OSDev.org

I was wondering what where the worst misconceptions about OSdev that the members of this forum started with?
Mine:
1) It was easy
2) I didn't understand the bootloader idea, I thought that the BIOS just loaded the whole disk (well I didn't actually, I just didn't think about it )
3) I tried to compile my OS code into a windows executable and executed them in windows (invariably causing an memory access out off bands error) (God knows why I did that... )

Oh and my worst misconception when I started programming was writing the source code into a txt file and then changing the extension to exe...
Thankfully my dad quickly showed me what I was doing wrong or I wouldn't know how to code today (God knows what my hobby would be without programming...)
Jules

I thought (when I was 10) that colonel was spelled kernel, and an OS kernel commanded the apps because of it's name.

I thought that a GUI was the biggest most important part, but then I started reading.

-JL

My biggest misconception was that I would have enough time to work on my OS.

I thought i would have my OS done in a year .

I thought I would be able to get a development team started.

Same old thing - I thought it would be easy and didn't realise how much time it would take.

I also thought that Windows/Linux binary compatibility would be easy to implement - never wondered why no-one else had done this yet!

Cheers,
Adam

I thought that it would be really easy to think up a glorious $ABSTRACTION, and once I had, everything else would become part of that $ABSTRACTION.

There is no such one idea. It's just a huge, insufferable mess. Everything is interdependent, and there is no clear order.

I thought the concepts I had in mind were so clear-cut and obvious that it wouldn't take much discussion or persuation to get a couple of people involved who thought likewise.

Ha-ha.

Anyways, the resulting discussion marathon taught me much, so I won't really complain.

I was wondering what where the worst misconceptions about OSdev that the members of this forum started with?

1) All hardware is as good following standards as the VGA (my fav piece of hardware). Only then came A20, broken CPUIDs, funky chipsets, and the rest of the lot.
2) I can leave things out of the design to get started quickly, then add them in later.
3) I was good at assembly...

Solar wrote:I thought the concepts I had in mind were so clear-cut and obvious that it wouldn't take much discussion or persuation to get a couple of people involved who thought likewise.

Hey, I'm using your stuff

Combuster wrote:

Solar wrote:I thought the concepts I had in mind were so clear-cut and obvious that it wouldn't take much discussion or persuation to get a couple of people involved who thought likewise.

Hey, I'm using your stuff

The whole OS project, not the fragment of it that I considered "do-able" for a single person after the OS project desintegrated.

1) That I would be able to read the Linux sourcecode, to get tips on how to do stuff. omfg what an ungodly hacked piece o' crap ....

2) That I would have basic functionality within a year. (Until I read a poll on here about "how long it takes ....")

3) I've been wanting to do this for 20 years, and have been making a list of what I wanted the OS to do, and how I thought it should work. I thought the list would be helpful. 99.7% of the concepts on the list have already been thrown out and replaced with better things, and I've looked at the list maybe 3 times in the last 2 years.

4) I didn't realize that debugging hardware timing problems and other issues would be 500 times harder than debugging software.

My biggest misconception was that the applications for each OS were written differently(e.g. with different bytecodes and instruction encodings). I guess I thought this because it was the only explanation I had of why windows executables didn't work on linux. Needless to say, I soon figured out that applications for most common operating systems area compiled into the same thing as a regular application. Because of this misconception, I wrote my first kernels in ASM and then, when I wanted to use C, I tried to make the compiler output asm because I thought that if I let it produce a regular file, It would be in that weird operating system bytecode. Thinking about it now, it was a pretty weird idea, but I still believed it.

bewing wrote:2) That I would have basic functionality within a year. (Until I read a poll on here about "how long it takes ....").

What do you define as basic functionality?
Jules

That even though lots of hardware and stuff wouldn't be supported yet, that I could begin to use it for my day-to-day OS. Write little apps to play mp3s. Have a basic html browser. Text editor. Etc.

I absolute didn't expect such extreme bugy hardware and backward compatibility crap. I also did not expect that so many hardware we use is a "top secret black box".

In fact I believe Taiwan has the best possibilities to spy on other countries. They can put the spyware in you network card.

World domination for Taiwan!!!