[SOURCE] A* Pathing with 3D Viewer

[LegacyAX]

wow, Mad Respect from me bro. Great work and thank you. +Rep

[Herleybob]

Lol, i have been.... Its making me mad.

[Apoc]

PPathers actual pathfinder, just plain sucks. Period. I think all the actual developers would agree. (It works, for sure, but the computation needed for it is a royal pain in the ass.)

Step 1) Split the parsed triangles into regions. (Note; you can literally remove 80% of the triangle mesh, by dumbing it down into larger regions.)

Step 2) Implement a better A* engine (or even Dijkstra engine!). It can be done much better, and much faster than it currently is.

Step 3) Well... there is no real step 3... Profit maybe?

[LegacyAX]

I can compile everything fine and all, but when I run it , it will calculate the path... But the 3D model view doesnt show up... is that not included in the package..or...

Im running WinXP SP3.

Below is a screen shot...



The error pops up about 2-3 seconds after I click the "search" button.

Thanks for the app and help

[darrensmith0125]

I can compile everything fine and all, but when I run it , it will calculate the path... But the 3D model view doesnt show up... is that not included in the package..or...

Im running WinXP SP3.

The error pops up about 2-3 seconds after I click the "search" button.

Thanks for the app and help

Hi, the error is coming from XNA, the XNA window is where the 3D view is shown it is opened when you click search. I assume you installed XNA 3.0. (not 3.1 or 1.0) You could see if XNA is working on your PC by running one of the simple samples like this one... Tutorial 1: Displaying a 3D Model on the Screen you will need to remove the XBox tutorial if it is included in the sample solution. If the sample works then it must be something I have done wrong, if it doesn't work then it must be your PC.

Edit: It might also be a problem with DirectX, are you using Direct X 9, please try reinstalling it.
http://www.gamedev.net/community/for...opic_id=434910

[xamxxl]

i dont know how to start it... maybe im to silly but pls say me whyt i have to download... ich have this programm but i dont know how to run it...
pls post mir the files i need to start ist--- im german an my english is bad so i cant read you first instruction

[LegacyAX]

well mt Dev PC cant support the Direct3D , But my laptop can atleast handle the program and second window to load but I dont get any image... just a blank blue screen in the window...

Ill download Direct X 9

Im wondering if this can be used to calculate paths from any XYZ to anouther.

I like how its calculating in plain text so atleast I can see all the data...

You think you could make it output the actual XYZ of the path it takes?

Thanks again for this. It really has opened my eyes to a whole dif, way of pathing. Very inspiring.

[darrensmith0125]

You think you could make it output the actual XYZ of the path it takes?

See in frmTest.btnUseSearchPathThread_Click, at the end of the method...
pathSearchThread.path is the path it has found. If you want to dump the steps found to the log:

Replace:

            if (pathSearchThread.path != null)
            {
                this.Text = "Finished: " + pathSearchThread.path.Count();
            }

With:

            if (pathSearchThread.path != null)
            {
                this.Text = "Finished: " + pathSearchThread.path.Count();
                for (int i = 0; i < pathSearchThread.path.Count(); i++)
                {
                    Logger.WriteToLog(pathSearchThread.path.Get(i).ToString());
                }
            }

[Thidan]

Rather interesting.
Not 100% useable for me though, but awesome post indeed.

[obox]

i think it was your SS i saw in the screenshot thread.. nice

[KOS0937]

To make a few things clear: A* is a Dijkstra algorithm. Though the basic Dijkstra is non-informed.
A* is very quick, but only in 'common' cases. If there is no path at all - or only a very strange one - it is as bad as checking EVERY possible combination of moves before realizing that there is no way to reach the destination.

To avoid such an inconvenience coders usually devide the graph into connected-components and only allow the search for two nodes within the same connected component.
As mentioned before, every component might be seperated into regions so that algorithms know wich region to pass very quickly before calculating the whole path.

Btw the size of the graph does not matter. Only how many neighbouring nodes each node has and how many steps the final path will have.

Everybody who wants to design a good pathing algorithm should read up on the topic first (for example in "AI Game Programming Wisdom Part I" - but there are a million and one books on the topic).

[suicidity]

Screw A* and forget Dijkstra's Algorithm. I made my own that is very efficient and is targeted towards making bots.

Nawh, but if it came down to it I choose Dijkstra over A*; I'm simple like that.

[Maxmaximuss2]

Lol this is gonna be so funny, since now I had been off the forum for studying self-protection hacking and now I should be able to do whatever I want to , anywayes if someone need help with any kind of program then ofc. I will help you all, but this program is a good one I belive, but it can be cracked in 2 mins for eksample by me . But nice post .

I will perhaps consider some programming for your noobs around this forum to perhaps hacking some accounts, since its easy .

//Maxmaximuss2

[vulcanaoc]

Lol this is gonna be so funny, since now I had been off the forum for studying self-protection hacking and now I should be able to do whatever I want to , anywayes if someone need help with any kind of program then ofc. I will help you all, but this program is a good one I belive, but it can be cracked in 2 mins for eksample by me . But nice post .

I will perhaps consider some programming for your noobs around this forum to perhaps hacking some accounts, since its easy .

//Maxmaximuss2

wtf?
)(Filrer)(

[FenixTX2]

Really nice, +rep for the contribution.