It is time to consider how we handle the legacy of the Troubles, before it unravels the peace process and undermines one of the most important and successful institutions to emerge from the Good Friday agreement, the office of the police ombudsman for Northern Ireland.
Nuala O'Loan, whose term of office expires in the autumn, has been remarkably successful in building confidence in the police among nationalists. The very existence of her office was one of the main factors influencing first the Social Democratic and Labour party then Sinn Féin to support policing.
Yet she managed to do this without alienating the unionist community, whence come the majority of the complaints she receives. People now feel confidence that these complaints will be properly investigated and the police will operate within guidelines.
Last week she ruled that police were justified in using CS gas in Coleraine in 2004, and in shooting a dissident republican in Belfast in 2002. These findings were widely accepted, not least because the public knew that if the police had exceeded their powers, O'Loan would not have hesitated to discipline them.
The use of force by the Police Service of Northern Ireland does not produce the sort of flare-ups we used to see. People are content to wait for the ombudsman's rulings, and there is a growing confidence she is promoting best practice within the police service.
The figures speak for themselves. A survey carried out by the Northern Ireland Statistics and Research Agency in January and February showed widespread support for her office, with 84% believing she is independent and 85% saying she is fair.
Despite this, there has been an almost complete breakdown in relations between the ombudsman and the representatives of retired and serving officers. The Police Federation for Northern Ireland has severed all links with the ombudsman, while the Northern Ireland Retired Police Officers' Association has issued a 76-page report condemning her methods and judgments.
Remarkably, this serious breach in relations does not involve any criticism of the ombudsman's core role of holding the PSNI to account for its actions. It concerns her attempts to investigate historic cases, notably the Ballast inquiry into the activities of police informers within the UVF in north Belfast.
As originally envisaged, the ombudsman was never supposed to delve more than two years into the past. Her office's historic role was bolted on almost as an afterthought to cover exceptional cases. The Omagh investigation was the first big one and then came Ballast.
Two more reports will be issued in the coming weeks on the murder of the solicitor Rosemary Nelson and on the 1972 Claudy bombing in which the Catholic church and the British government stand accused of covering up the alleged role of a priest in the murder of nine people.
Other areas, potentially more explosive than Ballast, are in the early stage of investigation. They centre on the handling of three security force agents within the IRA Kevin Fulton, Freddie Scappaticci, known as "Stakeknife", and a man known as "the Hawk". All three inquries will go deep into the IRA's security set-up and investigate allegations of collusion between its members and the security forces. They will bring O'Loan and her investigators up against the hidden hand of MI5, which she cannot compel to co-operate, but which set the parameters for much of what went on in Northern Ireland during the Troubles.
All this is putting an immense strain on O'Loan's office and its resources. But what other system can be found that will command the same or greater respect?
Last week a group of former RUC officers, led by Chris Albiston, a former assistant commissioner, did the rounds of ministers, embassies and policymakers in London arguing that disclosures and allegations by O'Loan could endanger not just their reputations but also national security. In meetings at the Irish and American embassies, as well as with MPs, they argued that techniques for handling agents, for replacing explosives by inert substances, and for otherwise disrupting violent conspiracies were endangered by the detailed examination of Special Branch methodology posted on the ombudsman's website.
Behind the officers' arguments lay a subtext the methods used in Northern Ireland were dictated by MI5 and the government. MI5 directly funded RUC agents within the paramilitary groups, demanding regular reports and often sitting in on management committees to handle difficult agents, such as Fulton. It was the Home Office and MI5 that communicated the government's intelligence requirements. The 76-page report produced by the retired officers says: "The overriding priority set by HMG at the time [was] to save life through the effective gathering, assessment, analysis and exploitation of intelligence as opposed to obtaining arrests and perhaps convictions after the offence."
Meanwhile, O'Loan and the retired police officers have pointed out it was standard practice during the Troubles to destroy notes of intelligence operations after they were completed. This was mainly to prevent the identities of agents, or the details of intelligence-gathering methods, being disclosed in court.
This sort of world, where arrests and convictions were not the top priority of a section of the police service and where notes were destroyed so they could not turn up in evidence, puts O'Loan in the same position as Sam Tyler in the TV drama Life on Mars. There, a modern-day detective is plunged back into 1973, where he tries to apply the standards of today's policing.
Another drawback to the involvement of the ombudsman in historic investigations is that when whistle-blowing police officers come forward, they often end up in the frame themselves. Systems and procedures they applied in the past lay them open to criticism, and the ombudsman has arrested a number of them. As a result, few are now willing to come forward.
Clearly there has to be a better way, and it does not lie in public inquiries that have proved an expensive diversion satisfying nobody. It is £180m and counting for Bloody Sunday, and five smaller inquiries that have yet to sit for a single day have already cost £18.5m. The cost is equivalent to "two police training colleges and a hospital", as Sir Hugh Orde, PSNI chief constable, has pointed out.
The best way forward is an inquisitorial truth recovery commission, which would draw on the resources of the ombudsman and the PSNI's historic inquiries team, but with independence and new powers. It would have to be able to grant immunity to those who co-operated and have powers to set the requirement for truth above justice, where that is the only way to get at the truth.
It would also need the unquestioned authority to delve into MI5, army and police records, in the same way as the Stevens inquiry into collusion was able to do. Or as the current Independent Monitoring Commission is able to do in relation to the paramilitary ceasefires. If it was to command confidence in the absence of public hearings, a commission would have to include a range of interests, including a victim's representative, an intelligence expert, a human-rights lawyer and a historian who could collectively make the necessary judgments on what could and couldn't be safely disclosed.
There would clearly need to be an international element in deciding what could be published, so the public would have confidence that where findings were published in a way intended to protect life or some vital interest, it was not simply a government cover-up of wrongdoing.
In this respect they might draw on the expertise of Marianne Birthler, the German official who heads a commission to codify and decide a publication policy on the records of the former East German secret police. While there are no facile comparisons, Birthler and her office have had to weigh some of the kind of issues as may emerge from a genuine truth recovery process in Northern Ireland.
They include, of course, the involvement of senior serving politicians, and public servants, in intelligence operations.