OS Development Time

[Gaz]

Ok, first post on here so a Big HI to everyone. Excellent forum on OS dev, been reading posts for quite some time.
Anyhow, onto the main question...
For my final year college project, I am deciding to produce a small OS, one capable of not a lot generally. Would I really have enough time to create it in say, 32 weeks? This being said, average working time on the OS would be about 2 hours a day for the 32 weeks?
Would this be sufficent or am I better off saving this for a personal project?

Thanks
Gaz

[Tom]

I don't think it'll be enough time...I program my FritzOS two hours a day, and I don't get much done...due to tests and v86 mode

[Tom]

if you really really work at it, you mabe could.

[frank]

it depends on:

- how you code + coding experience
(For example, I didn't know how to seprate a word into 2 bytes and that took me some time to find out)
- what your os is going to support (is it going to have a filesystem? does it load a kernel or is it only a bootsector? does it support harddisk/floppydrives etc)

Best,
Frank

[Gaz]

When I say small, I mean small. Basic printf function, basic keyboard input etc etc ect. Yes it would load a kernel, no it wouldnt have its own filesystem, for now a fat file system on the bootsector. No need for hardrive or floppy disk support within the OS itself.
Really you can say its not a fully featured OS, not compared to some people create. But this is merely for a college project where time is limited.. I hope to take it onto University where features such as floppy/hard drive support, its own file system etc etc would be implemented.

I might just give it a shot, just for the experience really, and the learning value

Gaz

[anubis]

Hi
I am also trying to write mine... But i have a college time of just 12weeks, with a preety tough schedules and exams. But still i managed to write basic keyboards routines, display routines, paging(Protected mode) and supporting FAT12 from floppy and a bootloader.
This work took 8 weeks with about a hour or half a day coding, no more. Hoping to complete some more in the left 4 weeks....

;D

[Pype.Clicker]

a generic rule: do always assume that it's going to take twice the time you thought it was about to take

[crazybuddha]

a generic rule: do always assume that it's going to take twice the time you thought it was about to take

As ever, pype is dead on target. This is a reliable method of estimating software development generally. Once you've broken down the problem and made reasonable estimates on accomplishing each task, multiply it all by two and you will be pretty close to the actual time. At least, this has been consistently true in my experience (and observation of others).