Thanks...I asked Ken about it (he would've replied to your other post),
but I'm unable to recreate the issue at all (let alone reliably) on
my iMac...hence the curiosity obviously ;) If Ken is correct and it's
a race situation in the code/OSX, then a whole heap of variables become
involved which will make it hard to chase down...
...Ken makes the observation in the other thread that duplicate bottles
should work, and it's there to be used ; I agree with that, but, I never
use it, especially so when trying to get a problematic app going...ie; a
duplication of a premade/prepared bottle containing a bunch of runtime
dependencies is not something I do in triage here...I have many case
examples that clued me to the possibility of getting bogus results doing
things this way...ie; just one example, the Impulse game client -- it
requires .NET 2.0. but .NET 2.0 has to be installed -after- the win32
client software, or you get a murky crash installing the Impulse client
software (and I would never had found that duplicating a bottle) ; there
was another notable situation in my mind, where I was under the impression
the game needed a runtime dep, but upon closer inspection, it didn't need
it at all : in that case, had I have kept duplicating the bottle, I would've
kept hitting the same error, so the duplicate bottles function is something
I tend not to use -- that shouldn't mean it doesn't work, it should, I'm just
pointing out why I avoid it's use, especially when trying to get some app
running (this para furthers Ken's thoughts on this aspect).
Also, when trying to get something running, it is always possible that
some stale process spawn of the test is hanging around, so I've learnt
to routinely quit/force quit a bottle I'm working with to avoid this
possibility...(although I've no idea if such is participle to the issue
here - like I say, I can't recreate it ;)
Apols for the slightly OT post here <g>...
Cheers1