Is Your OS Based On Another?

[xxchrisxx]

How many of you have based your OS on something that exists already? For example, Darwin (based on Mach and FreeBSD).

How many are doing something new, that hasn't been done before?

[Solar]

Pro-POS is a "from scratch, no legacy" approach, i.e. not based on another.

[Tim]

Mobius doesn't try to copy anything, but my design has been influenced by features I like in other operating system.

[Whatever5k]

I am really trying to make Flick being independable from any other OS. It is actually working until now.

[Solar]

Reading Tim's reply, I'd like to add that, of course, Pro-POS is also influenced by past experiences, but we just take inspiration from other OS', not code.

[BI lazy]

Well, BlueIllusion is microkernel based design. And since I have THE Tanenbaum book in my Claws, Minix has kinda influence on how I do some things.

It's as you say: Inspiration, but not code.

[Pype.Clicker]

Clicker's microkernel is built 100% from scratch, but some drivers (IDE disks, RTL network card etc.) are somehow based on Tim's job on The Mobius (with his agreement, last time i asked him )

I might import some VBE and VM86 stuff from BlueIllusion (with the agreement of Beyond Infinity, if my memory is still refreshed often enough)

Some Ultra-DMA drivers and other PCI work could be imported and translated stuff from Linux ... I see no reason to type every character myself, and i kinda failed to create a Clicker Development Team community ... so my current approach is to check things that are avl. in other hobby OS and ask their maintainer if i can adapt their work to the Clicker environment

The module-based approach makes it simpler to keep a track of what is based on who's job

[Schol-R-LEA]

The 'warm-up' system I am (failing to) work on, LoIS, does not specifically mimic any existing OS, though it is a relatively conventional one in most regards. For anumber of reasons, this has largely stalled, and with it my whole development process.

My planned second system, ENOS, not merely copies GNU EMACS but is in fact built upon it as the main system core; the name stands for EMACS Native Operating System, and that's precisely what it will be if I ever get around to it. It is meant as something of a proof-of-concept rather than a truly complete system, though with EMACS as it's user interface, it should be fairly usable. Obviously, it will be a GPL'ed system, and I'd turn code rights over to the FSF once I've gotten it minimally working.

My long-term goal, currently code-named Thelema, is a pretty radical departure from most existing systems, both internally and externally. The chances of me getting anywhere with it are slim, however.

[pini]

Mine is from scratch too (just as my linux ) but I admit I get inspired by some features in other OSes

[BI lazy]

*pypes memory refresher*: that's ok.

since I 'm now busy with a project at work (somehow even a geek has to earn his living), I lack time to translate some of the german identifiers to english. Damn. One should stick to standards if doing an OS instead of sitting on old behaviours like a fat toad in the cold.

stay safe and don't worry.

[Curufir]

Well like everyone (Nobody just woke up one day and wrote their OS...did you? ) I've look at ideas from existing OS where I can, but generally only as information about what has already been done, how other people solve these problems, what is possible etc.

I wouldn't say that my effort is based on any other OS. The real thing shaping it at the moment is my target platform and use of assembly language (C would be more efficient, but I'm in no rush and I like assembly language). Target platform is 386 with a vga compatible card and 4mb of ram. That hardware limitation (I have my reasons, the primary one being that my target userbase doesn't have access to modern hardware) colours everything I do because it imposes seriously scarce resources (My one justification for using assembly ;D).

I think that the programming language you use affects the coding thought process (This is actually a real phenomena with spoken/written languages when it comes to problem solving), even with closely related languages. Someone programming in C++ is likely to approach a problem differently than someone programming in C, and their solutions may well be different as well.

[mr. xsism]

heck, i'm not gonnna lie about it. i look over as much code as possible. i like to see how other people have approached problems that i have had. I don't copy their code, but i do assimilate basic ideas and methods into my code. An example is how people handle paging. Is it one big function or a few more spcific smaller functions.

If you ever take anyones code you should give credit. It shows your uberness and it is kind

mr. x

[Karig]

I am starting to work on my own OS (also called Karig ), and I plan on using Chuck Moore's colorForth compiler code for loading and running programs. Other than that, I expect that the bulk of my code will be original with me.

Full explanation on the Web: "What I Want" at http://www.karig.net/0005.html

[mr. xsism]

kool site. Did you get that name from StarCraft?

ColorForth, i've heard OF it but never looked at it. i'll look at it soon. Could you debrief me on it ?

mr. xsism

[Karig]

Gotta make this reply quick coz i have to go to work in about four minutes.

"Karig" is Dutch for "small." I just found the name in a language dictionary and liked it.

ColorForth is both a programming language and a programming environment that works a bit like an OS. It displays words in different colors according to the word's function -- one color to execute the word at runtime, another to execute the word at compile-time (usually a word that adds machine code to the end of the program being compiled), and so on. (A variation on colorForth could use fonts or even punctuation instead of color.) Chuck Moore intended colorForth to be the simplest, most efficient language he could make. Link: http://www.colorforth.com/cf.html

Gotta run.