Repairing the Growing Point

Having identified the form of the breakdown, the system selects an alternative way of achieving the desired behavior. This alternative may include running the same code again or it may use alternative implementations which appear more likely to work given the nature of the breakdown. Let us look at the problem of the stuck growing point, and see what must take place in order to choose the best way of continuing the computation. These are a few alternate methods to achieve the original computation.

Try, try again.

This is plausible when there is reason to believe that the cause of the problem is transitory. For example, if pheremone values are time-varying, or if processors are mobile, then simply waiting for conditions to appear in which it is safe to proceed with the original computation makes sense.

Increasing the size of the growing point.

If there aren't enough good choices to choose from then simply increasing the size of the growing point may be enough to reduce the sparseness of the environment. Perhaps sparseness is a problem at one scale, but not at another.

Working around the bad spot.

As long as we are making progress towards the goal, taking a detour may be the most expedient way of solving the problem. For example, diatropic and plagiotropic tropisms will suffice if orthotropic tropisms run into a wall.

Backing up.

Sometimes you paint yourself into a corner. By reversing the tropism, you might find that a better opportunity presents itself. More on this in the next section.

Ignoring the problem.

Sometimes it is better to do nothing at all. Other factors or higher-level directives might view a stuck processor as a good thing, or at least better than the alternatives. There may be cases when knowing where not to repair wins.