HowD y'all...

What a 666'd ordeal...I developed this code which read RAN
statistical
counters from an Ericsson RNC...sucked them up into a convenient
data
structure...a linked list of an Managed Object type with name, counter
values, time-stamp...

the program returned an error at a
fstream<<counterValue<<flush
statement...

I put in some break points as I'd
only ever encountered such weirdness
due to some sorta memory
release/allocation failures...

When I put in the break-points...the code lost
it's way...it just went into
assembly mode...would crash but not tell me
where...

if I moved the break points...it moved the crash...

I was
thinking of putting the code on my unix box and working there so as to
see what
would happen...

but decided, in the interests of work progress to fall back
to the previous
version of the code...where upon when I was trying to unzip
stats XML
files...I discovered I was out of disk-space....

It only goes to
show...weird, un-experienced problems...are seldom as complex
as you
imagine...and perhaps you should trust your programs error code
handling...fstreams and the rest...

Cheers...I