OSDev.org

Dear All,
The new OS design for so-called "next generation of hybrid OS" is attached in this file.
All of the mentioned ideas in this document have appeared in many Graphic OS and applications. I just think to mix all them to a new generation of OS. This may enhance the user's interaction so much.
You can see here the mixture of
- Windows: this new design based on Windows
- AIX, AutoCAD: every action by mouse have a corresponding command line
- Bart PE: option for fast OS
- HP Windows: fast response from sleep state
- Visual Studio: the way to work with objects via property or method..
This is small extraction:
Disadvantages of using mouse in Windows:
• When using mouse to locate a menu and start an application or function, we have to look at the mouse to see the exact position so that we are distracted by this action
• Thinking progress becomes slowly .
• Mouse used by only hand with active index finger to click to run a task.
The frequency to use mouse reduce considerably much
New OS
• We could use keyboard in almost all tasks or call the menu/command to come and appear to serve us.
• Thinking is not stopped or distracted by viewing.
• Could use all fingers when typing, thus the speed of INPUT is faster.
....
However, my new design needs the huge improvement from input devices as you may see in this attached document. One special requirement is new KEY, which call the command line whenever pressed.
Hope to here from some expert here.
Rgds,
Tran Quang Thinh

This sounds more like a window manager than an OS.. see http://wiki.osdev.org/Beginner_Mistakes

Also, it seems like you are just describing using a command line or keyboard shortcut system, which is what you find in many UNIX-like systems already. It's really just the difference between a GUI and a CLI - the CLI is faster for most tasks, but harder to use because you need to know what exactly you want.

earlz wrote:This sounds more like a window manager than an OS.. see http://wiki.osdev.org/Beginner_Mistakes

Thank you for your reply. When I explain this first time with my friend, he also said like you. However, to think more carefully you may find out:
1. Windows 3.1/95 could be GUI DOS manager
2. My design need to change the inside kernel of OS in order to support some things like: object oriented management of application, cross-command between applications (such as Manual OLE, pipeline command)
3. Hardware support for new kind of buffer of keyboard (such as to differentiate between real key press and simulated key-press to transfer keys from special-key to OS)
There are also other things I could explain more to explain it is not Windows manager.
ThinhTQ

thinhtq wrote:

earlz wrote:This sounds more like a window manager than an OS.. see http://wiki.osdev.org/Beginner_Mistakes

Thank you for your reply. When I explain this first time with my friend, he also said like you. However, to think more carefully you may find out:
1. Windows 3.1/95 could be GUI DOS manager
2. My design need to change the inside kernel of OS in order to support some things like: object oriented management of application, cross-command between applications (such as Manual OLE, pipeline command)
3. Hardware support for new kind of buffer of keyboard (such as to differentiate between real key press and simulated key-press to transfer keys from special-key to OS)
There are also other things I could explain more to explain it is not Windows manager.
ThinhTQ

Then you need a virtual keyboard and mouse driver and/or a new API for controlling GUI elements. I see no ring 0 code here.

Have you looked into making your own Linux Distribution?

earlz wrote:Have you looked into making your own Linux Distribution?

Sorry, I have no idea about that. I just work much with Windows and AIX. I would see it more and explain about your thinking.
In my knowledge, I just need a small improvement in current keyboard hardware here. You can find it simple that combined or stimulated key is just a key with NO TIME SLOT or SLIDED TIME for any insertion of other keys (from users).
ThinhTQ

thinhtq wrote:

earlz wrote:Have you looked into making your own Linux Distribution?

Sorry, I have no idea about that. I just work much with Windows and AIX. I would see it more and explain about your thinking.
In my knowledge, I just need a small improvement in current keyboard hardware here. You can find it simple that combined or stimulated key is just a key with NO TIME SLOT or SLIDED TIME for any insertion of other keys (from users).
ThinhTQ

I'm pretty sure all you need is your own linux module and/or a kernel module to do this "simulated key" thing.

I don't understand what you mean by no time slot here.

earlz wrote: I'm pretty sure all you need is your own linux module and/or a kernel module to do this "simulated key" thing.
I don't understand what you mean by no time slot here.

