OSDev.org

Hi,

I've been using this forum for quite a few years now. For my OS, I found that the Intel and AMD guides were the most usefull resources and then the osdev wiki helps to kickstart when you have no idea where to begin looking. Although I did use the forums many times, I ended up being able to find the information I needed in technical documentation and now my OS is in state that I would consider pretty much acceptable.

I noticed that many beginners, when asking questions, often get responses like "let me google that for you" or other answers that may read as arrogant by advanced users. Although I have never posted on this forum, in other forums I try to answer those questions even if the answer can easily be found on google. I do this because when I google something, the first hit is very often a link to a forum where a person has asked the same question I googled but someone told him to google it.... So I try to always answer because I think that information should be duplicated as much as possible to that it gets easy to find.

I understand that a poster might be seen as lazy because he did not take the time to google it, but in my opinion, that's what forums are for. We can always google everything. There are VERY good OS design books out there, the linux source code is everywhere, and Intel/AMD has very good documentation so technically, we don't need this forum at all. But sometimes when going through 50 pages of documentation and not finding what we need (maybe because we don't know exactly what to look for) then it's nice to be able to ask someone else in a forum.

I guess I just find that sometimes (not always), there is a hostile culture here that could "scare" beginners. I like forums (and also reddit) where people gladly answer a beginner's question. For me, if the answer is easy, I am just so happy to respond to the person because I am really passionate about this subject. And if I find that answering that person's question is just a waste of my time, then I will just not answer the post.

So I guess I should lead by example and start answering questions today

It's really a question of educational philosophy. Some believe that education is a matter of imparting facts to recipients. Others view education as the process of encouraging people to think, and to research questions themselves. Another factor is that some derive pleasure from demonstrating their knowledge whilst others derive pleasure from increasing the ability of beginners to develop their learning abilities.

Your view of the educational process will inevitably colour your response to simple questions.

That's interesting. Definitely something to think about. Something I should add is that in the workplace, we are encouraged to talk to each other and transfer the knowledge as fast as possible. It would be counter-productive to tell a new-hire to just go through all the company documentation to understand the product. Instead, he would be encouraged to do so, but we would also any questions he has when wanting to shortcut the documentation to start fixing bugs right away. We often feel that getting the new-hire to jump on bug-fixing and asking questions around is the easiest/fastest route. Soon enough he will want more information and he will find that the documentation is a great resource.

I guess that if a person's philosophy is to encourage reasearching questions, then maybe there is a way to answer in a non-arrogant way (to me, telling someone to google something sounds arrogant most of the time). Maybe by helping the OP formulate the question or to guide the OP in the thought process (I've seen this acutally on these threads, and it's a very interesting way of answering). Or a person can simply not answer the OP if he feels that the OP should research more.

BTW, I'm not complaining about the quality of these forums, I am just trying to start a sane debate that could help make the forums better because I see a lot of potential on this website. I started writing my OS before osdev existed (and before google) so I am very grateful that we now have such a great resource online.

I think a work situation, where time is money and a lot of custom software will be in use, is different to a forum like this where people are doing it for fun. But if someone at work kept asking trivial questions of the other developers I think they would soon be looking for another job.

I don't actually see many "I'll Google that for you" type answers. I do see "Read the Intel manual", often answered with "I can't be bothered to read that", which seems to be an entirely appropriate answer and is far more helpful - in the long run - than just giving the answer. Agreed sometimes people can be a little short when the poster has obviously made little or no effort to research their question - we're all human, and don't always appreciate laziness. We have had some truly horrendous posters here from time to time who post about every trivial problem that they have.

In similar vein, a number of people post their code here, with a question like "my OS triple-faults", expecting someone else to do the debugging for them. That's not, IMO, how it works. If you can't be bothered to debug your own code why are you bothering to write it in the first place? (The short answer is that often the poster hasn't written anything, but has just cobbled together code from various sources expecting it to work.)

I like OS development because it's a challenge and a good intellectual exercise. I wouldn't want anyone else to do the work for me.

Public forum messages can serve you to identify people that can team up with you. Then you can private message them and contact them outside the forum through their contact information.

For example I can assure you that I can answer the questions you and other people have as long as I know it, and I can help you implement complete functions because I will be needing them too.

Just identify the people who want to team up more for general direct cooperation and who have contact information. You can ask who wants to do such general cooperation. That could be better and call for more people than wanting others to help you in a way that is too specific as that will allow them to help you in the same or similar areas to those of their projects where they are currently working at or that already have a code library that they can share with you.

For example, I can help you if you want. You could contact me outside the forum and see what problems are to solve. I have so far been contacted by 1 person from Peru around 5 years ago who wanted to team up with me. I helped him a little then he just went away from IRC, he no longer logged in. But I helped as much as I could in a generic way.

