OSDev.org

I am not a programmer. (I can do a little basic...) So, I'm not going to be creating a new operating system. I do, however, have some thoughts on the subject for anyone setting out on such a quest…

The following should be at the very core of the ideal OP.

*Learning. Designed to learn... especially how to most efficiently accommodate the user(s). Self programming/expanding capabilities.

*Security! Built in, intelligent anti-malware. (Of course, Microsoft won't do that because then they couldn't sell their anti-malware programs and recharge every year...)

*Voice Recognition/response for everything. Not to exclude mouse, keyboard, touch screen, ect.

*Software assimilation. (related to the "learning" function) Instead of installing new software from a disk or download, the OS should scan the installer, and incorporate the functions of the software into it's own system. This would maintain the look and feel of the OS (i.e. LCARS) throughout the computing experience. It would also prevent software clashes and crashes. (The menu(s) would only show commands/actions applicable to the open file.)

*A single program. Is there any real reason for an OP (or any program) to come in a million pieces like Windows?

*This one may be a little far fetched…. The OS should not have a specific look (GUI?), but rather the look and feel that the user prefers. (i.e. LCARS, Windows, Mac, ect.) or something all together different, like a 3D rendition of your favorite TV series. For example, I might want to see the bridge of the U.S.S. Enterprise from the viewpoint of the Captain’s chair. Then use voice commands to different characters for different functions. In this scenario, opening a file would direct my gaze to the view screen which would zoom full screen upon loading the requested file. Or to one of the many displays throughout the room preloaded with frequently used files.

HARDWARE:
Now, this is going beyond the bounds of this forum, but I have to get it out there…
All computers to date are using 100 year old processors…
…A cluster of 200 year old transistors. Yes, the transistors are getting ever smaller (microscopic) but still the same binary limitations.
If we must use them still, why only 1 or 2 square inches? See Isaac Asimov…
I’d love to have a twin tower system. One normal one, but without the processor. The second, (built like a miniature refrigerator), to house a massive processor(s).
While I’m at it; why do they limit how much I can expand the Ram memory? Just so I have to upgrade the motherboard first?
Anyway, back to the transistors… I wonder if it would be possible to create an artificial Neuron?
Yes, I am talking about building a brain here. Instead of having memory storage, Ram, and Processing spread all over a motherboard. It’s no wonder computers crash so much… Plus giving the possibility of more than 2 dimensional (binary) ‘thinking’.

ARTIFICIAL NEURON:
Look at an existing neuron (including DNA). Replace the Carbon with Silicon. Mathematically determine what elements would coincide with that change and make the appropriate replacements….
Well hey, I’m not a chemist or an engineer.

Look here for more of my software ideas: http://www.autoitscript.com/forum/index ... ware+ideas
Thank you in advance for your responces.

Amerigo wrote:*Learning. Designed to learn... especially how to most efficiently accommodate the user(s). Self programming/expanding capabilities.

If you think of a computer as a 'tool', which it in all actuality simply is, then AI only complicates your tool. Picture yourself using a hammer, and then it suddenly decides to do something else to 'suit' you better. IMHO, a good tool will strive to do what it's told, as efficiently as possible.

Amerigo wrote:*Security! Built in, intelligent anti-malware. (Of course, Microsoft won't do that because then they couldn't sell their anti-malware programs and recharge every year...)

Malware is for Windows, you silly goose. Who needs anti-malware?

Amerigo wrote:*Voice Recognition/response for everything. Not to exclude mouse, keyboard, touch screen, ect.

Going back to the tool design, voice recognition with current computers just does not work as well as a good 'ol keyboard. Thus, seeing as it is inherently no more efficient than anything else, there is no reason to have it.

Amerigo wrote:*Software assimilation. (related to the "learning" function) Instead of installing new software from a disk or download, the OS should scan the installer, and incorporate the functions of the software into it's own system. This would maintain the look and feel of the OS (i.e. LCARS) throughout the computing experience. It would also prevent software clashes and crashes. (The menu(s) would only show commands/actions applicable to the open file.)

Assimilation would cause instabilities. (The kernel should have as little code as possible in it; the less code, the less bugs directly in the kernel: move them out into user land!)

Amerigo wrote:*A single program. Is there any real reason for an OP (or any program) to come in a million pieces like Windows?

A single program? Perhaps you mean a small, non-bloated OS?

Amerigo wrote:*This one may be a little far fetched…. The OS should not have a specific look (GUI?), but rather the look and feel that the user prefers.

