OSDev.org

@Combuster: You're correct about Mode-X, reading the documentation once-more I got to know that Mode-X works on all VGA Cards, but Unreal mode does cause problem for Cyrix Processors.

Are you seriously going to pick up on RDOS' worst traits?

Lol.

Combuster's Signature wrote:"After you finish the first 90% of a project, you have to finish the other 90%." - Michael Abrash

Michael Abrash

ModeX

If I remember correctly didn't he discover ModeX?

sid123 wrote:but Unreal mode does cause problem for Cyrix Processors.

I found something that gets close, but it isn't actually unreal mode nor allowed according to the manuals. Therefore, citation needed.

If I remember correctly didn't he discover ModeX?

What Abrash is missing in hacking skills of the bigger pros he makes up in being able to write about it. Mode X is not his invention as far as I know, but he is the one who wrote the first known book on it. (If you read his work more, you'll see he credits ideas and help on several other people in his writing. John Carmack is one of the few actually mentioned by name, although I don't think he's the original source for Mode X either)

And I have no idea how many people individually invented unchained modes back in that day, because it's a bit of the logical consequence of experimenting with attempts to fix the memory issue in mode 13. I could say that I "invented" undocumented modes of my own as well, because I have no idea if someone else has had that same eureka moment somewhere else.

Combuster wrote:

sid123 wrote:This is probably the largest problem with undocumented stuff like Unreal Mode, ModeX, and other hacks, that you can't expect reliable behaviour

Are you seriously going to pick up on RDOS' worst traits?

RDOS doesn't support Mode-X, never had and never will.

sid123 wrote:And please don't blame me for that huge whitespace around the image, I am a MS-Paint guy , GIMP's quite complicated though

For an MS Paint replacement, try Kolourpaint if you don't mind using a KDE program.

Kevin wrote:(...) For an MS Paint replacement, try Kolourpaint if you don't mind using a KDE program.

If you mind using a KDE application, use Pinta. It's also better than Windows Paint, which simply is a piece of sh*t, nothing more.

It crashed the emulator itself
IDK how did it get that loop.
Learnt a new lesson in my life. Wanted to share it.

lol, Do I know the difference between scrollin' and destroying the screen
Didn't expect such colorful output

Can't tell what's wrong but I think I'm reading past the amount of memory I have available, sadly I'm not really even sure which mode I'm in right now, I have a GDT and an IDT setup, last time I tried using BIOS interrupts to set the cursor position, things failed so I think I'm in protected mode, but anyway, this is what happens:

I can't seem to figure out why it might be calling interrupt 6 though. I'll try to figure it out before actually making a topic for it here.

hgoel0974 wrote:(...) I think I'm in protected mode (...)

You think you're in protected mode? You think? If you're using GRUB, you definitely are in protected mode (if you haven't switched back to real mode, which is, AFAIK, not that easy).

no92 wrote:

hgoel0974 wrote:(...) I think I'm in protected mode (...)

You think you're in protected mode? You think? If you're using GRUB, you definitely are in protected mode (if you haven't switched back to real mode, which is, AFAIK, not that easy).

lol thought so, I was sure until I found out about how GRUB does certain things differently for linux, thanks for confirming it for me though

Well, someone really doesn't want me to write a floppy driver

Good News: Interrupts, Exceptions work well
Bad News: lol.

Me trying to port netsurf,

I uncovered a few design flaws in my stdio implementation when I added fmemopen() so I set out to correct them. Naturally, small errors in stdio can lead to great wtf moments:



That's just me booting my system and typing the ls command. It turned out that I was flushing a bit more than just the intended buffer fragment.

I also ran into a few problems when I was adding thread-local storage support. Turns out init gets a bit touchy if errno doesn't work properly:

I believe I might have forgotten to implement, well, ANYTHING related to usermode before attempting to switch into usermode (no updated GDT, or TSS's). The product of my stupidity. At least it looks decent, somewhat.

SSE2 Alpha Blending gone wrong, back to debugging. Atleast I got my MMX Alpha Blending going correctly :p