OSDev.org

That and they for some reason demand that you use COM interfaces to build a kernel. Can you say overkill?

This is picking up based on votes. I'm in "not interested", but the yes has actually picked up. I'm DEFINATLY keeping an eye on this.

That and they for some reason demand that you use COM interfaces to build a kernel. Can you say overkill?

I dislike the programming environment of Window$. It's sloppy at best. I only write window$ stuff in VB, because there's not much to worry about .

Cheers, DH.

Hello....

@Brenden : I think people interseted in this project should send there message here itself. Hence other can also know the project team. It would be more private to send these things in email.
I will start the introduction phase...

Member ID : 002 (001 will be Brenden)
Name : Viral Patel
Current Status : Student
Current Project : TAJ (http://www.geocities.com/taj_os)
Interested fields: Device Drivers, GUI

Hi,

viral wrote:@Brenden : I think people interseted in this project should send there message here itself. Hence other can also know the project team. It would be more private to send these things in email.
I will start the introduction phase...

Currently, 2 other people have also expressed an interest by email. I have sent both of them a list of several suggestions for things that they could choose to work on (with a note saying that if they don't like anything listed they can let me know and we'll sort out something else).

I will also send this initial list of suggestions to people who express their interest here using the personal messages system.

@Viral: I tried sending it three times but I got a time-out the first 2 times and didn't know if it was sent or not. My apologies if you've got 3 copies of it...

From past experience (and the experiences of others), it's likely that some of the people who initially express interest won't actually contribute anything. This is common - the same thing can be seen for most projects. In this case it's better for these people if the original expression of interest is private, so that they can change their mind without consequence.

Hopefully everyone who expresses interest will find something they'd like to do. As soon as anyone lets me know what they'd like to do I'll put together a "Developers Page" on my web site showing which tasks are still available and who has been assigned to each. When a developer completes a task they will be added to the credits page and a good description of what they've done will be put on the Developers Page (so that if "Fred" does a lot of work he'd have a long list of things under his name on the developers page, for example).


Thanks,

Brendan

Just for the records, will this actually be running under the moniker "Mega-Tokyo Community OS"? I don't think that would be a good idea, as the votes imply it would represent less than 30% of "us"...

Solar wrote: Just for the records, will this actually be running under the moniker "Mega-Tokyo Community OS"? I don't think that would be a good idea, as the votes imply it would represent less than 30% of "us"...

I'd say the votes imply it represents at least 30% of us, and that are just the people that trust it to go well. I for instance voted "No, not gonna work" while in fact I do support a project that tries to make it work. I just doubt that it'll work.

It would at least be a positive thing for them to have some form of a name to link it to Mega-Tokyo, although I would say that calling it THE mega-tokyo OS would be wrong.

Well, I voted "not interested", but I wouldn't mind if they would be using my PDCLib some day. So, in a way, I'm working towards them, too.

My point is, how many people came here, and posted their ideas, and got help in setting up their OS project - and why is it this project that is so special that it will be "the" or "a" Mega-Tokyo OS?

I fear two things:

1) That this forum and the Wiki starts sliding towards supporting this project specifically instead of providing generic information (e.g., a bias towards this OS in examples);

2) that this forum and the Wiki will be judged by the merits of this specific OS (and as we all know, the majority of OS projects fails miserably or has very little to show after years of work).

Call me doomsayer, I just get into position for a "told you so".

Solar wrote: 1) That this forum and the Wiki starts sliding towards supporting this project specifically instead of providing generic information (e.g., a bias towards this OS in examples);

2) that this forum and the Wiki will be judged by the merits of this specific OS (and as we all know, the majority of OS projects fails miserably or has very little to show after years of work).

Which is why I'd rather not see it as "THE" MT OS. If it just refers to Mega-tokyo as place of development but under its own name, I think this would not happen as quickly.

Call me doomsayer, I just get into position for a "told you so".

I'd rather ensure those things don't happen than stand over a pile of rubble saying "told you so". I'm not going to vote for you as advisor of any country with atomic missiles.

