Ok, group TPU toy OS.

I'll put forward my ideas (their quite lengthy, and I wanna do a bit of reading on construction of OS's first before I put it together as a piece of writing. I'd like a simple declaration of Yes or No from each of the regulars here whether they'd consider it(hopefully this won't lead to one of the '@' guys saying I don't think it fits in with the long standing goals of TPU, I'd set it up(a couple of bulletin boards and an outline and various parts that could be tackled) on its own site so it wouldn't clog up this one, but still keep it strongly related to this site. Yes, No Anyone?

Sorry, I couldn't follow what you said......

Now, a Lisp operating system would be a nice idea...(oskit anyone?)

OK, I've been fart-arse-ing around on TPU for the past seven weeks.

This is gonna be my contribution to the development of this website.

I want to see how many here will work together on a group operating system.

If you will consider it leave a message saying YES.

If you won't consider it at all leave a message saying NO.

If I get enough YESes I will go away read for a week or two on OS development.

Then I will set up another website, with strong ties to TPU on exactly what it will be, how it will be done, and in what language.(Obviously I won't say "THIS IS GONNA BE WRITTEN IN C WHETHER YOU LIKE IT OR NOT", I'll have a bulletin board where those interested can discuss what exactly we are gonna make)

I don't want suggestions now(not meaning to be rude there), a simple YES/NO count on whether I should spend a couple of hours writing up my exact ideas on a TPU group OS.

A Note for the @Guys. I know you want to see the Interest/Learning groups take off, that is why I would set up the discussion and planning on another site, it wouldn't waste your bandwidth or time. I know quite well that a group project, specifically this OS wouldn't fit into your long standing vision for the future of TPU.

Oh and sorry for always writting unclear messages. Now I'm at home with my own computer I'll get into the habit of writing structured, clear messages.

Thanks
-RDD Regretfuldaydreamer (I'm considering a new idenitiy does "Wimpodude" or "Wimpogimp" sound better(Theres a long story behind those two names))

No.

No.
This has been tried to have been done by people in the past. Perhaps if the 'project groups' or whatever was suggested in the previous thread is implemented, you can set up something like this.

No. (But, then, you didn't expect me to want to participate in it, did you? :P)

You can feel free to use TPU's resources to find people interested in this and advertise the project, but it's not going to be an "official" part of TPU, especially without your being able to provide proof that you will have a full working OS in X amount of time. It really comes down to something like I just mentioned, since EVERY SINGLE outside project that has been officially affiliated with TPU has failed miserably to date, and so it's clear that they just aren't worth it, unless something about their execution changes greatly.

1) Whats been the problem with them and how have they failed?

2) I couldn't have proof I'd have a working OS in X amount of time, it's for learning, rather than commercial uses, therefore there wouldn't be deadlines.

3) Does anyone have any suggestions on how to make a project like this work?

1) AFAIK most projects fail for one of 2 reasons. loss of interest or inability to handle a certain problem.
I don't know there is much that can be done against the first, perhaps gathering a sufficent large group or keeping the project small enough.
The other one can be prevented by picking a project that is suitable to the experience of the group and spending sufficent time on planning.

3) Gather a small group of experienced programmers and compose a design document. See the Quick start guides, or read Code Complete for more details.

Hmmm ...

Ideally, I'd only want about five or six working on it ... too many cooks spoil the broth.

The loose interest thing. That happens to me after about a week of coding on a project. I'll normally regain my interest then for another week after two or three weeks. Maybe if we had some kind of rotation system, we work on it for one or two weeks a month, and then during the other two weeks, we leave our m9inds to wander. More often the than not the solutions to a problem come to me when I'm thinking about something totally unrelated.

You forgot one reason red, lack of time. I start school wednesday, and I'll be working at least 30 hours a week, I'll barely have enough time to eat and use the bathroom let alone program for no reason.

Yeah, school's a big problem for me this year, I've got a fuckload of coursework over the next four months and then I'll have exams all through June and May(GCSE's)

Indeed. I'll have to see how much time I'm going to have to spend on studying next week.

I hate school in a its far too much work way ... if it was simply go and get an hours homework andhave only a few exams, I wouldn't care. They overwork you.

OK, so what about the rotation system, anyone care to diss it or provide an alternative.

I'll do it, as I'm sure Psion would, but only if you write it in SML :-)

Thats 2 .. now we need a couple more.

That why we should have 'project groups' like the learning groups, people organize projects like this and drop in and help out. Kinda like sourceforge.

