Sorry mate we all have to write a login at some point, MS has various templates and samples that help and standardise the login functionality but there are just too many different environments and requirements to automate it. Never underestimate the power of human stupidity RAH

Sorry for the double post. What was I thinking! This will not work: To identify the actual path from the first point to the second point, save (in a solution array that has the same size as the array) the x-y coordinates (or point ID of some kind) of the originating point that was being processed when this new point was flooded, into the array position for the new point. Do this for points starting from the first point (the first color). For the points starting from the second point (the second color), save the new point ID in the current point array position so the path is consistent. To walk the path, start at the second point in the solution array and backtrack to the first point. If you are processing blocks of points in a maze (i.e., 8x8), then you only need a solution array big enough for the number of blocks, not big enough for the number of pixels. What it should read is: To identify the actual path from the first point to the second point, save (in a solution array that has the same size as the array) the x-y coordinates (or point ID of some kind) of the current point that was being processed when this new point was flooded, into the array position for the new point. Do this for all flooded points starting from either direction. When the collision occurs between the two wave fronts, you have two adjacent positions marked with different flood fill colors. For each of these positions, backtrack to the originating point and mark the positions with a shortest path color, then scan the entire array and mark all positions that are in either flood fill color with the normal color of the hallway. You will be left with the walls, the halls, the start and end points, and the shortest path between the points. If you are processing blocks of points in a maze (i.e., 8x8 blocks in a 800x800 maze), then you only need a solution array big enough for the number of blocks, not big enough for the number of pixels. Dave.

Look up the maze routing algorithm - it may work for what you're trying to do. Here is an applet[^] that shows you the basic steps of the search (the forward "wave" and the backtracking).

Actually, no. I never heard the term, and there isn't one in my native language. I just remembered that the outdoor toilet (an improvised shack, if you want) at my grandparents never had any real toilet paper, only old school books and almanacs from the communist era. They were cool to look at - lots of pictures and the almanacs had some pretty useful stuff and good jokes in them, but they were not very soft on my bums and I always criticized my grandma for that. Sometimes I would get some notebooks too, and figured they were my father's. Nice historical past in there. Awwww. The good old days... :)

Rama Krishna Vavilala wrote: But I do want to fully read it one day just for getting some practice to improve my mental prowess. :-D I second that. If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler. -- Alfonso the Wise, 13th Century King of Castile. This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong. -- Iain Clarke [My articles]

Ralph_2 wrote: Monica Bellucci Matrix, semi-transparent latex dress :omg: .........must go now :suss: ...........back in a bit :laugh: Dave Who am I?: http://www.bebo.com/daveauld/ or http://www.dave-auld.net/

Nice.

alikalik wrote: Where a C++ realization of the algorithm (for files, not for memory) can be found ? How would you do it without a computer? If you had two stacks of paper containing names, how would you go about merging those two stacks together into one? Now if you are having trouble with accessing arrays and swapping items in them, don't bother with sorting until you get over that hurdle. "Old age is like a bank account. You withdraw later in life what you have deposited along the way." - Unknown "Fireproof doesn't mean the fire will never come. It means when the fire comes that you will be able to withstand it." - Michael Simmons

Hello, There is a Forum for QT[^] and you may possibly get better help there (with regards to QT). :) It is a crappy thing, but it's life -^ Carlo Pallini

Nobody knows about, even Wikipedia [^], [^] (and you surely won't find links to C/C++ implementations at the bottom of the pages). :) If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler. -- Alfonso the Wise, 13th Century King of Castile. This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong. -- Iain Clarke [My articles]

I found, that it is possible only to iterate in this way: multiset s; multiset::iterator it; it = s.begin(); for (int i = 0; i < s.size(); ++i) it++; I tried to do the following: it += k; But that doesn't work. Is there any oppotunity to "jump" to the position?

Depends on the usage patterns. If the number of lookups is significantly higher than the number of inserts, then I'd just use a sorted vector (as my experiences indicate that a binary search on a sorted vector is significantly faster than use of STL's associative containers in C++). If inserts are more significant than that, use a std::multiset as suggested elsewhere. If using a sorted vector, it helps to delay the sort operation until required, using something like below, rather than blindly sorting on every insert. template<class Type> class SortedVector { public: SortedVector() sorted_(true) {} void insert(Type const& value) { sorted_ = false; data_.push_back(value); } typename std::vector<Type>::const_iterator find(Type const& value) const { if (!sorted_) { std::sort(data_.begin(), data_.end()); sorted_ = true; } std::vector<Type>::const_iterator found = std::lower_bound(data_.begin(), data_.end(), value); return (*found == value)?found:data_.end(); } private: mutable bool sorted_; mutable std::vector<Type> data_; }; Java, Basic, who cares - it's all a bunch of tree-hugging hippy cr*p

I'm creating add-in for MS Excel. The argument is a value in a cell. So for instance, if I call MyFn(B11) and MyFn(B12), where B11 and B12 are cells, the result will be wrong.

Have you read this part? This message is sent when the list box of a combo box is about to be made visible. The parent window of the combo box receives this message through the WM_COMMAND message.