Thank you very much for your interest in my ideas. However, I think you may have not read carefully my design in the text file. The main idea of the design is:
- Help user to maintain point of thinking than active windows or visible point
- Call the menu/command to come to us rather than we have to look for menus
You know when running application, you can use parameters as: outlook /safe; shutdown \\machinexx /r /t: 50
However, in Windows Application, there are less applications to support parameters (especially method parameter), and also the parameter is not complicated.
In my Design, I aim to enhance the parameter side of applications. This is a kind of message input to application, that helps you to work with application even when it is inactive (hidden from active windows). For example, when you are working with excel file, you can print winword file via name of it (enter the command like: wordfile1.print to printer1) without activating winword windows.
Stimulated key driver here is not enough if the core of OS or application do not really support those features. This driver even make the system worse.
To give you clearer view of stimulated input, I could show you this example. When you want to stop music while working on email (Outlook, for instance) you can:
1. Press command button to make command line appear (CMD)
2. Type: player1.stop
3. Press command button to Run
This seems more efficient than you need to switch to media player and look in to the stop button of application naming player1, then click mouse on that. So the serial of keys would be: CMD-PLAYER1-STOP-CMD
When you are in the player1 application, you can type: CMD-STOP-CMD
From this point I suggest an application key, for example, when you press this hardware key, there is serial of key CMD-STOP-CMD send to key buffer. The 3th hardware producer could make a key like button or switch to connect to our main keyboard. When press or turn on this key, it send a series of key as: CMD-STOP-CMD. This series of keys is one by one without any delay time for any insertion of key (that is what I say NO TIME SLOT or SLIDED TIME for other keys ).
And this STOP key could be use for any applications which support STOP function. To think about it you would see this helpful if there is button NEWTAB as a scrolled button. When you scroll this button by one degree, it will send command like: CMD-TAB-CMD, so the active button would be changed from menu/button to next menu/button (as we use TAB key). You can use application effectively.
Kindly show more concerns then I could explain more.
Regards,
Tran Quang Thinh

thinhtq wrote: In my Design, I aim to enhance the parameter side of applications. This is a kind of message input to application, that helps you to work with application even when it is inactive (hidden from active windows). For example, when you are working with excel file, you can print winword file via name of it (enter the command like: wordfile1.print to printer1) without activating winword windows.

Why would be pressing 27+ keys (in case of the "wordfile1.print" example) be quicker than three clicks (one to activate word processor application, one to activate the menu and another for the Print command)?

thinhtq wrote: Stimulated key driver here is not enough if the core of OS or application do not really support those features. This driver even make the system worse.
To give you clearer view of stimulated input, I could show you this example. When you want to stop music while working on email (Outlook, for instance) you can:
1. Press command button to make command line appear (CMD)
2. Type: player1.stop
3. Press command button to Run
This seems more efficient than you need to switch to media player and look in to the stop button of application naming player1, then click mouse on that. So the serial of keys would be: CMD-PLAYER1-STOP-CMD
When you are in the player1 application, you can type: CMD-STOP-CMD

Same applies here. Also see below.

thinhtq wrote: From this point I suggest an application key, for example, when you press this hardware key, there is serial of key CMD-STOP-CMD send to key buffer. The 3th hardware producer could make a key like button or switch to connect to our main keyboard. When press or turn on this key, it send a series of key as: CMD-STOP-CMD. This series of keys is one by one without any delay time for any insertion of key (that is what I say NO TIME SLOT or SLIDED TIME for other keys ).

You just described the functionality of those "multimedia" keyboards, which come with programmable keys. I don't own one so I can't comment on how well they actually work, and whether some proprietary utility is required to "wire" the functions to the keys.

thinhtq wrote: And this STOP key could be use for any applications which support STOP function.

How does the system know which application I want to stop. For example, I sometimes watch a TV series while the music player is streaming an online radio. If I press this hypothetical STOP button, which application gets the message, the video player or the music player?

thinhtq wrote:To think about it you would see this helpful if there is button NEWTAB as a scrolled button. When you scroll this button by one degree, it will send command like: CMD-TAB-CMD, so the active button would be changed from menu/button to next menu/button (as we use TAB key). You can use application effectively.
Kindly show more concerns then I could explain more.
Regards,
Tran Quang Thinh

I can't even interpret what you are talking about here. If you refer to the "focus" (dotted selection rectangle in case of Windows UI) jumping from one control to another (in this case to the next button or whatever), this already exists since... Like Windows 3.1 or probably before.

