>>>>> Robert Collins <email address hidden> writes:
<snip/>
> An alternative I proposed a while back was a reader thread and a
> writer thread,
+1, while threads may turn out to be a little harder to debug, splitting
the work and testing each part should help *not* getting there by
catching the problems earlier.
>>>>> Robert Collins <email address hidden> writes:
<snip/>
> An alternative I proposed a while back was a reader thread and a
> writer thread,
+1, while threads may turn out to be a little harder to debug, splitting
the work and testing each part should help *not* getting there by
catching the problems earlier.
Vincent