reference comparision

[Ulf Wiger (TN/EAB)]

Lennart Ohman wrote:

> One can on the other hand discus the relevance of doing > and 
> < comparisons on a reference...

If references couldn't be ordered, you couldn't use them as
keys in orddicts, for example. And if there weren't a 
consistent global ordering, you couldn't exchange such orddicts
between nodes. That would violate the principle of least 
astonishment.  (:

BR,
Ulf W