Re: order of destruction of variable by the garbage collector

["Peter Hill" <corwin@...>]

Mark Hamburg:
> Beware, however, with respect to the weak table dependency mechanism. If
> it just provides acyclic dependencies you are fine. If, however, it is
> used for more general references, then I think partially weak tables can
> result in uncollectable cycles.

RLake:
> Perhaps we will see a solution to this problem in a future Lua version. It
> appears that the issue crops up more often than was originally thought
> (including by me).

Hear hear! I second that!
Once more this problem causes unexpected & undesired results.   :-(

The problem with leaving an abberation like this unattended to is that it IS
likely to keep cropping up and biting people more than expected. And while
it is not actually an "error" per se, it IS a break with the basic principle
of weak links (which is why it causes problems).

Didn't you actually implement a solution some time in the past? I personally
think the problem should be fixed & debugged before 5.1 adds the additional
incremental complexity to the gc.

*cheers*
Peter Hill.