matma rex in the first bit about file what file does it mean

The Inkscape file you're working from.

i typed it in it said try again

Are you sure you typed it like this?
C:\Users\My Stuff\Maps\Map.svg

You need to give the exact location of the file.

Moros is right - either type in the full path of file, or copy both the file and my program into the same folder.

It's not worked perfectly... Connection missing between Yakimovo and Medkovets (Montana Province)

Yup, it doesn't work correct with multi-part territories, or those with holes in them, you have to do those manually.

It's written in original post, isn't it?

moros i did it manually i just need a name making one.

E88's example is not a holey or multi territory connection..

E88B3EC, thanks for report, I fixed the connection in map version 1.0.1 (currently pending public). I wonder why it didn't work, but I don't have time to investigate now.

I tried to use the tool which worked great until i got to the upload part and the it won't upload is says It didn't work and to report the error any ideas?

Uh, try reporting the error. Did you take a screenshot? What did it sey? (Apart from mthat it didn't work.) I can't fix it if I don't know which part fails.

okay here is what itt says word for word

Starting up, be patient...

This tool will allow you to automatically create
connections on your map.

It will have to ask you for your e-mail and password. This data
is not remembered or used for any malicious purposes,
it's simply required to confirm that it's your map
and you have the right to edit it.

It will also need the map's id. You can get it
by clicking "Get Link for Sharing",
the id is the number displayed at the end of public link.

Enter the name of file to work on:
> C:\Users\Living Room\Documents\Nate's Files\Warlight Maps\Map 5.svg

And now the map's and your data: [warning - your password will be visible]
Map id: North America (Medium)
E-mail: *********
Password: ***

Working... this may take some time. Go have a cup of tea or something.
This map appears to have 408 territories.
Loading data...
Press enter to exit.|oooooooooooooooooooooooooooooooooooooooo | ETA: 00:00:00

It says a whole lot of other stuff as well but i can't copy it fast enough before the program closes. However it used to get to the part where it uploads and then it says ** it didn't work there might be a problem with your connection please report it. Can you help please

It closes as soon as you press ENTER.

I know that but i go back to check my map and it still has connections missing here is the map you should recognize it click here
Ex. the Central Mountains of Virginia and Southwestern Virgina Mountains are still not connected even though i used the tool.

Oh, but it did add some connections? Then it worked, mostly. It's not perfect, unfortunately, and you might've hit a corner case - just add the connection yourself.

Matma, it's a similar issue that E88 reported earlier...

This program running for several hours was something that really amazed me. As such, I decided to take a look at your source code and see how you did this. Though I don't really know ruby, I was able to make up a lot out of what I saw. I hope you don't mind me taking a moment to post my findings here.

First off, let me say that the way you reduced the problem to collision detection of bounding boxes. It is a very sound strategy and is likely to generate rather good results.

The first thing that stood out in my opinion was the fact that you are lazily generating the bounding boxes, even though you know up front that you are going to need them for each and every territory. Though this does work, just calculating the bounding boxes first removes a point where you can make mistakes (the laziness). I should add that my lack of knowledge when it comes to ruby made sure that I didn't really understand how you were doing the laziness, but I got the intention.

The other part is the actual collision detection. It is a very simplistic algorithm running in O(n^2). Because collision detection is so important in things like games, there are lots of efficient algorithms for this, including some that are still rather simple and can be implemented easily (especially when you can make some assumptions about the input, like you can with warlight maps). However, it is not entirely clear if making a more efficient solution here is actually going to improve your solution (especially since n < 3500). However, with the change suggested above, where you make the determining of the bounding boxes static instead of dynamic, it will be possible to have a progress bar for each of the two, which will make it easy to test which of the two is your bottle neck in real world examples.

I hope you can do something with my comments, but I won't be offended if you don't. I just thought I would share my findings with you.

Didn't study the details, but my gut instinct says that going from O(n^2) to O(n log n) will very likely be faster, even for small n.
Remember, even if it's ten times as slow for n = 5, it will still be near-instantaneous, whereas reducing the running time by 50% for n = 100 or n = 1000 will save you hours, which will be very "noticeable". ;)