![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
symbioidljIt's long, but I'd like to hear some feedback from my techy type friends on this issue. I think you all have some good insight, and can set me on some right tracks on these issue. Thanks...
--------
I was thinking today about tabbed interfaces, and about my post to bugzilla.
one of the issues they were having was conflict with various OS keymapping(not the technically correct term for what I'm thinking of, really) I never thought that much about how cross platform design entails so much complication in terms of key-combos/interface interaction.
Of course, each OS/environment has it's own unique interface standards(HIGs as they're called -- Human Interface Guidelines). I must say I'm also using the term OS incorrectly here, as well. I should say "GUI" or "Window Manager"
Anyways, I had already been thinking about an OS that uses a tabbed interface as it's main way of working with the system. Something similar to the old mac systems with one bar on the top. I never got that, because I was so used to working with windows, but really it's an elegant design(does OSX have that same setup? I've not really looked).
I also have always thought that a windowed environment for a system that has strong CLI capabilities should have some sort of simple terminal always accessible.
I imagine a small bar across the bottom of the screen where most modern systems have ripped off the Win95 "start" bar.... Just one line, that could be expanded to a full terminal with a keystroke or two. One of the biggest problems is that most modern keyboards have a windows key on them, but I can envision that as a sort of key to enable a quickjump from a window to the CLI. I think it would be good if, perhaps, the scancode of that key is defined through ISO or whatever org designs such things. It shouldn't be tied to windows, and it would be good if it was given a name. Maybe some standards body has already defined it well enough. It just needs a good, universal name if it doesn't have one.
I know there's the "meta" in linux/unix, and "option" in macs... usually, i think, however, meta is considered the "ctrl" key. I could be wrong. Whatever the "windows" key is called or given as a picture(I think I've seen some linux keyboards with Tux the Penguin on it instead of a window)
Anyways, this key could be used to switch between CLI and GUI.
Now, for tabbed interface. I was thinking that ultimately, it's absurd for each app to have to implement it's own version of tabbing in a GUI. There should be an API for tabs. Then all programs would have a good standard interface for it. I'm sure there should be discussion of the basics UI designers would like to see in a tabbed interface, with much room for expansion.
I was also thinking about the conflict issues between OS and App key-combos, and that the best thing to do would have the OS implement, say only "Ctrl" and "Ctrl-Alt" as it's basic "starter" command, and Apps could have the "Alt" and "Alt-Ctrl" or something, as starters to open up their key-space. Now, it's possible, for example, that one could keep the exact same keybindings currently in place, and merely allow alt-ctrl to open up the input to allow for such commands as "Ctrl-f"(find, for example), since it's so common.
Or, perhaps, one could find the most universal command-combos, and integrate them into the API itself, and then when an app calls it, it would call the OS-level function(such that Ctrl-F would bring forth the search function)
Now, I know M$ is already working on this concept of metadata with longhorn/winfs system, combining a db with a search tool to find massive quantities of data.
This idea holds much promise, and Linux already has a powerful search function(slocate) as it is... running "updatedb" allows one to update the list of files on a system. Now, automate this process and allow a simpler interface that can be done from the command line, in the same way a text-box google search would work, and you have something very similar.
I believe it's Miguel de Icaza is working on this issue. I just looked. He is. It's called iFolder(from, of course, Novell), and a guy named Seth Nickellson is working on Gnome Storage, which is similar. The big issue is creating XML style metadata for everything on your system.
I know that Novell bought Ximian and Suse. Ximian has strong ties to Gnome(Miguel de Icaza, I believe was a founder of gnome, if not the original architect)... I have a feeling that something tremendous will come out of the gnomesphere regarding these technologies, and Novell may be the corporate force that can really bring it to the fore. I think Novell is dedicated to Open Source technology, and that's a good thing. Linux will be much more competitive against M$ with their backing.
Now, these are good technologies. I love tabbed interfaces(really, what are they but MDIs?) I'm curious as to what Gnome 2.6 is going to enable with their new "spatial" finder(which really is similar to the old way of doing things, ie: non-webbased file manager), but implementing the idea that people have a strong visual-spatial memory and that using this can help people organize data better.