If you got bored you could just puruse the other projects or something

I would be amazed if there would be a single "project group" that lasted a month in actually doing anything but "planning."

gian, what is it you're sure I'd do?

It would still offer a good point of learning, for people who like more hands on stuff.

mop, you're missing the distinction between people undertaking projects and having special support for those on this web site.

Why not work on an existing OS? The GNU system always needs help...someone should write an audio server for the Hurd, or aid in the port to l4 once the first l4hurd release is made. The TODO list for the Hurd has things that almost anyone can do in it.

Or, if you really have to have your own OS, you could use OSKIT and make an entirely Scheme based multiuser operating system. I _may_ be writing a Scheme compiler soon (I need something to do as a project this year for class).

"what is it you're sure I'd do?"
Participate in this OSdev project if it were written in SML.

You seem to forget that I'm a full time student who is going to take a class in operating systems anyway. :P

No! Have the OS use built in java byte code interpretation!

Or qbasic.

I somehow don't think an OS can be written in QBasic, you can't even compile with that, I tried to compile a 300kb file with QuickBASIC recently and it failed miseraably(The interpretor worked fine though)

Now RedX is starting to speak my language.

Which one of the two of you is the QBasic freak?

I like QBasic but your too obsesive, if your programming skills are actually anything you'll realise how crap QBasic is in reality(I still like playing around with it once in a while though, it only took me about two nights to learn.)

Occasionally I answer stupid questions with stupid answers.
I havn't used Qbasic for years.

me neighter. But it's something fun to mention.

I do suspect that one certain commercial OS is written in qbasic.

Naw, I suspect the one your thinking of is written in GW - BASIC

Back to the OS

I specifially chose an OS for a group proect because

-Very few here will have developed one - so it'll attract a lot of attention - as it'll be new to quite a few

- It sticks out, out of hundreds of possible group projects as one different from anything else.

- The only other thing that I think would make a good group project would be a compiler, which is another story, and we'd have to decide which language to make a compiler for(Need I say any more about compilers!)

- I wouldn't want to do GNU work because the point of this would be to be a beneficial learning project(Sorry if I'm politically correct - Applied Programming Interest Project - please stop the nonsense about trying to decide a new name for the learning groups - I didn't just hint something there, did I? ;)) for all the members of TPU.

It would be a very good learning point, I've always wanted to get into a lot more low level stuff

I think the OS you guys are thinking of is called Windows 95, 98, ME, 2000, XP, and I also think it was written in logo.

No, If it was written in Logo, the graphics might be quicker.

zzzzing!

In
-Mop
-Gian
-Unknown_Lamer

Not In
-Psion
-Neuman
-CodeRed

Not Sure Of:
-RedX

I'd like to know from the following are you in or out:
-FatalDragon
-diegoeskryptic
-CViper
-buzgub
-Faisal - If he turns up again - Id expect help from since he made such a fussy thread about his own OS
-DragonWolf
-Spyder_Y3K
-sphinX
-Linux_Penguin
-AnyoneEB
-Nyiah
-taubz
-metamorphic

I have attempted OS development before -- read a few books, written pages and pages of assembly. I got very frustrated in the end with not being able to find the exact resources that I needed. It's not like learning a programming language where there are dozens of tutorials available. Sometimes you have to struggle with technical documentation that seems incredibly abstract.

I wouldn't mind helping, but perhaps only in an advisory capacity. My life at the moment is far too busy to be devoting time to a programming project that would take even a professional programming team a couple years to complete.

My first suggestion to you is, if you decide to go ahead with this project, plan all the nitty gritty details of the os first -- the big things: microkernel/monolithic, how device drivers will work, how you will organised ring 0 (kernel) and other (user) space. Also, plan down to the little things also, ie, where will you put the stack during the bootload process, where will you load the kernel, should you enter protected mode first and then boot the kernel, or boot the kernel and then enter protected mode? Will you use a 2 stage boot process, or can you do it in one?

All these things need to be thought of before the first line of code is written.

Ok, This'll probably damage my chances of getting this project going but i'll say it any way:

Very few people here have experience in OS development. I have none. I know it'd be far to big to attempt on my own, thats why I want a group with me, so that you and I are not alone in learning. Which brings me onto, once I have enough people(5 or more to be exact, then I will go and read OS Development FAQ's and tutorials, set up the OS's development website , add links to OS FAQ's, everyone in reads them, and then when we have enough collective knowledge; at around I estimate Christmas, then we start planning.

