[Coin-discuss] lots of failed assertion with Sample2 on Solaris8

[Lou Hafer]

Sven,

	(John, tune in, please  :-).

	Yep, I know just what you're talking about. I've been working on sbb
on Solaris 8 using Sun (Forte) workshop, running the miplib test set using
the sbb main program. I see the !numberDualScaled assertion violated for the
miplib problems arki001, bell5, fiber, gesa2, harp2, khb05250, and mod011.

	The command line I'm using for sbb is
 
sbb -solver clp -loglevel 2 -time 600 \
    -directory /cs/mitacs1/lou/COIN/Sbb/Samples/miplib3/ -import arki001.gz \
    -probing on -gomory on -knapsack on -oddhole on -rounding on -branch

(A sort of dumb default --- turn everything on and see what blows up.)

	dylp (the solver I'm responsible for) has its own problems, and I'd
decided to pursue them first, then come back to this if John hadn't picked up
on it.

	John, what is this assertion complaining about? On the Linux system
that's available to me (RedHat 9), I see the same assertion failure for
arki001, bell5, gesa2, harp2, and mod011.

							Lou