Candy wrote: I'd rather ensure those things don't happen than stand over a pile of rubble saying "told you so".

I'd rather, too. Call me desillusioned - I've been through the worst with the Amiga "community", and since have cut back on my hopes of anybody listening to my rants.

I'm not going to vote for you as advisor of any country with atomic missiles.

A relief - I wouldn't want that job anyway.

Hi,

Solar wrote:Well, I voted "not interested", but I wouldn't mind if they would be using my PDCLib some day. So, in a way, I'm working towards them, too.

Ironically, I haven't voted at all.

Solar wrote:I fear two things:

1) That this forum and the Wiki starts sliding towards supporting this project specifically instead of providing generic information (e.g., a bias towards this OS in examples);

2) that this forum and the Wiki will be judged by the merits of this specific OS (and as we all know, the majority of OS projects fails miserably or has very little to show after years of work).

I would hope that the forum and Wiki are used as a valuable source of information, and that members of the OS project contribute to these resources in the same manner as always - occasionally adding some (OS neutral) information to the Wiki (like descriptions of hardware, etc), occasionally seeking answers or suggestions in the forums, and occasionally helping others in the forums.

IMHO the project should be "A Community OS", rather than "A Mega-Tokyo OS" or "The Mega-Tokyo OS". There are several reasons for this, including Solar's concerns.

Hopefully, eventually, the OS will grow to the point where we need people from many other areas - applications programmers, graphics artists, games programmers, compiler developers, web/java developers, system administrators, etc. These things are outside the scope of the The Mega-Tokyo OS Developer's forum.

Additionally, there has already been a bit of interest in this project from communities outside of Mega-Tokyo. IMHO the project can benefit from the experiences and diversity of other people and other communities, and shouldn't really be restricted to Mega-Tokyo only (even though Mega-Tokyo will always be a focal point).

Anyway, comments and suggestions on this welcome...


Cheers,

Brendan

Hello..
This is for people who are in this project... Even if it looks awk, we sould discuss over platforms:

I use Windows XP for OSDev. Textpad as editor, DJGPP as compiler. We should have something common, so that we all can compile new OS on our system...

@Brenden: Is there something you wanna say (like Red Hat 7 Linux, gcc... ld... vi... etc) ?

Hello....
Sorry I just forget to add something, so i initiated new post.

We should have some common commenting techniques. I suggest something like Java. Like:

/**
* @author: Mahatma Gandhi
* @date: 01-01-2006
* @arg: pointer to source & dest strings
* @return: 0 on Success
*/
int strcpy(char *str1, char *str2)
{
   int i;
   for(i=0; str2 != '\0'; i++) //loop till end of str2
    str1 = str2;

    str1 = '\0'; //embed null

}

This will be useful when we will create some documentation. Like in java we use 'javadoc". Its not compulsory that we should have this only.. anything that we all agree.....

Hi,

viral wrote: Hello..
This is for people who are in this project... Even if it looks awk, we sould discuss over platforms:

I use Windows XP for OSDev. Textpad as editor, DJGPP as compiler. We should have something common, so that we all can compile new OS on our system...

@Brenden: Is there something you wanna say (like Red Hat 7 Linux, gcc... ld... vi... etc) ?

Currently, all source code is written for NASM (and YASM for 64 bit stuff), and these should work on all platforms (and with any text editor - I've never had problems with NASM and CR/LF). At the moment most of the code is in real mode or some strange mixture of real mode and protected mode (or long mode), so most C compilers would have difficulty generating anything to suit.

Eventually (when it supports normal processes - device drivers, applications, etc) almost all tools will be able to create binaries to suit it. In this case you'd be able to use anything you like and include the resulting binary into the boot image, but I'd prefer open source tools where possible. It'll probably take 6 months or so to get that far though.

viral wrote:We should have some common commenting techniques. I suggest something like Java. Like:

I'm thinking that when the boot code and kernels are ready people can have total control of their own binaries (either one person per binary or one group of people per binary). Each binary can use completely different tools, completely different languages and completely different commenting.

