X-OS 0.3.2

aladdin · Post by **aladdin** » Sun May 09, 2004 9:04 am

here is the last version of my OS

The source code :
http://xos.freezee.org/downloads/xossrc_08052004.tar.gz

The floppy image :
http://xos.freezee.org/downloads//xosimg_08052004.tar.gz

please report any bug or mistake ...

DennisCGc · Post by **DennisCGc** » Sun May 09, 2004 1:05 pm

It works well, but you changed the US with the French keyboard layout

aladdin · Post by **aladdin** » Mon May 10, 2004 4:04 am

i'm using a french keyboard becose it's a french OS, but i will add a keymap support soon

Pype.Clicker · Post by **Pype.Clicker** » Mon May 10, 2004 6:01 am

if i may afford a comment, i'm not sure people who wrote all the documents you offer in your "documentation" section will like to see their job presented as if they were bound to X-OS ...

I know you don't intend to rip off anyone, but your page miss appropriate credits, imho ...

aladdin · Post by **aladdin** » Tue May 11, 2004 2:50 am

I'll try to contact authors, but the problem is, that most document are too old, and contact adresses are no more valid ??

Also, there is no copyright on documents.

Candy · Post by **Candy** » Tue May 11, 2004 2:54 am

aladdin wrote: Also, there is no copyright on documents.

there's more copyright on documents than on anything else, check your legal adviser. Documents are the prime point why copyright has been invented (you do know what "copy" stands for right?).

Pype.Clicker · Post by **Pype.Clicker** » Tue May 11, 2004 4:00 am

@candy: i guess what he means is that there's no text like "(C) Tim Robinson, all rights reserved" on the initial thing he provides... but afaik, this is *default* if no copyright notice is written!

Having the permission to mirror a file is one thing, but what tickles me here is that those documents appear to be 'project-related documentation', as if they were written explicitly for X-OS or by cooperators of the X-OS project... A link to the place where the original could be found should at least be offered, and the local copy be explicitly tagged as "mirror copy", imho.

"Rendez ? C?sar ce qui appartient ? C?sar ... (JC)"

Pype.Clicker · Post by **Pype.Clicker** » Tue May 11, 2004 4:16 am

Well, enough legal crap. Let's give the code a try.
- the CPU load remains at 100% while the shell is running. I guess there must be a nasty busyloop somewhere ...
- Nice splashscreen while loading ... on BOCHS, though, it took more time for the splashcreen to appear than for starting the system, so imho, it should be kept for later

- When scrolling the console, strange stuff occur (some characters are randomly replaced by other ones (i'll try to include a snapshot)
- the keyboard map is not yet quite correct. 'M' in qwerty should be ',?' in AZERTY, rather than ';.' iirc. As a result, you don't have '.' available (M is mapped twice!)
- the 'arrows' are wrongly decoded as numbers.

DennisCGc · Post by **DennisCGc** » Tue May 11, 2004 5:53 am

- When scrolling the console, strange stuff occur (some characters are randomly replaced by other ones (i'll try to include a snapshot)

I think this is Bochs its fault, because I had some crazy stuff too in Bochs, but whn I restarted, some just disappeared.

Pype.Clicker · Post by **Pype.Clicker** » Tue May 11, 2004 6:16 am

considering the "weird" code

Code: Select all

k_int0:
     cli
     call idt_int0
     mov al,0x20
     out 0x20,al
     sti
     iret

i'd rather say BOCHS is innocent. the interrupt handler is trashing at least the lower part of eax register ... and possibly more depending on what occurs in idt_int0 ...

moreover, it looks like 'int0' was the 'division by zero' and therefore requires *no* acknowledge to the PIC ... I suggest you read a good tutorial about PIC and "Interrupts & Exceptions" ...

Therefore, whatever you're doing when an IRQ arises is subject to random alteration of computed values ... very ugly (imho)

DennisCGc · Post by **DennisCGc** » Tue May 11, 2004 6:20 am

Hmm, okay, I didn't have a look at the source ::)

aladdin · Post by **aladdin** » Tue May 11, 2004 8:49 am

thank you for your interest.

the strange think about random characters on console is that, when I run X-OS in 80x25 text mode, there is no problem.
the problem apear with 80x40 text mode ??
also, there is, no problem with a real computer...

and about the splash screen, it was a suggestion from some, people who tested XOS and suggested a function that show a logo at startup, u can disable it by editing 32b/kernel32/setup.asm ...

...i'll chech my int handler, and read some docs about PIC

finally ... about documents, i think i'll change the title , or add a comment to precise that the document are not mine, and not about X-OS project

Slasher · Post by **Slasher** » Wed May 12, 2004 5:08 pm

how do you run the programs on the floppy?
I tried everything to no avail.
ps I cant read french anymore :'( (i use to when I was younger, I've forgotten how to now

)

aladdin · Post by **aladdin** » Thu May 13, 2004 4:01 am

how do you run the programs on the floppy?

there is not yet a support for running programms cose i'm working on multitasking...
there is also some probl?ms in the floppy driver, sometimes it dont read some sectors so unformation is bad.

i have a basic support of running programs in 16bit mode (witch is disabled in this version ::) but i will reenable it soon

)

ps I cant read french anymore (i use to when I was younger, I've forgotten how to now )

i'm working on a translation of kernel messages

Pype.Clicker · Post by **Pype.Clicker** » Thu May 13, 2004 6:49 am

actually, it's not surprising that no problem of screwed display arise on a real PC. With CPU@1GHz and IRQ0@100Hz, there are little chance the register trashing occurs during something else than the idle polling loop ...

With a BOCHS@100KHz, things are totally different

OSDev.org

X-OS 0.3.2

X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2

Re:X-OS 0.3.2