OSDev.org

The idea that a newbie forum wouldn't be read by the more experienced members needs to be questioned: if the more experienced members weren't interested in helping newcomers, I don't think they'd be helping them now either. They help out because people are generally helpful - they like to help others get past difficult obstacles because they know how it feels to be stuck for weeks trying to track down a hard bug or when they lack some vital piece of knowledge and don't even realise that they lack it. The time when help is most needed is when someone new to OS dev is just trying to get their foot in the door to the point where they can then see what they're doing such that they are able to work out how to get round the rest of the problems more easily - the hardest stage is when you keep getting nothing back and can't work out why you're getting crash after crash. The people who help out also like to be seen in a good light - it's a good advert for both themselves and for their own OS project if they help others, but I'm sure the motivation just to be of help is much stronger. The trouble comes in where time is in short supply and people are trying to keep up with reading the more important new content on the forum, because time and time again they find themselves clicking into threads written by incompetent and lazy people who couldn't be bothered to read the manuals for their tools, and then they get angry, and yet on a good day when they aren't so pressed for time they might not mind at all. By encouraging all such threads from newbies and lazy gits to go into a dedicated subforum, people will only read through it when they're in a good mood and not feel that their time is being wasted whenever they come upon a complete cretin or a child at play. I think the newbies would get just as much help as they do now, and with a huge reduction in the amount of abuse heaped upon them when they get things wrong.

Why not just try out the idea and see how it goes? If someone incompetent or lazy dares to post outside of the new subforum, that thread can be shifted into it without any comment being made, and if that person keeps posting outside of it they can be told to quit it or be banned. If a newbie posts something of quality and sufficiently advanced, it can be promoted into a more serious subforum and the person who posted it will realise that they've maybe been accepted as a quality member of the forum, though they'll soon learn if they don't keep up that standard that their posts will be moved back into the newbie forum - it's more likely that they'd still automatically post anything they aren't sure about into the newbie forum for a while anyway just to keep on the safe side, so I doubt that any more threads would need to be moved than already happens.

I also like the idea of there being a newbie/lazy subforum where you are allowed to take shortcuts and simply ask people to point you towards the best source of info on x/y/z without worrying about having a strip torn off you by people who tell you to google for it - it took me a year to find links to proper information about VESA because every time I googled VESA I found tons of unhelpful junk (some of which suggested you had to pay to get access to the vital documents) and so I kept giving up. Had I realised that googling terms like VBE3 would lead to instant success and that the documents were available for free, it would have been dead easy, but I didn't know that back then - I only realised where I'd gone wrong when I saw someone else being attacked for daring to ask for help finding information about VESA. It wasn't at all important to me at the time to be able to do graphics (and still isn't), so maybe I gave up too easily, but as soon as I had access to the VBE3 stuff I was able to add proper graphics capability to my OS (though not yet in the version that's available online as I have still got to make it fully safe so that people can't set impossible modes and maybe blow up a monitor). Now I can scroll the screen in any direction at speed using hardware scrolling while the cursor stays still and doesn't even flicker, but I could have been there a year earlier if I'd felt able to ask for help without fear of being blasted.

If it becomes policy that laziness is acceptable, I'll probably do exactly what smart questions say hackers do: write an lengthy insult, backspace it away, then mark the topic as read.

For the people who've seen the Sorcerer's apprentice from Fantasia could do that to refer to most newbies out there: if you want something that's way too difficult for you, it will bite you hard. People don't simply need answers, they need skills and the occasional protection from themselves. (for an excellent example of this, see the crosscompiler consensus JamesM recently put into dispute)

@JamesM: sorry, I got confused with my own algorithm What I wrote is totally wrong. I checked my actual code, and I do not copy physical memory at all. I record free spaces in fragments, not allocated ones. Therefore no copy needed, if several threads exit, it became fragmented that's true, but as soon as new threads starts, it will decrease fragmentation. That's not a problem frames are mixed, since paging takes care of that.
(10,1) - now 3 threads exists, it would be
(0,2),(3,1),(5,2),(10,1) - and a new thread starts that requires 4 frames
(6,1),(10,1)
So in the worst case some threads will be blocked when performing exit(), wait until defragmenter makes some space, and continue vanishing.

@piranha: yes, you are probably right. I took dumb criticism personally, but I shouldn't. I'll change that. I think the proper quote for this is: "Don't argue with dumbs, because you'll lower to their level, where you'll be defeated by their routine."

@rdos: I apologize to hijacking your thread. Didn't intend to.

piranha wrote:You were the person to bring up C yet again, as I seem to recall that you do fairly often.

Not so. Look at the timestamps. I took a forum pause, and when I returned to post a few things, I once more got into the same kind of off-putting discussions. This time I think the pause will be much longer.

One problem here is that apparently we all got our egos bruised once too often. As long as we reserve the moral high ground for ourselves, and at the same moment deny it those we are talking with, that is bound to elicit hostile responses.

rdos wrote:I have every right to defend my design decisions when they are attacked. If people find that an arrogant attitude, I'll just let them. I don't care.

@rdos:

It's not about defending anything. This is not a contest that the best man wins. You don't have to convince us your design is better than any other. We don't want to convince you otherwise, either. I don't think I, or anyone else, ever told you to change your language of choice, or your memory model, on this board.

But there is a fine line when someone starts making blanket claims, like "A is better than B", or "you cannot do C in D". Those are factual claims, and the person will be asked to provide verifyable proof. And if no proof is presented, or the proof is not logically sound, people will react poorly to those claims.

That's the way scientific discourse is done. Look at how the scientific community reacts towards creationists, 9/11 conspiracy theories and the like. If you can't prove what you claim, or your "proof" is full of holes, you'll be met with disregard, argument, or ridicule. The latter is unfortunate, but understandable, given that the internet is inclined toward more hostile responses than you'd get at a dinner party...

turdus wrote:"Don't argue with dumbs, because you'll lower to their level, where you'll be defeated by their routine."

@turdus:

Trying to be graceful about something to cool it down works much better if you don't call the other people "dumb" with the same breath.

----

@all:

I've had a PM talk with burbach, with whom I had a run-in in another thread - simply because we were both testy to begin with.

I think we're all on a hair's trigger, and have been for quite some while. There have been many suggestions on how to restructure what we're doing here, none of which really reached consensus. But one thing we all could aspire for, and do so without having to change anything but our own approach, is to handle this board here with some longanimity. It's just a website after all.

We've all been trying to "educate" each other so much that we've all become bitter about it. While I still don't believe in cosseting beginners who shouldn't be here in the first place, I came to realize that terseness is probably much better than trying to acid-spit people in the right direction, for all of us, in the long run. In the very least, it makes responses shorter, so we can carry on doing productive things.

Rdos, I hope you keep posting as you have been. I liked reading your blog-like description of your multicore implementation and your journey into ACPI land. The thing is that not many others do what you did and describe your findings in such detail. I found it very valuable. In my opinion there is far too little discussion about real OS design.

I must admit that it does annoy me a bit when you blame all the worlds problems on the C programming language but not enough to ever make me stop reading.