Kill the Wiki

[Omega]

Sorry, I was just rambling on there. Really into CS at the moment, kind of kills a lot of my free time; it is a fun break from coding. Just thought others might do the same. Sorry to kind of hijack my own thread, disregard it if you must. And, the STFW question was a serious question, it was used in this thread and I am not too *hip* on Internet lingo I suppose and too lazy and disinterested to google it. I just assumed it is like STFU; I know it can't be good whatever it is. haha

Alright, I'm off to go shoot people. Have a nice day.

[EDIT]OK, I googled it: Search The Freaking Web. That's mean. lol[/EDIT]

[Jeko]

I think the wiki mustn't be a simple tutorial. It must be a container of resources abount operating system development. Of course in the wiki there can be also a simple tutorial, but it will be a page, or a section, but not the entire wiki.

[Mr.Confuzed]

I agree that the wiki needs more tutorial content. I'm not sure if it is possible, but a tutorial namespace could be useful, or at least a tutorial category.

vst_0201 describes a massive, structured tutorial covering OSDev with detailed examples, like a course. This would be AMAZING for those of us who want to learn how write an OS. There's a problem, though. Everyone who works through the course will learn how to write the same OS. If you're here to write an OS, chances are you want it to be your OS, not the same thing as everyone else.

I propose an alternative: We should continue adding tutorial and informative content to the wiki, but attempt to give it a backbone structure. We could do this by adding a multiple page guide. For example:

Start Here
01. What is OS Development?
02. Setting up the Working Environment
03. Media Resources

Bootstrapping
01. Booting with GRUB
02. Rolling your Own

Kernel Development
01. Setting up the Environment
02. Memory Management
03. Writing Generic Drivers

Debugging
01. Dumping the Symbols
02. Debugging with Bochs

References
01. Opcodes
02. Scancodes
03. VGA Colors
04. Intel Manual
05. Ralph Brown's INT

The guide would have to be theoretical, but would link to more specific pages, which would probably link to tutorials or examples detailing exactly how to implement that method, should you choose it. If we link to multiple examples, then the user is given choices, which encourages thinking about the topic, and may even result in a different approach altogether. IMO, this will prove more useful and intuitive to the entire OSDev community.

In an attempt to avoid RTFM and STFW threads, how about including a short introduction on the main page - which links into said guide - followed by the index format we currently have, for quick reference.

Requesting opinions,
Matt

[Omega]

Just give up, because they don't understand this concept. They think the wiki is the best thing they ever saw, they love it, and it is not going to go anywhere. It will continue to get bloated and more and more useless over time, but they don't see that right now though they have been told.

Chase you should change the slogan man, because you know good and well that this is not the place to start. All the lame Internet abbreviations being used like STFW and RTFM do not support your campaign. The slogan should be, "The place to go when there are no other alternatives.", because that's what it really is. If you never programmed in ASM or C before and you still want to try out OS Dev, then why should you be turned away? If they say, "Let me see code", and if you have it, then why not reveal it? This method works well for web developers who frequently post fully working code with even examples on how to use it. It is called Learning by Example. You see; you get the code, compile it, test it, learn it, modify it, and implement it. That is how a lot of people learn and I myself am one of those people so it kind of pisses me off when I read threads where someone has asked for code and they get flamed; its kind of weak in my opinion and I just don't like that sort of 'can't-do' attitude.

Now there are a lot of great coders here and even more great topics with helpful information, I am sure a guide as I have proposed in this thread could easily be made in a short amount of time. But, you don't want to hear that because it means you have made a mistake and have to start again, and who with an ego will ever admit he is wrong and which one of you here with authority are without ego?

[chase]

Now there are a lot of great coders here and even more great topics with helpful information, I am sure a guide as I have proposed in this thread could easily be made in a short amount of time. But, you don't want to hear that because it means you have made a mistake and have to start again, and who with an ego will ever admit he is wrong and which one of you here with authority are without ego?

You see that the wiki should be one thing and other people see that it should be what it already is. Both groups are right because there is a need for both structures. I already bought this idea up with the moderators befor this thread was ever started. So it's not that we are not hearing you, it's that you aren't saying anything that we don't already know. It will happen at some point but it's going to take a while. You are wrong when you say that something like this can be done in a short amount of time. I am a professional programming instructor for Sun and I've written several of their courses. I know exactly how much work it is to undertake something like this. When we start this it isn't going to me some small little tutorial, it's going to be something this size of a decent book. If you think community book writing is easy just look at how well everyone in this thread agrees. We have not made a mistake with the current wiki, it will continue to exist and be worked on even after we start working on the book-like guide.

You are wrong to think that your way is the only to do things. I've been teaching for 7 years, I know how people learn and that not everybody learns the same. Do we have everyones learning style covered yet? No. However this still is the best place to start for OS developers.

[Jeko]

Just give up, because they don't understand this concept. They think the wiki is the best thing they ever saw, they love it, and it is not going to go anywhere. It will continue to get bloated and more and more useless over time, but they don't see that right now though they have been told.

