have to rewrite the whole damn disk to do updates (again, like CD-RW). I think that one of the best kept secrets of ebay are those cheap 128 MB dynamo drives. For 30 bucks (if you have a fast scsi card) you can get the equivalent storage of a zip drive with greater reliability for a third of the cost.
P.S. MO now also has a firewire interface as well
so we can make it suck less.
the paperclip cannot be far behind