There's so many options with UI, and new ideas to work with, it just boggles my mind, and I think if I could choose some sort of area of research, it would be Computer-Human Interface Research, but data organizational models, graphic and spatial psychology, and other such things.
I wonder what ever happened to David Gelerntners lifestreams?
I think a lot of researchers are on the right track, but they're so isolated, that it take something in the real world to show them how to implement it(and hence the corporate thing, for both financing and large scale implementation and needs for the corporation to have it done)... Also, so many pioneers are a bit too early. Look at Mark Pesce, for example, and VRML. The idea was interesting, but the processing power wasn't there, and it was perhaps a bit too complex. now we have the complex 3-D renderes and speed capable of really giving a good solid interactive experience. Perhaps VRML should have been compiled into a sort of byte-code that can be more efficiently interpreted. I don't know if he just never thought of it, or if it was just logistically complex, or what.
I also think that desktop interfaces need to be radically altered. Not quite sure how, and I think Jef Raskin has some interesting ideas, but he seems a bit pigheaded about things sometimes, and his interfaces can be a bit "too" radical sometimes.
I wonder if there's good research programs at the UW for human-computer interfaces. Well, it's fucking late late late.
Please feel free to discuss any idea here. If you have any questions, any (constructive) critiques, and insight into what could work and couldn't and make it better on any of these ideas, I'd appreciate it very much.
--------
I was thinking today about tabbed interfaces, and about my post to bugzilla.
one of the issues they were having was conflict with various OS keymapping(not the technically correct term for what I'm thinking of, really) I never thought that much about how cross platform design entails so much complication in terms of key-combos/interface interaction.
Of course, each OS/environment has it's own unique interface standards(HIGs as they're called -- Human Interface Guidelines). I must say I'm also using the term OS incorrectly here, as well. I should say "GUI" or "Window Manager"
Anyways, I had already been thinking about an OS that uses a tabbed interface as it's main way of working with the system. Something similar to the old mac systems with one bar on the top. I never got that, because I was so used to working with windows, but really it's an elegant design(does OSX have that same setup? I've not really looked).
I also have always thought that a windowed environment for a system that has strong CLI capabilities should have some sort of simple terminal always accessible.
I imagine a small bar across the bottom of the screen where most modern systems have ripped off the Win95 "start" bar.... Just one line, that could be expanded to a full terminal with a keystroke or two. One of the biggest problems is that most modern keyboards have a windows key on them, but I can envision that as a sort of key to enable a quickjump from a window to the CLI. I think it would be good if, perhaps, the scancode of that key is defined through ISO or whatever org designs such things. It shouldn't be tied to windows, and it would be good if it was given a name. Maybe some standards body has already defined it well enough. It just needs a good, universal name if it doesn't have one.
I know there's the "meta" in linux/unix, and "option" in macs... usually, i think, however, meta is considered the "ctrl" key. I could be wrong. Whatever the "windows" key is called or given as a picture(I think I've seen some linux keyboards with Tux the Penguin on it instead of a window)
Anyways, this key could be used to switch between CLI and GUI.
Now, for tabbed interface. I was thinking that ultimately, it's absurd for each app to have to implement it's own version of tabbing in a GUI. There should be an API for tabs. Then all programs would have a good standard interface for it. I'm sure there should be discussion of the basics UI designers would like to see in a tabbed interface, with much room for expansion.
I was also thinking about the conflict issues between OS and App key-combos, and that the best thing to do would have the OS implement, say only "Ctrl" and "Ctrl-Alt" as it's basic "starter" command, and Apps could have the "Alt" and "Alt-Ctrl" or something, as starters to open up their key-space. Now, it's possible, for example, that one could keep the exact same keybindings currently in place, and merely allow alt-ctrl to open up the input to allow for such commands as "Ctrl-f"(find, for example), since it's so common.
Or, perhaps, one could find the most universal command-combos, and integrate them into the API itself, and then when an app calls it, it would call the OS-level function(such that Ctrl-F would bring forth the search function)
Now, I know M$ is already working on this concept of metadata with longhorn/winfs system, combining a db with a search tool to find massive quantities of data.
This idea holds much promise, and Linux already has a powerful search function(slocate) as it is... running "updatedb" allows one to update the list of files on a system. Now, automate this process and allow a simpler interface that can be done from the command line, in the same way a text-box google search would work, and you have something very similar.
