Is LCARS a pratical interface?

[Jadzia]

A few random thoughts:

(1) One of the important factors for ease of use is to make functions accessible in as few user operations as possible, while at the same time making sure the number of options available at any one time is manageable so as to not overwhelm the user. There is a balance to find between breadth and depth.

(2) Related to this, we don't want to be spending forever scrolling through lists with cursor keys. This is something I find very tedious, and even with a mouse it's tedious but less so. It was undoubtedly one of the reasons mice became popular. Proposing a new interface without mice you'd have to address this.

While well designed menu systems can work well for (1), they don't work for (2). Icons and toolbars help to relieve us from the tedium of menus.

(3) As far as headache producing displays go, there's no reason to do high saturation rainbow colours on black if that proves undesirable. Simplicity doesn't demand we use that colour combination. This bbs uses white text on grey, which is relatively easy on the eye. While the simple 'edit' and 'quote' buttons could stand out more than they do, the use of colour doesn't need to be extreme.

I'll also point out, the use of text labels for these buttons. As a moderator I have an edit button on every post, not only my own, and I have once or twice hit the edit button when I meant to hit the quote button and not realised until it was too late. The use of icons over text labels might actually make the controls of a gui safer.

 

[sojourner]

Try to imagine a Photoshop, for example, or anything other then simple Notepad in an LCARS form -- you'd have nowhere to put the actions you'd do through the menus, toolbars, action palettes -- all these informational and interactive elements that you need can't be put anywhere in an LCARS system.

Good point. Thought about it two ways.

One, LCARS as we see it on the teev is for running a complex, multi system vessel, not for sending emails or prettying up your face. As Jadzia implies, it's a quick interface for a fast moving system. LCARS could have a right now real world application in things like cruise ships or power stations, to mention nothing of the military. Or, to put it another way, no matter how juiced up it is, you can't run a ship using a keyboard and a mouse, no matter what the game makers tell you.

Two, things like Photoshop or Notepad could run in a virtual window, either with a foldaway or touchscreen keyboard.

Having worked on the bridge of many cruiseships I can say this is true. You also need a joystick to go with the keyboard and mouse!

 

[Australis]

^ I hadn't thought of that!

 

[Bisz]

One, LCARS as we see it on the teev is for running a complex, multi system vessel, not for sending emails or prettying up your face. As Jadzia implies, it's a quick interface for a fast moving system. LCARS could have a right now real world application in things like cruise ships or power stations, to mention nothing of the military.

I work for a large car manufacturer and the control panels for our very complex and interconnected welding equipment are in essence LCARS touch screens. Of course they don't look as pretty, but aside from the purely cosmetic differences the functionality is pretty much the same.

Would I want to use LCARS to surf the net? No.

Would I want to use LCARS to reset a fault and master up an automated welding line? You bet!

 

[Crewman47]

Try to imagine a Photoshop, for example, or anything other then simple Notepad in an LCARS form -- you'd have nowhere to put the actions you'd do through the menus, toolbars, action palettes -- all these informational and interactive elements that you need can't be put anywhere in an LCARS system.

Good point. Thought about it two ways.

One, LCARS as we see it on the teev is for running a complex, multi system vessel, not for sending emails or prettying up your face. As Jadzia implies, it's a quick interface for a fast moving system. LCARS could have a right now real world application in things like cruise ships or power stations, to mention nothing of the military. Or, to put it another way, no matter how juiced up it is, you can't run a ship using a keyboard and a mouse, no matter what the game makers tell you.

Two, things like Photoshop or Notepad could run in a virtual window, either with a foldaway or touchscreen keyboard.

Having worked on the bridge of many cruiseships I can say this is true. You also need a joystick to go with the keyboard and mouse!

I was just thinking there that the only time that we see LCARS in use are on Starfleet installations and vessels, for all we know civilian home computers might use something completly different like what we have with Windows and Macs.

Just a question though but what sort of interface do present miltary vessels use on there computers, is it Windows or something different? Maybe it's the same in Star Trek.

 

[Jadzia]

I think the military use windows xp. But they're intending to move to vista, or so I've heard. Windows 7 is too new to take risks with, so they always are one generation behind the rest of society.

