[invalid] Issues with RoutePlanner-beta

[tomreyn]

M0ellemeister and I - both on unixoid systems (FreeBSD + Ubuntu) - played a network game with two AIs and RoutePlanner-beta on r1203.

The game did not crash or get out of synch, so I contribute that part to cross architecture games exclusively. However, there were some other issues:

* One of my workers whom I had ordered to walk to some distant place was stuck in the middle, walking down a field from top to bottom (north to south), and once he reached the lower end of that field, would magically disappear, then reappear at the top of this field, then walking the same way again, then disappear and reappear again... a vicious circle.

* Several of M0ellemeister had ordered several workers to walk to another place (probably also a longer walk), and they went in this direction, but after a while they stopped in the middle of their walk. Manually giving them new orders after this worked, though.

[m0ellemeister]

* Several of M0ellemeister had ordered several workers to walk to another place (probably also a longer walk), and they went in this direction, but after a while they stopped in the middle of their walk. Manually giving them new orders after this worked, though.

This was happening several times on the same way to the same destination. Every time the workers stopped near by a stone. (Not mineable one).

[ultifd]

So we tried it today, same thing.
Here's my debug log: http://www.mediafire.com/?iuan55akbzqmo38

[tomreyn]

And mine's here: megaglest_btrunk+r1206_d20101103-012515UTC_ptomreyntrunk_trouteplanner-timeout.tar.gz.

What we did today was to play a 3-person 3-bots game with Routeplanner-beta. Ultifd was on Windows, m0ellemeister was on BSD and I was on Linux. I was hosting. As we expected to happen, Ultifd timed out a few minutes into the game. There was no segmentation fault, so there is no stacktrace, but the tiomeout on cross architecture (*nix vs. Windows) games is very reproducible.