[quote="ru2aqare"]
Why would be pressing 27+ keys (in case of the "wordfile1.print" example) be quicker than three clicks (one to activate word processor application, one to activate the menu and another for the Print command)?

You have a good point, but this is an example. I just want to show two things if you read my design: when typing, you do not have to see the screen and can do it while thinking, you can maintain your thinking. The speed would be better if you could use 10 finger much more than moving your hand to find the mouse.

You just described the functionality of those "multimedia" keyboards, which come with programmable keys. I don't own one so I can't comment on how well they actually work, and whether some proprietary utility is required to "wire" the functions to the keys.


Yes, you are true. This means the way I bring a lot of actions (pull, push, drag..) to objects in system (you could see in my design). For example, you can program a key with message: SIZE + 1, when you scroll this key, the selected font's size could be increased one by one. If the application supports lots of command lines related to menu (such as AutoCAD), the function key is very helpful. Especially, when you work with some graphic interactive applications or games.

How does the system know which application I want to stop. For example, I sometimes watch a TV series while the music player is streaming an online radio. If I press this hypothetical STOP button, which application gets the message, the video player or the music player?


Kindly read my design again, the STOP function is activated with current active windows in case you do not have other directed arguments.


I can't even interpret what you are talking about here. If you refer to the "focus" (dotted selection rectangle in case of Windows UI) jumping from one control to another (in this case to the next button or whatever), this already exists since... Like Windows 3.1 or probably before

You are right, but those activities are not connected to visible hardware keys. Moreover, this keys designed to follow the imagination of producer, their arranges are setup by users. It would be easier for users to work with application.

Don't use colours, especially red.. just awful.

http://forum.osdev.org/viewtopic.php?f=1&t=16944

As for a hybrid between "Unix" and "Windows", honestly.. you should be ashamed of yourself.

I think the biggest problem with the ideas proposed so far is the lack of empirical evidence for some fundamental assumptions. For example:

thinhtq wrote:You have a good point, but this is an example. I just want to show two things if you read my design: when typing, you do not have to see the screen and can do it while thinking, you can maintain your thinking. The speed would be better if you could use 10 finger much more than moving your hand to find the mouse.

First of all: where is the user looking, if not at the screen, when (s)he is typing? Supposing they know how to type, they shouldn't have to look at the keyboard. In any formal touchtyping lesson, the user is indeed taught to look at the screen, however, the important issue is how most users actually behave when using the computer. I think you're making the assumption here that reaching for the mouse necessarily requires the user to fully divert attention and thus breaks his/her concentration. This is probably not actually the case for users who have for a long time habitually used a mouse to operate a computer.

I think your design could benefit if you used axioms from cited research rather than your own, presumably untested observations. That is, unless this project is mostly intended for your own use only. No offense intended, just my 2 cents.

thinhtq wrote:Yes, you are true. This means the way I bring a lot of actions (pull, push, drag..) to objects in system (you could see in my design). For example, you can program a key with message: SIZE + 1, when you scroll this key, the selected font's size could be increased one by one. If the application supports lots of command lines related to menu (such as AutoCAD), the function key is very helpful. Especially, when you work with some graphic interactive applications or games.

thinhtq wrote: Kindly read my design again, the STOP function is activated with current active windows in case you do not have other directed arguments.

I admit not having read your design document. However. There are only a limited set of function keys on the keyboard (most of them have ten or twelve), as opposed to tens (or hundreds) of commands that can be placed in menus. Also, not every command is meaningful to every application. To stick with your example, the graphic designer application can't interpret a "stop playback" command - and you have just wasted a key.

