OSDev.org

I hope so, but I sometimes have to wonder...

It really does seem as if the majority of people coming to the site are skipping the forum entirely, and worse, many seem to refuse to use the information in it.

Mind you, most of the people who come here don't seem to really even use the fora - many seem to join to ask a single question, get an answer, and then never come back.

(Though many also seem to ask a question, get rebuffed by the experienced members and either told to read the wiki or search the fora history, or are informed that they are asking a question that either cannot be answered with code or doesn't even make sense, at which point the storm off in a huff rather than taking the advice to heart. While some of this is definitely up to the overly-harsh "git good" attitudes on the fora, it usually seems to primarily be caused by an unstated assumption on the newcomers' parts that learning OS development is going to be exactly like copying a bunch of JavaScript snippets off of SO, just more of it, and get angry when you tell them otherwise.)

What's more, there seems to be too little interest in maintaining, updating and expanding it in the on the part of those who already know. This is understandable - those old enough to have the necessary experience generally have more pressing demands on their time, and have to either scale back on the hobby or drop out of it altogether - it is nonetheless a serious problem, as the group is small enough that even when all the experienced members are free to actively participate, there is only just barely has a quorum of what is needed to keep it up to date.

(For my own part, I can't do much more despite having a lot of time on my hands - due to the reasons I have that time. I cannot work on much of anything for more than about an hour at a time - either personal goals, including basic self-maintenance, or professional work, which I simply cannot risk committing to because I would never be able to accomplish any agreed on goals - without self-doubt and fear of failure cause my work to grind to a halt. I am slowly getting better, but most of the time, the most I can do is rattle on idiotically on message boards - just like I am doing now.)

I think it is still a valuable resource, but I also think we need to consider ways to address these issues.

If the mods wish to move this to the General Ramblings forum, I would have no objection, but I did think it fit on this forum as it directly addresses the wiki's issues.

Hope you don't mind me throwing in my 2c on this. I haven't contributed to the wiki, I'd like to but I am simply not knowledgeable enough. The wiki is incredibly useful though, it's unparalleled as a repository of knowledge on a very niche, but utterly integral part of software engineering. If it disappeared, the barrier of entry for hobby OS dev would be much higher.

One thing to keep in mind, the contributors to the wiki are by necessity a small portion of it's users. The majority of people using the wiki are going to be people who don't know enough to contribute. Additionally, the people who get the best use out of the wiki won't post here. I mean, if all the info they need is on the wiki, why ask a question? The more effective the wiki is (and the archived conversations on the forum) the less people have to ask. Personally I've only ever had to post one question, because the wiki has all the information I've needed to do everything I've done so far, and bug fixing rarely requires a subject matter expert, just dedication.

If I may offer a personal observation I think that very little is happening in OS dev in general. Not to say no-one is doing good or interesting work but that there is very little new information to put up on the wiki, the fundamentals aren't changing. The pages already there are kind of good enough for a dedicated newbie to access the necessary information, so there isn't much motivation for any of the tiny number of people with the requisite knowledge to put in the huge amount of work to get a fairly limited improvement in quality.

This can kind of be seen in the posts in the development sub-forum. Most of the posts fall into small categories, and mostly they're people who have bugs in the programmatically complex areas of the x86, UHCI, PIC and GDT. Once the bugs are found and fixed, there's little left to discuss because this stuff is basically all very well understood. There's no incentive to collaborate because somebody has already uncovered all the necessary knowledge and documented it somewhere out there, maybe in the Linux source, maybe the wiki or maybe the forum. All places people are expected to check (rightly so) on their own initiative.

To summarize my rambling, there's no low hanging fruit left on the OS tree. There is good enough documentation for the basic concepts necessary to develop an OS for x86 (MMU, VGA, ect.). The stuff which consumes the time of the hobbyist OS developer that isn't on the wiki is so complex and specific that general purpose documentation would be uselessly shallow (like graphics hardware), stuff that's specific to the OS itself (API design) and stuff that's outside the scope of OS development (application code).

I have used the wiki very often, keep using it frequently, and have contributed a few bits and pieces. There are few areas, though, where I consider myself experienced enough to contribute much relevant information. I hope that I will soon find the time and knowledge to contribute more to non-x86 topics (mostly ARM, Raspberry Pi, MIPS for now) which might be interesting.

Wow, I am really thrilled to here all of this. Seriously, I have been getting a bit worried about it, and while some of that is due to my own self-doubts about how important or useful my own contributions to it have been, it does help me put them - and the wiki as a whole - into more perspective. Thanks to all of you.