For example, you might write a "spell checker module" in C++, someone else might write a module for syntax highlighting in Pascal and a third person might write a text editor in Forth. These seperate modules would use the messaging system at run time to work together and create a text editor with spell checking and syntax highlighting. As long as the messaging interface for each module is well documented no-one will really need to see any of the source code for other binaries/modules.

Apart from the boot code and micro-kernels, the entire system can be done like this.


Cheers,

Brendan

Hi,

CopperMan wrote:Making OS together it's a good idea.

Ok, Brendan, I have read your proposal and I think you'll be a good project leader!

Thanks, but whether I make a good leader or not is something that none of us will know for a while....

CopperMan wrote:Your changes to proposal (attachment): d:\my_proposal.txt

... and ...

And after that we can start to design an OS.

This may be one of the most common misconceptions about my proposal. The idea is to start with my existing design, rather than spending 6 months trying to decide if "Hello World" should be followed by a full stop or an exclamation mark.

This means that most of the fundamental design work is already done - it was started some 10 years ago. There's already been many different implementations of this design, with each implementation being more ambitious and with the design improving along the way.

This also means that we can start from the code I've already done, which is mostly just boot code. This code supports hyper-threading, SMP and NUMA, will setup the computer for a 32 bit kernel (plain paging), 36 bit kernel (PAE paging) and/or a 64 bit kernel (long mode), and is designed to be modular (where people can add or remove "boot modules" that are run by the boot manager before the kernel starts running).

CopperMan wrote:It'll be good if you make a page (at your or at mega-tokyo web-site) for registering people.

Currently we are at the beginning of a "transitional" stage. When people say they'd like to become part of the project I send them a list of things that they could choose to work on in the hope that they will either choose something listed or negotiate something else.

If any of these people actually do start working on anything, I will begin converting the existing stuff into a community oriented project. This includes creating a Developer's Page listing details of people who are doing something, what they are working on and a list of anything they have contributed. The Developer's Page will also list jobs that can be done and be used to co-ordinate things. This transition will also include changing the licencing on existing source code, creating instructions for building the project from source (list of tools, etc), setting up something for developers to discuss things, expanding the developers documentation, etc.

If this doesn't happen (i.e. if everyone says they want to help but never actually do anything), then the project won't be converted into a community project and will continue as it is now.

CopperMan wrote:Making OS together it's a good idea.

I guess that is the main point - making an OS together is a good idea, but several people trying to go in several different directions at the same time will never work.


Cheers,

Brendan

It seems to me that this thread has reached the point where it should leave place to another one. Whether it's a GoodIdea or not has now been quite extensively discussed ... It appears you guys want to do the stuff. You're welcome

I suggest people that want to contribute "stick" that on their MT profile (such as "MTOS" keyword tagged in your signature, on your avatar, whatever). Feel free to open a new wiki page to log "important" threads in the design that you could have in the forum, or to list names of contributors/supporters.

http://www.osdev.org/osfaq2/index.php/OsdevingTogether

Keep in mind, however, that the FAQ altogether is an open place and that it should remain full of useful hints even for people not being part of the OsdevingTogether effort.

I suggest once everyone know the contact info of everyone else that you settle a IRC session (or something alike, but imho a thread would be less appropriate) to accomodate on the bare design guidelines, e.g.
- what language to use, what language might be used, what language will not be used.
- what bootloader to use (e.g. grub or not grub ?)
- what license do you agree on (e.g. GPL, BSD, public domain ?)
- micro ? monolithic ? or something else ?

Ranking against each other is probably not a good idea, though something like Sourceforge skills profile (adapted to your needs) might be useful... even mine appears over-ranked to me nowadays: i'm certainly not a wizard in OO programming, for instance ... merely competent.

But discussing this (skillz rating) here is pointless, isn't it ? so let's drop that off.