Meanwhile, apartment blocks in Moscow were being blown up as frequently as buildings were a quarter of a century ago in Belfast and the Asia/Pacific countries were hurriedly cobbling together a peace making/keeping force for East Timor.

Intelligence gathering was notoriously difficult in Belfast in the 1970s but it cannot be so difficult today in Indonesia; a large number of people, including United Nations staff and me, knew as soon as the Government announced a referendum that this was part of a retributive not a generative process. The only reasonable conclusion is that the intelligence was gathered but ignored. In which case, why bother gathering it?

There has always been a tension between the need for secret operations and accountability - one cannot believe that the exploding cigars and other CIA life imitating art devices of the early 1960s would have been possible under more recent Senate scrutiny, though we did have 'Irangate' - but there is no tension between secret operations and open reporting.

As, presumably, everyone takes it for granted that the CIA has spies in Indonesia, including Dili, nobody would have been compromised if the intelligence reports had been published on receipt. Instead of which we are left with a tangle of ethical barbed wire about military equipment sales and training and, more to the point, a quarter of a million dead, innocent people. This is not, after all, a matter of ethics; had the democratic powers been totally indifferent they would not now be intervening. What matters is that the timing was clearly wrong and we now do not know whether this was bad intelligence or good intelligence ignored.

Now that the Cold War has become less hard-edged it is often argued that the real purpose of espionage is commercial but this should, surely, be left to competing international companies.

Much more to the point, intelligence services should concentrate on the messy, intractable issues such as the fault line between Islam and the Northern blocs from China to Serbia, the significance of economic reform in China, the breakdown of government in Russia and Indonesia and the impact of the African HIV/AIDS infection pool.

We can then, effective democracy willing, have a sensible discussion about how far our foreign policies can be ethical and how far they must be pragmatic and, incidentally, whether these are useful distinctions.

It will likely turn out that intelligence is better conducted by economists than paramilitary playboys. Thus far during this decade, the governing establishments have been shocked by the seriously destabilising economic performance of Japan, they have comprehensively mishandled the Soviet Union's break with communism and now have almost certainly - puzzling paradox - brought about and not recognised the almost certain disintegration of Indonesia. Alongside these we may put the chronic misunderstanding of China. All in all, there isn't much to be said for Western Intelligence or intelligence.

These matters are all far too serious and tedious; perhaps that is why the undercover (covers?) agents are neither shaken nor stirred by such great events and choose to imitate the fancies of John le CarrÈ. If only somebody would tell me whether my initial reaction to Diana's death was correct or just a piece of culturally ingrained conspiracy theory.