btw, here (http://www.mech.upatras.gr/~robgroup/ol ... index.html) you can find a demo with source code (it was in Delphi [the algorithms are in the respective folder]) for various path planning algorithms you can use on a grid

don't remember if I had also implemented A* in that demo, probably not (I see it has BestFirst, BreadthFirst and DepthFirst)

UPDATE: it doesn't allow me to post URLs here, so replace the last part with

~robgroup/old_site/search/index.html

If there was a way to visually link items and make directional graphs (with an arrow pointing towards the direction and two arrow heads for both directions [to avoid making two links in that case]), and you could query the node connectivity, one could more easily do path planning on the graph

It would be nice if there was a Shuffle behavior, where I could make a container (held by some background component or something), assign various sprites to it and then tell the container (or the parent object that has the container [if I understand how containers work]) to shuffle the positions of the items held (it would be a Shuffle behavior on the parent I think if it only holds one container [is it a property?] at a time)

I'd like to be able to set a time interval for the auto-shuffle (and be able to set that to 0 to stop shuffling) and also to be able to manually tell it to shuffle the item locations

this is very handy for memory games (e.g. where you have a grid of turned-over cards and you want to find pairs - they need to be shuffled at the beginning and possibly also periodically for making it harder)

not sure if there is such behavior already and I guess I need to learn the Javascript API (or can I make behaviors using eventsheets?) to implement it myself (is there any example of how to make a new behavior)?

The image editor although looking pretty at first look, seems to be very limited. There isn't even a tool to make an ellipse [holding down say SHIFT to force a circle] (one can only do boxes [haven't checked if can hold SHIFT to make squares btw])

Would be nice if that image editor had a button to open the image in an external tool (saving any pending changes first) and monitor the file in the filesystem (Windows have shell notifications for that, don't need to use polling) to update when the file is changed (so that you could even keep the external tool open)

what is the current status with translation? I'd like to do a Greek translation (so that a Greek kid can use the UI)

is there some Pootle server for contributing translations or do they use Transifex, or GetLocalization or similar service?

Another thing I could use is Dictionary, if only it had a mode to work as a MultiValueDictionary, where it allows one to add multiple values for a given key and returns a list of values for the key if it has more than one value

thanks for the replies,

found this one too that sounds interesting:

Data storage objects in a container

----

It's possible to add data storage objects like Array and Dictionary to a container with another object. Despite the fact these objects are invisible, a separate instance of the object is still created for each container. This allows you to have a dedicated Array or Dictionary for each instance of an object. This can be very useful as an advanced substitute for instance variables, such as if a very large number of variables is necessary, or if variables need to be dynamically added and removed.

scirra. com /manual/150/containers

indeed, had recently read about that but it skipped my mind

Another thing is that I need to be able to go the other way too, from clicking something on the display to finding the correct cell in the datamodel (the array). Can I keep some Tag value on visual objects or use their Name or something to point back to the cell's number? (I mean x in (x,*,*))

And in the (i,0,0) position I'd store some info on where the node is on the UI or some reference to a graphics object (are there such?). If I can only keep numbers in the array then I'd need (K,2,N) dimension so that at (I,0,0) I store X and at (I,1,0) I store Y coordinate of the node on the display

I guess I could use an array anyway, of (K,1,N) dimension - I could store a number 0-(K-1) in each cell to mark a one way link. For two way links would store two times, as two one-way links

It's a turn-based chase game, where nodes can have say up to N links connecting them to other nodes and there are N-1 enemies chasing a player that moves over the links on the graph nodes. The nodes+links (graph) is the underlying data structure, the display can be some drawing where the links aren't necessarily direct lines (can show up like say paths in some forest from a top view)