We in the software world call those 'themes', however, no theme system I know of allows Star-Trek simulations.

Amerigo wrote:HARDWARE:

The 'modern' computer is not quite as old as you would believe, and seeing as we are nowhere near to knowing exactly how the brain works, emulating it would be impossible.

Believe me, the problems you see with the OS from a user standpoint are there for a reason.

I am not a programmer. (I can do a little basic...) So, I'm not going to be creating a new operating system. I do, however, have some thoughts on the subject for anyone setting out on such a quest…

If you want such an OS, make one.

The following should be at the very core of the ideal OP.

Idealism is very personalized.

*Learning. Designed to learn... especially how to most efficiently accommodate the user(s). Self programming/expanding capabilities.

Yes, basic learning could be useful. Of course, this is for a higher level system rather than a low-level one, which is what many of us are working on.

*Security! Built in, intelligent anti-malware. (Of course, Microsoft won't do that because then they couldn't sell their anti-malware programs and recharge every year...)

You want me to put malware detection into my kernel? So that means that my kernel would have to interface with several programs running, and I don't even know if they're there. Now, viruses inside an executable would be a bit more logical to look for, but again, this ould cause a very unstable kernel, or if build right, a slow one.

*Voice Recognition/response for everything. Not to exclude mouse, keyboard, touch screen, ect.

Programs can do this, not the kernel.

*Software assimilation. (related to the "learning" function) Instead of installing new software from a disk or download, the OS should scan the installer, and incorporate the functions of the software into it's own system. This would maintain the look and feel of the OS (i.e. LCARS) throughout the computing experience. It would also prevent software clashes and crashes. (The menu(s) would only show commands/actions applicable to the open file.)

You want the program to become part of the kernel? Such a thing already exists, it's called a module. The problem is that the program would have unlimited access to resources because it would be running in ring 0. Or maybe you want to separate the kernel modules somehow, but that of course is very unstable, negating your earlier argument for security.

*A single program. Is there any real reason for an OP (or any program) to come in a million pieces like Windows?

Yes, stability, security and speed.
Speed because the kernel can be loaded and only load parts that it needs when it needs them.
Security because the kernel can stop crashing programs without crashing itself.
Stability, see Security.

*This one may be a little far fetched…. The OS should not have a specific look (GUI?), but rather the look and feel that the user prefers. (i.e. LCARS, Windows, Mac, ect.) or something all together different, like a 3D rendition of your favorite TV series. For example, I might want to see the bridge of the U.S.S. Enterprise from the viewpoint of the Captain’s chair. Then use voice commands to different characters for different functions. In this scenario, opening a file would direct my gaze to the view screen which would zoom full screen upon loading the requested file. Or to one of the many displays throughout the room preloaded with frequently used files.

Most of that isn't really the operating system, but hardware. Plus, it would be slow as hell from all of the rendering it would have to do. Granted, graphics isn't a strong point for me, but still. Lets stay with the current technology. Also, once again, you need a stable, fast kernel before you can successfully create a functioning GUI.

Hardware: Not an OS.

ARTIFICIAL NEURON:
Look at an existing neuron (including DNA). Replace the Carbon with Silicon. Mathematically determine what elements would coincide with that change and make the appropriate replacements….
Well hey, I’m not a chemist or an engineer.

How does this relate?

Overall, about 1/2 of your features I really really wouldn't want, would not work well or just in general make little sense. If you want to create such an OS, please do so. But I disagree with you. Please learn a little more about operating systems before you suggest such things.

Well, Alboin beat me to it.
I agree with Alboin though, I realize that you see problems with it from the user-standpoint, but many of them really need to be there, as they are the best of the options that have been thought of (and not all have been thought of, btw).

-JL

Thanks for the input.
Perhaps one could have a "master program" separate of the "small, unbloated kernel" that would assimulate the functions of other software?
A third part to handle the GUI(s)/Shell(s)?

One big problem I have with this is upgradability, as to upgrade one small thing you'd have to upgrade the entire system. Also, if you want to install a custom things, you'd have to write it custom, as this system wouldn't really be close to portable. And how would this be customized?

I think that the system should have the kernel, drivers, libraries and basic programs (ls, cat, echo) and the build a GUI on top of that.
The more you make bigger clumps of programs, the more you open yourself up to vulnerability and stability flaws (IMHO).

-JL

I'm not going to repeat what the others already said, hope they made it clear

Amerigo wrote:ARTIFICIAL NEURON:
Look at an existing neuron (including DNA). Replace the Carbon with Silicon. Mathematically determine what elements would coincide with that change and make the appropriate replacements….
Well hey, I’m not a chemist or an engineer.

