Anyone else been there???

[_mark()]

I started off a few months ago with a pretty neet idea for a CAD program. While starting some code for it, I had a pretty good idea for an application framework. While implementing a prototype of the app framework, it became clear that a java like language would enhance the whole programming experience. So while implementing the VM for that, it seemed like a desktop environment implmented for the VM would be cool. But of course this desktop should run in other environments also, so a good cross-platform library would be needed. Of course I would want to write this also. So some interfaces were schetched out, but while doing more corss-platform research, it became clear that what I really needed was a new compiler - one that supported my new language directly. Of course during this reseach, it became clear that to support this new language, I really needed an OS that had more advanced features, so the OS development research began. Most of my research has led me to the conclusion that the OS+other goodies could likely take 10 years or more to complete by myself.

So here I am. Begin from ground zero and develop an OS + tools etc from scratch. This could essentually take the rest of my life - since I do have a job and family. Or do I scratch all my really cool ideas (all of which are scetched out in legal pads - everthing from new FS and new exe formats, to high level user mode routines) and focus on something a little more usfull - like a CAD program....

The ideas just seem to be endless. Even new and better hardware designs have entered the legal pads in a couple places....

Am I alone in this? Or were you all in the same place once upon a time?

Thanks
Mark

[pskyboy]

LOL, your far from alone i have set out on this same journey. Its a funny thing isn't it you start with one thing think that would be nice and continue down the slippery slope to restructuring the way transistors are layed out on silicon.

Seriously i have done exactly the same thing and am co-developing my OS and new language. I have however thought about new processors and got into desigining them as well but had to stop myself somwhere. I still leak into doing a bit of work on a processor now and again the and my language needs one anyway.

If your interested in the language im developing let me know and ill do a post on it as i would be interested to here peoples ideas. You never know it may end up we can get a group of people together to implement this language so it a) actually happens or B) gets doen in a reasonable amount of time.


Peter

[_mark()]

Sure - I love to look. I'd prefer just to read any whitepapers you have - but source is fine if you do not have any.

Mark

[pskyboy]

Hey Mark

Your a bit ahead of me there, the white papers i have are very bad and mostly just me throwing ideas down so are highly criptic. Anyway here is a copy i sent to whyme_t which is a brief summary of the language.

Hey whyme_t, yeah I think we may well be thinking along the same lines. I was thinking of a Java type language but with lower level abilities.

It would basically be a fully object orientated language like Java and have a virtual machine. The big difference would be that it would have support for checked pointers, platform specific code and the virtual machine would really only be there as a way to allow for platform independent code. What I mean by this is when an application got installed, as almost all have to be now it would compile the source to machine code for that machine. The virtual machine however would allow for executables to be distributed and when you compile your source it would compile to the virtual machine. I was also thinking of having an ability in my OS so if you right click on a VM file it would have a compile to native code option. This could also be implemented for Linux and windows.

The nature of my language would mean that it could be used for OS and driver development but also remain highly platform independent like Java. Obviously there would be a few thinks to work out but all in all I think I have worked most the problems out and any left I?m sure can be fixed.

If you?re interested it might be worth combining ideas on languages and seeing if we could spec a language. I was originally going to right this language first and then do my OS in it but I got impatient and started my OS in c++. I have however written my C++ Kernel in a way that will hopefully be easy to convert to my language. I wouldn?t mind getting a group of people together to implement the language as this would allow for it to be done quicker and allow for high speed platform independent software.

Peter

[_mark()]

I had different ideas, none of which could really be described in a short. I plan on doing some formal whitepapers and getting my UML scetches into the putter soon. I'll try and post things here when I have something tangible which should include quite a bit of prototyped code.

Think about where M$ is heading with .NET though. My ideas take this farther in that the compiler esentially becomes a distributed app. As well as even parts of the OS. My ideas are not exactly appropriate for using the language to develope an OS, but that certainly gives me more to think about.

Thanks
Mark

[DynatOS]

I imagine alot of people start seeing the inefficiencies of current Operating Systems and APIs and graudally get disgusted to the point where their own low-level stuff would be better.