Chase you should change the slogan man, because you know good and well that this is not the place to start. All the lame Internet abbreviations being used like STFW and RTFM do not support your campaign. The slogan should be, "The place to go when there are no other alternatives.", because that's what it really is. If you never programmed in ASM or C before and you still want to try out OS Dev, then why should you be turned away? If they say, "Let me see code", and if you have it, then why not reveal it? This method works well for web developers who frequently post fully working code with even examples on how to use it. It is called Learning by Example. You see; you get the code, compile it, test it, learn it, modify it, and implement it. That is how a lot of people learn and I myself am one of those people so it kind of pisses me off when I read threads where someone has asked for code and they get flamed; its kind of weak in my opinion and I just don't like that sort of 'can't-do' attitude.

Now there are a lot of great coders here and even more great topics with helpful information, I am sure a guide as I have proposed in this thread could easily be made in a short amount of time. But, you don't want to hear that because it means you have made a mistake and have to start again, and who with an ego will ever admit he is wrong and which one of you here with authority are without ego?

I think that if you never programmed in ASM or C before you can't try out OSDev. If you don't know how to develop, how could you develop an OS?

However Learning by Example is a good thing...But in OS Development we need documentation and discussions about topics, we don't need to copy other code.
If yoy learn by example you can change the code, but the IDEA is the same. And our OSes must be different.

[Mr.Confuzed]

I have to say that this is the best place to start for OS developers, AFAIK. When I started OSDeving, I looked around online, but this was the most helpful place to me. Now I default to looking here first.

I think that if you never programmed in ASM or C before you can't try out OSDev. If you don't know how to develop, how could you develop an OS?

I disagree entirely. I learned (and am learning) assembly and OSDeving at the same time. I didn't know C at the time either. Of course, I learned it on my own instead of bugging you guys with stupid n00b questions. I'm not sure if I would recommend that method, but I liked it!

[Alboin]

I think

I think that if you never programmed in ASM or C before you can't try out OSDev. If you don't know how to develop, how could you develop an OS?

would be better as

I think that if you do not have programming experience you can't try out OSDev. If you don't know how to develop, how could you develop an OS?

Experience is a major factor, be it C#, Java, whatever. No experience, just learning C and writing a hello world program, is not enough to begin osdev. Sure, you can pick it up along the way, but you're going to ask stupid questions which would be better placed on a Java forum.

[Solar]

Just give up, because they don't understand this concept. They think the wiki is the best thing they ever saw, they love it, and it is not going to go anywhere. It will continue to get bloated and more and more useless over time, but they don't see that right now though they have been told.

Do you really think insulting the people who build all this is going to help your case?

Have you seen the Expanded_Main_Page?

[Omega]

Linus didn't know ASM or C before he started linux. That is my inspiration to keep going. This road is hard. You just run in circles, write lines and lines of code and then delete it all hours later because you realized it is garbage, etc. When parts of my OS start working I often sit in disbelief that it is working and I second guess code all the time. The wiki is cool, just let it be I suppose. It will get better.

[bewing]

Do you really think insulting the people who build all this is going to help your case?

Honestly, solar -- that sounds a little bit funny coming from you.

[suthers]

Linus didn't know ASM or C before he started linux. That is my inspiration to keep going. This road is hard. You just run in circles, write lines and lines of code and then delete it all hours later because you realized it is garbage, etc.

Citation please, I haven't been able to find this information anywhere....
Also, I have to agree, OSdev is not the way to start, because if you do, you'll need to be helped every step of the way....
As for learning by example, I think it's a good idea to share code, but if we do that to much just to help beginners, we'll end up with loads projects which are just effectively copy and pastes of all our example just pasted together, then after that, the guy won't know how to do anything, so his project will probably die....
Again, just my opinion...
Jules

[Omega]

Hi Suthers. He said it in this video, enjoy: http://video.google.com/videoplay?docid ... 39DQ&hl=en

It is pretty long but interesting enough to keep your attention; I thought.

[Alboin]

Linus didn't know ASM or C before he started linux.

In 1991, Linus was a second year student in college for his CS major. Considering it was 16 years ago, I'm pretty sure he knew C, and considering he had owned a Commodore VIC-20 and a Sinclair QL before ever owning an i386, both of which he used and worked on extensively, I'm also pretty sure he knew how to program assembly. (If not x86 assembly.)

Edit: ~8:00, He says he programmed in machine code on his v20. Somehow I don't think he knew nothing about assembly when he started Linux.

[suthers]

According to that video, he learned C the year before he started to program linux... in his second year of uni
He got a Vic 20, on which he coded 'in machine code', implying that he already understood the concept of the instructions, so using ASM would only be a small extra step from that...
He sais he got his vic 20 when he was 11, so in 1980..., he also says that he only figured out what a assembler was 5 years later, 1985, 6 years before he started coding linux....
Jules

edit: nice video by the way, interesting stuff...
HA, I've just lost nearly 1 and half hours of OSdev... (currently trying to get a cross compiler working, without it I can't continue, because GCC doesn't seem to care where LD puts the global variables, so I can't continue anymore without getting that to work....).