Ultra-difficult questions

I regularly come up with questions that no-one answers, presumably because they don't know the answer. I know we are all busy, but it might be nice to work out answers to such questions between us, if we all chip in with our views and guesses.

I am not the only person. Often when I do a search of the archives for the answer to my question, I find that someone has already asked the question...and also had no replies.

What do people think about finding a different way of handling such questions? Maybe we could have a problem solving forum with all such ultra-hard questions? Maybe we could have awards or "Guru-ratings" for the people who answer them.

I think we ought to find a way to encourage people to help answer such questions. What do you think?

Why if the reason nobody answers the question is because there is nothing understandable in the question?

Maybe what the person is trying to do is against the so-called "windows philosophy" or any other programming paradigm and maybe this issue should be discussed in a "design" forum.

Anyway... I support such a forum, but I wouldn't want questions such as the ones I mentionned aboved to be transferred to this new forum.

I'm kinda new to C++ but I come from a background where I used a variety of languages and designs have always been similar. I really like this forum and I try to help to the best of my ability, as well as finding answers to my problems. I think that design is a part of programming and Visual C++, from what I've seen so far, is the most powerful language I've ever experienced. So what I'm trying to say we cant keep the two apart, unless the question is specifically for design methods. A design forum in my opinion is a waste of server database space, we should share all knowledge including design related in this forum.

cheers
:cool:

I wasn't really talking about design issues so much as other things. For example undocumented features/bugs, or obscure stuff.

For example, If you use DeferWindowPos() to position a parent and child window in the same set, then the function just does nothing. No errors, no reason given, it just does nothing. This is not documented anywhere either. I raised it as a question, and only got an answer by searching the archives, finding someone who seemed to know stuff about deferWindowPos(), and sending him a PM. I guessed that the parent-child thing might be the reason, but I didn't KNOW. Now I do.

The voting seems to be in favour of a new forum. There could be a mechanism whereby if a post receives no replies within 12 hours (by which time it tends to drop onto the next page in forum view and never get seen again), then it automatically gets flagged, reviewed by a moderator, and if deemed suitable, moved to the ultra-hard questions forum. Alternatively if you as the poster are not happy with the replies given, you could flag it yourself for review by a moderator and moving to the ultra-hard forum.

What do people think? Any moderators/administrators want to comment?

I agree that the transferred post should be revised by a "so-called" expert to assess the hardness of the question, vs. somebody who didn't read MSDN or who is trying to do something completely the wrong way (wrong being of course used as: not the usual or intended use). Codeguru has already been for me a wonderful way of precisely finding information about undocumented features, espescially when I was trying to implement a TreeView with tabs, like the one in VC++... yes... it is in the library somewhere... but not in the MSDN... Codeguru is full of neat resources if you search hard enough.

That would be an added value to the already great set of resources CodeGuru already provides.

Slightly off the topic of this thread, but this tree view... I am busily implementing a tree view kind of thing. I am interested in this tree view you say is in the library but not documented. Are we talking about the MFC library or another library. If another library, I would love to know about it.

This is where I took the source from:

http://www.codeguru.com/docking/docking_window.shtml

by Cristi Posea

It is not code treeview related, but it show how to have docking windows VC++ style.

After that, I use HTREEITEM to build the tree.

Ha! I already built my own docking window code, very like VC++ (but actually slightly better : ) Plus it is written in Win32 and not MFC. I had a look at the MFC code for docking windows and boy does it suck. Way over complex compared to mine.

hehe

What is fun about subclassing is that you don't have to look at the code afterwards :) I never look at the base class code anymore. It works, is efficient for what I want it to do... that's perfect. I'm not a purist.

I think a separate forum to which unanswered, difficult questions are transferred is a neat idea. Also, we should encourage people with unanswered questions to post any answer they themselves later came up with. A couple of times, I have found unanswered posts that were relevant to what I was trying to do, and have emailed the folks, "Did you ever figure it out?" So don't be shy about answering your own question if nobody else does -- it can help someone else out.

I do that often. Also, if several different solutions are available it is a great idea to post so that later researchers can find different possible solutions and pick the best one for them.

I think the puzzle forum is a good idea. But why do we have this problem? A lot of questions are beginner questions. So the forum is flooded with them. People who don't even look at Google or MSDN. We discussed that before. Nothing against that, but this way questions get to the second page so fast. In the old forum it was possible to set a category. When I had an easy question, I would put it in "beginner", when it's about "COM" ... and so on.
So if you selected a special category, questions quite old still appeared on first page.
Maybe this would be a solution

I agree to a degree. Every time I re-visit this site I end up answering some newbies question about how to open and connect a TCP socket. I posted code here several times. They really ought to search the archives.

On the other hand if we have too many different forums, then there are a lot of forums to check, and it could be people don't get answers because other people don't have time to check all the forums. I tend to just look at the first page of the forum and skim a question that looks likely and has no replies, and then I answer it if I can. If there were 6 different forums, I might well just check one or two and some people would miss out on possible answers from me.

Even with more forums the unanswered questions would still happen, although they would be visible for longer. If we had a puzzle forum and some form of status reward then all those elite members who are total gurus could just skim that forum and answer questions without wasting their time answering newbie questions that us less but non-beginner people could answer.

It is a good point though. Beginners should search archives first and ask questions second.

For an example of an ultra-difficult question, see

http://www.codeguru.com/forum/showth...797#post535797

One of mine, but I would really appreciate a hand if anyone has any ideas.

Quote:

---

Originally posted by jontom
For an example of an ultra-difficult question, see

http://www.codeguru.com/forum/showth...797#post535797

One of mine, but I would really appreciate a hand if anyone has any ideas.

---

Here is the answer to your question:

Use VB to wrap the MSComm control into a new ActiveX control. Place your custom comm functionality in the new ActiveX control, and design its interface in a way so it's not a pain in the ... to use it from C++. I am using MSComm this way in an industrial program that drives custom external devices, and it works very good.

Back to your original question: Some "hard" questions remain unanswered, because no one knows the answer (I also have one, maybe I'll post it). Other "hard" questions simply "slip" down before the one who knows the answer has a chance to see them (to be honest: who has the time to scan for **every** new post, and read it?). Other "hard" questions are not hard at all, but asked in a manner that makes them inaccessible. There may also be other cases I didn't think of.

I would **not** support a forum for "hard" questions, because it would get misused (everyone who askes an idiotic question and does not get an answer will post the same question there). Filtering the "hard" questions in an efficient manner needs a large ammount of moderation work.

We used to have a rating system, but that also was imperfect, and the "ratings" stood in no relation with the difficulty of the question.

I ignore a question in three cases: 1) when it is [dumb -- homework -- the poster made no effort to find an answer]; 2)when I do not have a clue and 3) when I have no time.