I've robably just signed this projects death warrant by saying this, but I suppose I'd better be honest about it now, rather than later and appear a complete idiot then.

This plan blew up on the launchpad, it's too big, try something smaller, like adding features to an existing OS, as someone already suggested

I would only puruse an OS development group and occasioally contribute, because this is most likely going to go the way of the other OS projects.

Right, I guess it won't taske off, I'll do as you say and contribute to GNU, I'll comeback and suggest it again though, maybe in 8 - 12 months

i'm ready to help, but i don't have too much time to sit down and work (working on another major project for school).

But otherwise, yeah, i'm kinda interested

What? When did I agree to work on this? I merely suggested writing
Scheme operating system and using OSKit if you wrote the OS (OSKit
will make your life much easier because it deals with the boring stuff
people have already done 1000 times before and are probably the parts
that would make you say give up and go home). If you are doing a
Scheme operating system, I may be able to work on it sometime in the
middle of my school year (I am pondering if I should write a Scheme
compiler for my MST Research project this year).

The end of the year just happens to be when I would probably have a
semi-working compiler that I could get my teacher to allow me to
release GPLed (or LGPLed, depending on how I feel at the moment). I
just need to figure out how to do stuff like write a GCC frontend. I
am going to try to implement a simple integer arithmetic engine that
has +, -, *, /, and % with no variables and only one output
function--Disp to display the result of an expression, all using
prefix notation (the point of the exercise is not to make a usable
language but to make a useless language so I can see how to write a
GCC front end without worrying about much else). After that I'll start
with a basic Scheme system that supports only the "core" forms and
compiles Scheme using GCC or a runtime system. I plan to make the
compiler backend changable by using a different 'compile' function
depending on the context, basically the system would expand macros and
then call compile on the new scheme code which would convert the list
structure into something that could be ran or into a GCC tree
structure. The GCC tree backend would be used for static compilation
and the runtime version would be used for runtime calls of eval (I
know I'm a lamer that can't code terribly well, but I have a school
year to implement a working compiler...anyone with half a brain could
implement a basic Scheme compiler in a school year).

Anyway, if you decided to go with a Scheme operating system (although,
I have to be honest, I doubt that you or anyone else here would
actually go through with writing the OS after coming up with an idea
for it; I'm the same way--I get bored too easily), you would need (of
all things) a Scheme compiler. Hey wait, what was I planning to write
again? So, I guess you can put me down as a maybe. If other people say
they are going to write a Scheme operating system, I'll do the Scheme
compiler as my MST Research project. Since I'll be getting graded on
whether or not the compiler works, the compiler will work :-).

So what the hell is this Scheme operating system I keep talking about?
Most operating systems are designed in and for C. I want an operating
system designed for running Scheme or something similar to Scheme
(really more like Scheme on crack). So basically the core would be
written in C using OSKit to speed up development time and it would
support enough of POSIX using OSKit's C library to compile the OSKit
based stuff, GCC, and the Scheme compiler (a GCC frontend, hence the
need for GCC). X would probably be a good idea to use in the short
term, although using something like Fresco or writing a new display
system from the ground up in mostly Scheme would probably be a better
idea for the long run. Being basically-posix would be a good so that
the basic POSIX tools (ls, rm, cp, etc) could be used to quickly get a
working system. I think the long term goal should be to rewrite
basically everything in Scheme. I think that having an operating
system designed from the ground up to be used with Scheme programs
would be great. Since OSKit is a set of *libraries* they can be used
in anyway you want, e.g. to write a microkernel with servers. The real
use I was thinking about was using them to write the core of the OS
and then providing a higher level interface that Scheme programs would
use, allowing basically everything but a few things to be written in
Scheme.


I know no one else agrees with me that a Scheme system would be cool
and now you will all laugh at me and probably light me on fire (I
wrote this the night before I actually posted this so you have all
probably been crushed by the mighty Psion).

RDD: I wouldn't mind joining an OS group, but right now I'm pretty busy... ask after school's out

I said I'd start this up in the new year, I started this thread whn I was blissfully ignorant to how difficult it is to read OS tutorials.

Remember, if you use OSKit (maybe I should tell everyone where it is -- http://www.cs.utah.edu/flux/oskit/ ) you won't have to deal with nearly as many of the low level boring-been-down-1000-times-before stuff. If you want I could grab some stuff from the ACM library (yeah, I'm a loser that joined the ACM because I needed some papers on Lisp for a project last year, I really wish I didn't subscribe to the crap paper journals and just went with the portal...).

As I said the new year.