Well, as opposed to CISC and RISC CPU architectures, ZISC are the closest to what you're talking about. They're not so much used in day-to-day computing, though, more in robots.

Hi,

Alboin wrote:

Amerigo wrote:*Voice Recognition/response for everything. Not to exclude mouse, keyboard, touch screen, ect.

Going back to the tool design, voice recognition with current computers just does not work as well as a good 'ol keyboard. Thus, seeing as it is inherently no more efficient than anything else, there is no reason to have it.

A minor note... Voice recognition is an important accessibility feature (even a simple/common problem like arthritis can make typing impossible).


Cheers,

Brendan

Yet another thread I'll simply not read... OP, please don't use colours, there are dark styled themes on this forum.

I saw the LCARS in the title, went over three Star Trek series' and 4 movies in my head, and /facepalm'ed.

Speech recognition: great plan, but sadly the current software to do that sucks. A lot.

You should probably attempt something like this as a shell for Windows or something first.

Troy Martin wrote:You should probably attempt something like this as a shell for Windows or something first.

He's not planning on making an OS.

Amerigo wrote:Anyway, back to the transistors… I wonder if it would be possible to create an artificial Neuron?

To replace the transistor?... it'd take more than one... What's the advantage?

Amerigo wrote:ARTIFICIAL NEURON:
Look at an existing neuron (including DNA). Replace the Carbon with Silicon. Mathematically determine what elements would coincide with that change and make the appropriate replacements….

If you want a neuron, why not use a neuron? (ignoring issues like stem-cell research...)

Love4Boobies wrote:

Troy Martin wrote:You should probably attempt something like this as a shell for Windows or something first.

He's not planning on making an OS.

I know, I was just putting that out there to everyone.

On voice recognition:

Ever work in a cubicle farm? With someone who is always on the phone sitting near you? Imagine being surrounded by a group of people constantly talking to their computer. I know I'd be worried about my productivity. Keeping with the Star Trek point of view, ever notice that voice recognition is ubiquitous, but the crew rarely uses it? The cubicle factor is a part of it.

In my apartment, I can name two things I would want to have voice recognition. My TV* (no more remotes) and my phone (since it is an audio device anyway). I suppose I'd also want it for the car radio and such.

I suspect that voice recognition has an "uncanny valley" effect. For it to be successful in the mainstream market, it'd need to be like talking to a human. This doesn't occur with typing and the mouse because we don't really do such things anywhere else. Whereas we are used to communicating in a certain way with speech, anything that doesn't communicate like us would more than likely be written-off as unnatural.

At least that is my two cents on the matter.

* Imagine that saying "TV off" turned your TV off, or "TV volume up" turned the volume up. I could see a late night talk show host saying something like "TV volume up, BALLOON FETISH!"

I said at the beginning, "I am not a programmer". So, yes I am clueless...
I do know that an op can be simpler in construction because a Mac OS doesn't have a zillion files like the Windows folder does. (It has been a few years since I last seen an Apple OS)

One of the last responses was about a "cubical effect"... I don't think an office enviroment would allow any OP other than what their "professionals" put on their computers. We're talking private use at home here.

Anyway, I am getting a clue now as to why there is no true LCARS System available... (aside from copyright problems)
And yes, voice recongition, as is, does suck. That's why it needs a fresh start in/with an new OP.

I have a universal remote control that responds to voice commands (InVoca). It works better than my computer...

I see your point in not having the OS do everything, but I haven't seen much in responce to my "Master Program" idea... I will recap here:
A program that will scan software downloads/cd's and assimulate the functions of that program into itself rather than installing the software. When a file is opened; only the applicable menu options would be shown.
Why? There are many programs to do simular jobs (i.e. text editors) and every one has different attributes and special functions, but there isn't one that will do it all. Also, not all are skinable. With this program, you could do everything from surfing the web to photo editing all in the same skin (i.e. LCARS). This would potentially save a good deal of harddrive space. It would grow to be a rather large program, but still much smaller than having dozens of separate programs.

Look here for more of my software ideas: http://adsa.myfreeforum.org/New_Softwar ... out78.html

You're not a programmer? Then why are you here posting redonkulously hard to implement ideas with no code or anything to show for them?

LCARS is 24th century technology, and even then it often backfires or even damages itself. What do you think would happen with today's hardware?

M-Saunders (in a different thread) wrote:Hrm. I'm pretty sure now that this is an elaborate trolling operation