If you find that kind of people you would at least gain explaining more of what you are thinking in a chat or the like to actually help you. You could also do that here in the forum and the IRC chat, but that would require you to talk in terms of a more open source project so that anyone fully understands how to contribute to your code.

lockcmpxchg wrote:Something I should add is that in the workplace, we are encouraged to talk to each other and transfer the knowledge as fast as possible. It would be counter-productive to tell a new-hire to just go through all the company documentation to understand the product. Instead, he would be encouraged to do so, but we would also any questions he has when wanting to shortcut the documentation to start fixing bugs right away. We often feel that getting the new-hire to jump on bug-fixing and asking questions around is the easiest/fastest route.

Well... I think that is a sign that your documentation is not up to standards. IMHO.

For the software I am maintaining 9-to-5 (basically an interpreter for a domain-specific language), there are five pieces of documentation -- a developer guide on how to set up the dev environment, the internal structure of the project, how to build the software etc.; an general reference to the domain-specific language; two tutorials for the two different basic functionalities offered by it; and a reference for the DSL "library" offered with the software.

I would fully expect any newcomer to work through the developer guide to set up his system and get a rough understanding of the internals, then work through the tutorials to understand the product. He might ask questions when something is not clear, which I would take as a kind of bug report against the docs. But I will not sit down and walk him through introduction -- because that is what the developer guide and tutorials were written for. (The subject is quite involved, and I'd expect several weeks to go by before a newcomer got a grasp on "how things work". That would be weeks I'd be reduced to next to zero productivity if I were to walk him through. Not to mention the deficiencies of ad-hoc teaching, and the possibility of me being not there for a walkthrough, due to illness, accident, or misplaning.)

And even later on, I would expect that documentation is consulted if there's a chance it might answer the question. E.g. "how does the extract() command work?" is right there in the docs. (And, actually, I would consult the docs myself before answering the question to avoid giving wrong info -- so why don't you do the lookup?) "Why doesn't extract() work here when the docs seem to imply it does?" is a very valid question, on the other hand.

Similar applies here, or at StackOverflow. A question of "I read that X is supposed to do Y, but this code here seems to indicate otherwise" is usually well-received. A question of "how does X work?" is not, with the LMGTFY pointer indicating that the information is readily available, and that you should have checked before asking others to do the checking for you.

----

Besides, taking shortcuts into "fixing bugs" is a recipee for desaster, in my experience.

Well... I think that is a sign that your documentation is not up to standards. IMHO

I agree it could be good indicator that documentation is poor but we must not forget that each person can learn a different way. For example, the Intel developers manual are just remarkable. I've been reading those bibles since the 90's and they are, in my opinion, the best technical guides out there. But still, sometimes I've had a hard time understanding some things. And I think it might be because of the way I read it. and I was happy to find some posts (on this forum) where people had the same questions. some people answered to look at the intel docs (which I did, and maybe the OP also), and some people were nice enough to elaborate a little more and give more information about their personal experience. That ended up helping me in many cases.

Sometimes I see people asking questions here for information that is readily available and I think "maybe this guy needs someone to explain from a different angle".

For example, yesterday I googled some info about vesa and vbe, the first result is a wikipedia page. That page is not technical enough. I have every intention of reading the official docs but I'd like to read what some people who actually had to implement it as part of their OS have to say. The 2nd google result is a link to these forums. A person is replying (very arrogantly) to do a google search. But that is exactly what I did. And I was hoping that this link would provide me with the information I wanted. The way I look at forums, is that they are archives of most requested information.

I also found this post: http://forum.osdev.org/viewtopic.php?f= ... 05&start=0. The OP is asking a very basic question that he could have found the answer to on wikipedia. but someone politely answered his question anyway and guided him in the right direction. Very good, because that link ranks high on google and a lot of people will land there and will be able to continue their research. But why did the 3rd poster bothered to answer if he had nothing to contribute to the thread? I think that staying polite and to be helpful is always the key to a good discussion, otherwise a person should probably not participate in that discussion.

If you have to go back 8 years for an example I think the problem may be less acute than your OP implies.

iansjack wrote:If you have to go back 8 years for an example I think the problem may be less acute than your OP implies.

That depends on whether going back that far was in search of a polite answer or an impolite answer!

Perhaps it took that long just to find an answer?

Well as I mentioned, this was on the first page of a google search. I did not try to find this post. I could list other posts but it wouldn't be constructive to do so if people only read the posts just to find something wrong with each one of them. I was just trying to make a point.
I sense that maybe some of you think I'm trying to see problems where there are not. That's not what I am trying to do. I was simply initiating a healthy discussion about how to improve things. If the osdev community does not feel that such improvement is needed then that is ok with me. Just trying to help

Thank you.

I'm sure that if you Google just about any question on programming you can find posts that say the equivalent of RTFM. Although, as a matter of interest, I Googled "vga functions in protected mode", and the first link was a very helpful page in this wiki. I'd have been inclined to either give the "Google it" response (it's quicker than going through the details every time, and a lot more helpful) or not answer at all. And what happens if everyone does the latter? Bump - bump - bump ...

This site makes no secret of the fact that it is intended for experienced programmers who are expected to do a little research and reading before posting questions.