 

[Robert Maxwell]

One thing that hasn't been mentioned as far as I can tell is that the LCARS interface probably adapts in real-time to the task at hand. That's something today's computer interfaces don't do. They don't "learn" from what the user is doing except on a very basic level. The LCARS system seems very context-specific, which is exactly what you want in a user interface. Pretty icons are distracting eye candy. You want your data front and center. There is nothing at all wrong with a text-heavy interface. The text in an LCARS interface can be color-coded which can be just as useful (if not moreso) than iconography.

But I think the key piece we are missing is an interface that learns from user input and gradually adjusts for maximum efficiency. A system that predicts what the user will want to do with a high degree of accuracy. I'm thinking specifically about instances where time is a critical factor, such as a sudden encounter with an enemy ship. Worf shouldn't have to press a button to bring up a tactical scan of the ship--it should just happen because we've got an unknown ship closing in on us, and there is no way we wouldn't want an immediate tactical assessment. Additionally, based on the decided reasponse to the threat--disable, evade, destroy, etc.--you shouldn't have to adjust phaser strength or photorp yield. You just pick what level of response you want and hit the "fire" button.

So, no, I don't think LCARS is necessarily a good interface for today's computers, unless we're dealing with a limited range of applications. We just don't have the computing power or the AI for a fully flexible self-altering interface. Although that would be really awesome and I think it's something to aspire to. Static interfaces are just a limitation of today's technology and our development capabilities.

 

[Roshi]

Check out the "Pocket Unvierse" App for iPhone:

link.

 

[Trevacious]

And who's got the crappy job of wiping all those greasy fingerprints off the interface panels?

 

[Bill Morris]

One thing that hasn't been mentioned as far as I can tell is that the LCARS interface probably adapts in real-time to the task at hand. That's something today's computer interfaces don't do. They don't "learn" from what the user is doing except on a very basic level. The LCARS system seems very context-specific, which is exactly what you want in a user interface. Pretty icons are distracting eye candy. You want your data front and center. There is nothing at all wrong with a text-heavy interface. The text in an LCARS interface can be color-coded which can be just as useful (if not moreso) than iconography.

But I think the key piece we are missing is an interface that learns from user input and gradually adjusts for maximum efficiency. A system that predicts what the user will want to do with a high degree of accuracy. I'm thinking specifically about instances where time is a critical factor, such as a sudden encounter with an enemy ship. Worf shouldn't have to press a button to bring up a tactical scan of the ship--it should just happen because we've got an unknown ship closing in on us, and there is no way we wouldn't want an immediate tactical assessment. Additionally, based on the decided reasponse to the threat--disable, evade, destroy, etc.--you shouldn't have to adjust phaser strength or photorp yield. You just pick what level of response you want and hit the "fire" button.

So, no, I don't think LCARS is necessarily a good interface for today's computers, unless we're dealing with a limited range of applications. We just don't have the computing power or the AI for a fully flexible self-altering interface. Although that would be really awesome and I think it's something to aspire to. Static interfaces are just a limitation of today's technology and our development capabilities.

Windows and others, as you know, do some simple things like showing recent files/programs accessed, as well as sometimes anticipating where you would like a download to go. And search engines, etc. try to anticipate what you're looking for and offer suggestions.

I've tried to do some of that in LCARS 24. I finally got around to adding bookmarking of files in for the upcoming release, where it opens a text file scrolled to where you had it last time. Games open with any unfinished game still in progress. Puzzle servers remember the last one completed and go to the next. The last mode selected and various options for some programs are also saved to a log and envoked when opened the next time. It's not exactly AI, but the system takes a lot of notes and tries to please. For word processing, it does completion of a partial word typed when you hit the ALT key, by referring to what's above in the file and doing backward search or using a predefined macro. That's quicker and smoother than selecting a suggestion from a dropdown menu. So have been trying to implement what you suggest as much as possible.

 

[Robert Maxwell]

Yeah, that stuff is simple. I'm talking about a fully adaptive interface that figures out what elements you use most frequently and what ones you don't, and rearranges them accordingly. In addition, it would use contextual information from other systems to bring up the right display at the right time without any user prompting at all. Fairly advanced stuff that we just aren't doing today.