I believe it's Miguel de Icaza is working on this issue. I just looked. He is. It's called iFolder(from, of course, Novell), and a guy named Seth Nickellson is working on Gnome Storage, which is similar. The big issue is creating XML style metadata for everything on your system.
I know that Novell bought Ximian and Suse. Ximian has strong ties to Gnome(Miguel de Icaza, I believe was a founder of gnome, if not the original architect)... I have a feeling that something tremendous will come out of the gnomesphere regarding these technologies, and Novell may be the corporate force that can really bring it to the fore. I think Novell is dedicated to Open Source technology, and that's a good thing. Linux will be much more competitive against M$ with their backing.
Now, these are good technologies. I love tabbed interfaces(really, what are they but MDIs?) I'm curious as to what Gnome 2.6 is going to enable with their new "spatial" finder(which really is similar to the old way of doing things, ie: non-webbased file manager), but implementing the idea that people have a strong visual-spatial memory and that using this can help people organize data better.
There's so many options with UI, and new ideas to work with, it just boggles my mind, and I think if I could choose some sort of area of research, it would be Computer-Human Interface Research, but data organizational models, graphic and spatial psychology, and other such things.
I wonder what ever happened to David Gelerntners lifestreams?
I think a lot of researchers are on the right track, but they're so isolated, that it take something in the real world to show them how to implement it(and hence the corporate thing, for both financing and large scale implementation and needs for the corporation to have it done)... Also, so many pioneers are a bit too early. Look at Mark Pesce, for example, and VRML. The idea was interesting, but the processing power wasn't there, and it was perhaps a bit too complex. now we have the complex 3-D renderes and speed capable of really giving a good solid interactive experience. Perhaps VRML should have been compiled into a sort of byte-code that can be more efficiently interpreted. I don't know if he just never thought of it, or if it was just logistically complex, or what.
I also think that desktop interfaces need to be radically altered. Not quite sure how, and I think Jef Raskin has some interesting ideas, but he seems a bit pigheaded about things sometimes, and his interfaces can be a bit "too" radical sometimes.
I wonder if there's good research programs at the UW for human-computer interfaces. Well, it's fucking late late late.
Please feel free to discuss any idea here. If you have any questions, any (constructive) critiques, and insight into what could work and couldn't and make it better on any of these ideas, I'd appreciate it very much.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2004-05-04 08:58 am (UTC)Anyway, yes the menubar is still at the top on OS X just like on Mac Classic. But "Control" isn't the command key on Mac's, that's the, um, "Command" key (or "Open Apple" as I still call it). Thus Ctrl-F on Macs does not open up Find, Cmd-F does.
As far as tabbed interfaces, I think they're...odd. The entire idea of the Mac is not to have them, yet they're in Safari, because every other good browser has them, yet they totally break the Apple HIG. But then again, since people are used to the idea of tabs on browsers, they make sense, and although they could have done something differently, sometimes it's not about doing it differently, it's about doing what makes sense to people.
If you wanna do computer interfaces, design websites. I do Information Architecture at work all the time and it's not easy. The web is especially hard because if someone can't figure out how to use your site in about 10 seconds, they'll go somewhere else. I've seen lots of people push alternative interfaces, but mostly they're just eye candy, they don't actually help you get anything done, thus they're worthless. I think this is a UI designers biggest trap they can fall into, doing something different just for the sake of doing it is a no-no. Sometimes paradigms are set by others (say, Amazon) that aren't exactly great, but that everyone understands. The way around that is to slowly change to new paradigms so that people can follow the transition, not just throw them into something completely new.
I think you'll find both happened at Apple. If you look at System 1 vs System 9, they're very different, yet still fairly the same. The changes they made to the OS were slow and over the course of a long time and people accepted them. Then with OS X they changed things drastically and a lot of people had a fit. Some things were better, some were worse, but the fact that the change was so drastic really rubbed people wrong. It's taken a few years now for people to come to grips with it, and some still haven't (OS X adoption rate is only like 50% I believe). But again, if you look at OS 10.0 to 10.3, there have been a lot of changes, but they've been nice and gradual, nothing brash, and people have accepted those changes quite well.
I'm not sure where I'm going with this, but that's my observations on the matter.