DSL Ideas and Suggestions :: DSL

///

Cool post. Who doesn't think about these things all the time? (I know I do!)
I'll just respond to some of the points you mentioned, okay?

Quote

DSL A) has access to a list of drivers from its repository or the internet that will then be converted into an abstract driver. B) if it does not have access, it can "hack" and discover what the hardware is and create an abstract driver; this can be done with the help of Q&As and user input. For this, an intense assembly code will be required so to discover the unknown and to create the abstract driver.

Regarding the actual code, I must ask: If human programmers are having problems understanding close-sourced hardware, surely it would be even harder to write such a software program?

Quote

Initially, the mouse will appear and a background image fades in. In the workspace will appear eye-candy dialog boxes that have lists of commands, such as access to home and things to do. The workspace(not command-line and workspace selector) also called the "face" can be manipulated and reconstructed to the users liking with backgrounds, or fitted to a 3D shape.

Regarding this point, my question is: Could the same operations not be performed faster and more easily through the shell? What would be the advantages of such a graphical workspace?

Quote

A web browser will contain the basic browser construct, but more so, access to the deconstruction of the web page in terms of links, text, and protocols. From these we can create bookmarks, lists of links, and we can use and reroute this information if we need to do so.

This sounds exactly like the idea of the semantic web, which seems to have become a 'hot topic' in recent years.

Quote

If I want to see my network connections, and routing information, it would appear similarly to "JACK" connections . If I would want to see the information going through the network, I would have to interpret it with the basic constructs that I have available and, thus, create an application.

Regarding this point, do you not think that shell apps like the standard net-tools package (e.g. 'netstat', 'route', etc) and nmap already present a good visual representation of the network space?

Quote

These things to do are binded to the application ingredients. How can I interpret this information using graphs, sound etc...? How can I interpret images and sound? I will need a signal source, and use an FFT to extract the frequencies, but I will not have to know how an FFT works, but just the results of its use. From this, I should be able to construct an application, a new command, an icon.

Regarding this point, are not present-day multimedia codecs and media players already sufficient for this purpose? If not, what do you think is lacking?

////////

Quote

if it does not have access, it can "hack" and discover what the hardware is and create an abstract driver

And my car can run on water and jet Coca-Cola out the exhaust! Here in the real world, that isn't very feasible. Vendors don't care to show their proprietary systems to open source driver writers which causes those driver writers to come up with often really flaky kludges to get things working *partially* in Linux and BSD environments. So what makes you think all of a sudden open source driver writers can come up with a solution that magically writes what they can't and better than they do now? You wanna buy a bridge?

Quote

this can be done with the help of Q&As and user input.

You have three posts here as I write this. I don't know if you've seen some of the questions asked here. I think you're overly optimistic average users are capable of answering the kinds of questions that would be required to create a driver on the fly.

Quote

As s_i pointed out, that's semantic web-related and not something best handled within the operating system itself. One of the prevailing ideas is that the browser will be the main, if not only, interface for users. I think that's already the case since most people use browsers not only for web and e-mail service, but increasingly for what's traditionally been handled with separate applications. But that's where desegregation, aggregation, and integration of data will more likely occur than in an OS itself.
http://lucky13linux.wordpress.com/2007....butions

Quote

this is what Linux would be if it were represented as a GUI OS. In essence, it would also be damn small!

Not with all that 3d crap you wrote about. It would be damn huge and a drain on computing resources.

PS: Among the last words I want associated with my operating system is "avant garde."

Next Page...
original here.