I have no doubt you are working on some of that for your LCARS system but I think a lot of it is just plain beyond today's computers. GUI toolkits in particular rely on you building a fairly static interface that can change contexts but in general the layouts are not adaptive and certainly not predictive.

 

[Bill Morris]

What we're getting from mainstream now is nearly the opposite: unwanted popup ads, sudden reconfiguration of a Windows app or destruction of user data in response to an accidental wrong click or keypress, causing a crisis for the user. When Windows decides to throw up a firewall to block the user from uploading by FTP, etc., the option to turn it off doesn't come up front as you would like but is instead very well hidden. In that case, it is the opposite.

Another example is if you open Notepad or whatever not remembering that you've got it docked on the taskbar with the same file, then edit, save, and close the new one, your changes can accidently be destroyed if you later maximize the docked version, because its easy to save that old version (and may be prompted to do so), and there's no warning that you're overwriting your changes made with the later opened iteration of that same app and same file.

So just getting rid of those booby traps sounds like something to do first. I've already explained why sitting on an LCARS panel doesn't invoke accidental input.

 

[DiSiLLUSiON]

Those small things could be changed, yes, but a program running on today's computers can never achieve the level of context-sensitivity needed to be able to predict, to any large degree, what action the user wants to take next. And it's simple why, actually; in order for it to be able to predict more then a simple context, it also needs more input, and it's not getting that.

For example, If a PC had voice recognition abilities and a medium of intelligence, it could predict that, when the boss says to the employee "the deadline for the car project is now today", it would temporarily save and close all other files, bring up the files of that project and hide everything from sight that isn't geared toward completing that project.

(There are lots of other examples you can think of).

But Artificial Intelligence simply isn't good enough; even if those inputs would be connected, the computer can't make heads or tails of that information. Yes, you could design a system based on states of readiness, but that wouldn't do either; who's to say that, for example, when a deadline is moved forward, that every employee should work on that project? There are other projects with deadlines, too. Not only that, but the time it takes to input that state of readiness is the same as simply forwarding a mail to your employees; they know what to do.

So, firstly, we need artificial intelligence. Then we need to hook up all possible inputs of information. Then, we need to let the software use deductive reasoning to figure out what's going on. We also need it to know what possible actions to take, what the result of those actions would be a few steps ahead, and be able to actually choose the correct action. Then we would have to design an interface that can change context in an easy way. Problem is; you'd also have to devise ways for the interface to not change context; there are always small things you have to do in the mean time that don't have anything to do with a current project. Not only that, but sometimes you need software or information that, on the surface, seemingly has little to do with the project.

So there's still a ways to go; fortunately, we know a thing or to about interfaces these days. They won't use bright on black colors, they'll use good typography, they'll use good iconography, they'll use recognizable interface widgets, they'll use a natural method of user input and, as today, they have the ability to do a whole load of different operations.

However, the combination of all those points is everything LCARS is not, so I doubt it would be modeled after that. At least the graphical part of it; the interactive (you could say "mechanical") part is all a guessing game anyway, since LCARS, as seen on the show, doesn't exist. There's probably more to it then simply some buttons in bright colors , but we wouldn't know. As such, all we can debate about concerning LCARS is it's graphical elements, and those are woefully inadequate.

 

[msbae]

I work for a large car manufacturer and the control panels for our very complex and interconnected welding equipment are in essence LCARS touch screens. Of course they don't look as pretty, but aside from the purely cosmetic differences the functionality is pretty much the same.

Would I want to use LCARS to surf the net? No.

Would I want to use LCARS to reset a fault and master up an automated welding line? You bet!

Which one?

I think the military use windows xp. But they're intending to move to vista, or so I've heard. Windows 7 is too new to take risks with, so they always are one generation behind the rest of society.

The military uses a specially modified, ultra-secure version of XP Professional.

One thing that hasn't been mentioned as far as I can tell is that the LCARS interface probably adapts in real-time to the task at hand. That's something today's computer interfaces don't do. They don't "learn" from what the user is doing except on a very basic level. The LCARS system seems very context-specific, which is exactly what you want in a user interface. Pretty icons are distracting eye candy. You want your data front and center. There is nothing at all wrong with a text-heavy interface. The text in an LCARS interface can be color-coded which can be just as useful (if not moreso) than iconography.