Edit: after reading the design document, I would also argue with some (most?) of the presumptions of your design. To mention a few:
- "Letĺs imagine if you just push a button to call a command window, then you type ôgoto excel1ö you will be at Excel window. This is clearly save time of users."
I still think that clicking *twice* (or pressing alt+tab, two keys) beats typing "goto excel" (ten characters) every time. Even with auto-completion (type G, space for "OTO ", type E, space for "XCEL") I don't see the benefits. This gets worse if more than one instance of the same application is running.
- "The frequency to use mouse reduce considerably much. We could use keyboard in almost all tasks"
Ever tried selecting a non-rectangular area in Photoshop?
- "Every application could be considered as an object ... As a result, they may include two main parts: method and property. Methods could be considered as functions, which fulfills a specific task basing on parameters provided to them."
While this could be arguably true, how would you perform (to stick with the previous example) image manipulation in Photoshop? That can't be done without actually looking on the screen. Furthermore, to actually make use of this method/property system, *every* application would need a thorough rewrite. Sorry, but I just don't see that happening.
What you described here reminds me of batch files or those Unix programs that perform one specific task, and the results of one may be sent to the input of another.
- "User needs to find or look into menus then run after windows. So he becomes slave rather than master of Windows."
The new, so-called Ribbon user interface of Office 2oo7 was actually designed to address exactly this issue.
- "Key-set for application means we could make specialized keyboard for every application."
Try convincing keyboard manufacturers.
- "When we allow application object to have properties and methods (like .NET programming) and allow users to transfer message to application object properly help users to remember easily and manipulate on application."
Sorry, but I see this as requiring users to actually *program* their PCs. Not everyone is/wants to be a programmer just to be able to browse the internetz and watch silly *tube videos.
- "OS will have many modes (such as Normal and Safe mode currently), include those like Television, Game, MS Office, Printing Modes... Each mode allow OS to have access to specialized data ..."
So after having watched a movie I would need to restart the OS just to access for example Visual Studio?

And, finally, my favourite:
- "o Make user like Windows as drug thanks to learning and training progress"

Brynet-Inc wrote:Don't use colours, especially red.. just awful.

http://forum.osdev.org/viewtopic.php?f=1&t=16944

As for a hybrid between "Unix" and "Windows", honestly.. you should be ashamed of yourself.

Thank for your reminding as regards posting. It is really difficult to explain this with some people who have not spent time on working on both of OSs. Even my friend, a post-doctor who are working in IT in US, he also think this explaination as interface changes as he has never worked as real administrator.

madeofstaples wrote:I think the biggest problem with the ideas proposed so far is the lack of empirical evidence for some fundamental assumptions. For example:

thinhtq wrote:You have a good point, but this is an example. I just want to show two things if you read my design: when typing, you do not have to see the screen and can do it while thinking, you can maintain your thinking. The speed would be better if you could use 10 finger much more than moving your hand to find the mouse.

First of all: where is the user looking, if not at the screen, when (s)he is typing? Supposing they know how to type, they shouldn't have to look at the keyboard. In any formal touchtyping lesson, the user is indeed taught to look at the screen, however, the important issue is how most users actually behave when using the computer. I think you're making the assumption here that reaching for the mouse necessarily requires the user to fully divert attention and thus breaks his/her concentration. This is probably not actually the case for users who have for a long time habitually used a mouse to operate a computer.

If you are UNIX/Linux skilful users, I think you would never say like this. Especially, if you are expert in Windows support, I think you should remember the saying "try to use mouse as less as possible". And when the Windows user, who you are supporting, said that: "Oh, I cannot see what you are doing", I thing you are quite successful at that time.
My suggestion for you when you want to explain about speed to use mouse and keyboard is that:
1. Try to SMITTY/SMIT in xNUX system fluently and then tell some things.
2. Try to use some Oracle applications at your highest speed and then show me your feeling.
As far as you are not an addicted to Unix or Linux, it is difficult to explain more.

I think your design could benefit if you used axioms from cited research rather than your own, presumably untested observations. That is, unless this project is mostly intended for your own use only. No offense intended, just my 2 cents.

In fact, if my design is so easy, many people could find out it earlier. I do not have any cited research until now except for my experience. What I have in my experience is to witness and summarize activities of many skillful users. And the main thing I want to explain here is the improvement from ADMIN/USER SIDE, to mix best features of both Unix and Windows in new generation of OS. Developers in core OS could do this design for me as I not the genie who could make the OS!
Regards,
Tran Quang Thinh