OSDev.org

Name: Sendla Operating system

Team: Motron software

Link: http://motronsoftware.com

Description:

Sendla OS is an os specifically designed to have major stability, it is also designed to be able to run: Windows programs, Mac os programs, Linux programs, DOS programs, Beos programs, Unix Programs and many more. Sendla OS is designed to have major security systems running and an included antispyware engine, We are also making 3 modes with it ( Begginner, Advanced and Command) The beginner mode will help new users check their email, play a game of cards, and maybe browse a few websites, and write a document, Whilst the advanced mode will have administrive options to change, add drivers and install new programs, along with the command mode which will be like the MS DOS prompt in windows. The beginner mode will also feature a jargon buster to help beginners with error messages. These are what we have planned for the os, but for now we are mainly focusing on the kernel. If you are interested in developing with us, or have any queries, please contact me via MSN messenger on: [email protected] or [email protected] ,

If you have anymore suggestions for sendla os we'd love to hear from you.

-stevenup7002

___________

I'd just like to announce that the jobs are still open and anyone who wants to join just post or email or IM me!

funny =)

most people help them who help themselves,especially if they dont know them

we bought a domain name:

http://motronsoftware.com
http://sendla.motronsoftware.com
http://mg.motronsoftware.com
[email protected]

!!

sorry but i really think that there can be nothing error proofed, i mean think about it if a person messes up in application code you have to do something about it or as some illeterates would think it is crashing because you have to terminate it
its not if it crashes, its when it crashes

and thats a whole lota support for a newbie os dever

dream big but think small

Nothing can be crash proof but if it does crash it will only be the last thing, there will be a sequence somthing like this

1. Fatal error
2. Notify user
3. Search for fixes
4. Fix if found

1. Fatal error
2. Notify user
3. Search for fixes
4. Give warning to user about error
5. Stop the command from running
6. If very fatal, dump unsaved data to a file.
7. Close the application (When the app restarts you can reload what you had)
8. If Hardware cannot be run due to error reboot.
9. On mode select screen in the next session, it will say: The machine was rebooted due to a Fatal error, we are sorry for any inconvienience.

here would be what i would think
1. report exception and give user a chance to end or ignore
2. if they did not pick end then using error codes and what is pushed on the stack attempt to skip or fix the error
3. if it still cant be fixed post a fatal error and end it

i trully dont see how an os can crash especially if applications and drivers are in ring 3 but however if it was in ring 0 and did something like disable ints or changed something around like the idt then yea i could see how it would crash but I dont see how windows has a bluescreen instead of just rebooting(yes i know sometimes it does just reboot)

actually, windowsXP will reboot automatically, but only AFTER saving a memory dump the the HDD(if posible) and preparing a complete error report

its actually quite easy to stop it from rebooting on error:
just don't allow it to triple-fault

if you have a valid idt, with proper double-fault handler, it will never tripple-fault
btw: something alot of people don't realize: it won't reset on all computers -- some computers will simply freeze on tripple-fault (the chipset controlls what happens, not the CPU)

windowsXP simply checks the exceptions and for fatal crashes (unrecoverable) it drops back to a 'safe' program which saves the memory and error conditions and creates a error report and then reboots the computer (using the keyboard controller), after resuming it analizes the report, and then ask if you want to send the relevent portions to MS for analysis

on older OSs it would (instead of rebooting) show a simplified portion of the crash report onscreen, but MS realised most people don't need to see that (and wouldn't even understand it) so they set it to (by default - which can be changed) perform the only action posible after a critical error: reboot

Sendla OS will be released under GPL when we get a bug fixed, thus making it open source

thats cool

and why do i never get a double fault and i know my idt works because the only difference between my double fault and triple fault is that they say different messages

which actually i do error handling incorrect because on a triple fault the stack or even gdt may be corrupted so how am i going to call printf

Hello,

I would like to help out with the OS. I have been learning ASM fokr about half a year now, and I have been programming in C since I was eight. I think that I could set up some programs for the OS and make it compatible with other processors other than the x86 architecture. I think that I would also be able to set up a rough GUI.