Oh and between now and the new year, if you wanna be part of it, maybe an idea would be to gather a set of links together that you think will be particurally useful, then when we come together, gathering resources will be quicker

"ask after school's out"

Yeah R, ask in 45 years, after I retire and have enough time to undertake this daunting project, LOL

LOL - Yeah, maybe I'll leave it for 45 years(j/k) - An ideal think to keep me amused in my rich lengthy retirement.

But of course by then there will be no such thing as "Operating Systems"

If Microsoft have their way there will, except it'll be subscriptionware of $1000 a day.

No, by them MS will own the entire computer industry, they will outlaw building your own, and the only ones for sale will be made by MS, who, each year, will realese a processor 10mhz faster than the last and charge $200 for it, and if you don't buy it, your computer stops working.

Now we musn't get too pesemistic ... There'll always be Apple ... They're just starting to follow Microsofts trends, and they aim to double their market share in the next year or two.

Either that or linux will be adopted by the programmers, and the Mac OS will be adopted by the people who want to quickly check e-mail and do some work without partitioning something first and windows will be vanquished.

Then poor Bill Gates will finally go backrupt in 900 years.

Hmmm I've just gone the way of the unknown_lamer: I'm a pure Linux user now, completely free software. This is going to take some getting used to.

Once you do get used to it you'll never go back.

Woohoo! Free Software is great :) I just have to resist the urge to get doom 3 when it comes out (I'll play it on a friend's machine! My 500Mhz k6-2 and voodoo3 won't be able to play it anyway).

Well this is sorta nifty. I skimmed over this thread a couple of times, but only just now read what it's really all about.. Good thing I won't have to bother with an answer, as I'm not in the list. :o)

"and they aim to double their market share "
What are they planning? To clone their current users?

You see they only have a small share of the market, say 5%, so to double their market share they need to get 10% of it.

This will be a minor graze on Microsofts had, whereas for Apple, they will seem upgraded from a crapped out old banger of a car to a luxury private jet.

And for all you big Billy Haters, hes trying to downgrade his role in Microsoft and become more of a seen but not heard type, he feels he's more of a liability to the company now than an asset.

I don't care how big or evil of a corporation Microsoft is I would still take a windows based PC over a macintosh ANY DAY. Linux is alright, but I game, so I said f*ck it

I don't hate Microsoft cause they're big, I respect them for that, it's the methods(Heroin Economiccs) that I don't like, and lets face it, would you do what Billy did for $90,000,000,000?

I think so.

Still haven't forgiven ms for buying out bungie and releasing halo this far only on the xbox (supposed to come out 2003 for pc - yeah right; thats what they said 2000, 2001 & this year too).

Halo is a decent game, not much replay value though, very repetitive

I have a PS2, I've played it once, I haven't bought a memory card yet, it came with MGS2

Thats the same with most bungie games. still i like them (Marathon, anybody?)

btw: they already anounced Halo 2 [for the xbox]... *grrr*

I've played all the systems, but own none, and don't plan too, as it is, my PC still has the capability to play more complex games than their's, if any new ones come out anytime soon dammit. Marrowind is really cool, UT2003 is just about the only game that will push my system to the limit, if I ever get it (damn you EOE, you promised by the end of the month). My system stats (updated since last time I posted them)

1gig AMD Athlon (Overclocked to 1.3) on ABIT KR7A-Raid
512mb PC3200 DDR (Corsair XMS)
2x 20gb WD 7200rpm HDD's in Raid-0 Array
GeForce 4 TI4600 (Overclocked)
Sound System: 7.1 channel 500 watt Dolby Digital/DTS
---Mains: 2x 200w Sony 3 way towers
---Surrounds: 4x 150w 2 way bookshelves
---Center Channel: 1x 120w 3 way
---Subwoofer: 1x 200w 18" driver

As you can see, I spent the majority of my money on my sound system, not my computer, priorities... LOL

I get to use OS X (for now it is 10.1 until the magic upgrade elf makes 10.2 magically appear without the school paying for it) at school and it rocks. I like to run XDarwin because then I can use emacs, a bunch of xterms, and gaim under X (or really any other X app). XDarwin is rootless so my X and Aqua windows merge perfectly together and the X windows get shadows! I think it might be all surfaces get shadows because the menus in emacs have shadows, which rocks. Too bad it is an evil evil proprietary system! (hey, I use proprietary when I have no other choice, and OS X is better than Windows).

Right now I am using the AstonXP GUI for windows XP, Best shell I have ever seen