Fatal bug in r1867

[Pete Mack]

Nick's right. And if it is a memory bug, Linux valgrind should find it. (I know that it gets used from time to time to track down this kind of thing.)

Alternatively you could run it from gcc and find out where the crash occurs.

[d_m]

So valgrind did detect some shady memory access in pathfind.c.

I have committed a band-aid fix in r1887, which should help Windows users. That said, it would be good to have a better fix in place eventually.

Please let me know if this fixes the problems on Windows (or doesn't fix them, or causes other problems, ...)

[Pete Mack]

@Nick--
Up until Oct 22 I would have agreed with you--Windows is a drag. Windows 7 is what Vista should have been, but wasn't.

[Nick]

Windows 7 is what Vista should have been, but wasn't.

Yeah - I've used it a little, and it seems reasonable. I still prefer both OSX and Ubuntu so far, though.

[Faust]

r1887 seems to have fixed it.

[Nightmarjoo]

Going to bed now, but I'll check it when I wake up. Hope it's fixed, bug made playing angband depressing, I would flinch everytime I saw a brown #

You said you found some shady pathing, what exactly was the issue? Just seemed like a really odd/inconsistent bug. I wonder how it came about too, cuz in 1626 it didn't exist. What added it?


Alright, so far so good, can't recreate the bug in r1896, thanks.