But I think the key piece we are missing is an interface that learns from user input and gradually adjusts for maximum efficiency. A system that predicts what the user will want to do with a high degree of accuracy. I'm thinking specifically about instances where time is a critical factor, such as a sudden encounter with an enemy ship. Worf shouldn't have to press a button to bring up a tactical scan of the ship--it should just happen because we've got an unknown ship closing in on us, and there is no way we wouldn't want an immediate tactical assessment. Additionally, based on the decided reasponse to the threat--disable, evade, destroy, etc.--you shouldn't have to adjust phaser strength or photorp yield. You just pick what level of response you want and hit the "fire" button.

So, no, I don't think LCARS is necessarily a good interface for today's computers, unless we're dealing with a limited range of applications. We just don't have the computing power or the AI for a fully flexible self-altering interface. Although that would be really awesome and I think it's something to aspire to. Static interfaces are just a limitation of today's technology and our development capabilities.

Your description of LCARS is almost exactly like the one in the Next Gen tech manual. As a mission-specific OS for a starship, LCARS is great. For what I do on my desktop, not so much. Windows does that job quite well.

 

[Bisz]

Which one?

I take it you mean which car company? Toyota.

 

[bornwithapen]

hi people
just joined the site and it looks really interesting with plenty of intelligent people.
Regarding this topic as to whether LCARS would make a Realistic interface, the Answer i believe is a resounding yes, however the mentality of today's programmers needs to change, i have for the past year been working on a design for an actual Application that uses the LCARS interface exclusively, the Application is a Show control Program for nightclub Lighting, you know all those pretty moving lights that you see in a club, well i play with those for a living , that's right someone pays me to turn lights on and off to music, anyway, the only part of my job that i hate with as much passion as i have for the job is the horrible interface that i have to use to make the lights work.
Sadly its the same story on all the 12 programs that i had a look at, all wrapped up in a ridiculous amount of onscreen clutter, non intuitive, very steep learning curve etc etc yada yada, don't get me wrong im not stupid by any means, but i started lighting to enjoy it, not to have to learn some programmers idea of what lighting software should be like.
And this is where the beauty of LCARS comes into its own, if done correctly the LCARS interface can outperform any windows or mac program hands down, as long as the mentality is correct, what i mean by that is if the programmer thinks about the role of the software then a lot of what is placed in current software is irrelevant to the use of the program, im a lighting operator, that means i operate lights, that's it, i go in and play a keyboard and lights come on when i press the keys, i don't need to know what DMX channel is doing what at what particular smpte time code point, but the software i am using shows me all kinds of stuff that i don't or will ever need to know as an operator.
As i mentioned earlier i am designing a program that addresses these issues, things like asking the user for their name, hand preference and role in the club would then allow several things to take place, one the program could refer to the user by his or her first name, obviously this would be text based as computers dont have a decent speech generator(least i dont think they do), by knowing what hand the user prefers the program can be presented in a left or right handed version and by knowing what role the user occupies , the program can present only that which is needed.
as far as knowing what is used the most and adapting the interface to suit, that is a relatively simple thing to add to something like an LCARS system due to it being a Text based interface, its all down to the programmers mentality.
the other perfect thing about LCARS is its vector based, so in essence the computer now only has three things to give CPU power to, colour, vector lines and text, i think it would be like racing a veyron against a Mini.
the other beauty of LCARS is its aesthetics, there is something so much more attractive about the LCARS Layout than the clunky inelegant interfaces we all have to put up with.
so it's nothing to do with the power of computers, its nothing to do with can it be done, but simply its the programmers mentality that is stopping LCARS from becoming an actual application GUI.

 

[sojourner]

BRRaaaaiiinnnnsss.

 

[Robert Maxwell]

BRRaaaaiiinnnnsss.

No spam, please. The post above yours is legit, though I was tempted to spambot the poster at first until I read it.

It's on topic and it's not as if we have a lot of LCARS threads so I'll let this go.

But seriously, no "brains" comments.