Thinking within this frame of reference, it seems obvious the way to de-escalation would be to involve the rioters in France's formal economy. The trouble is that such involvement takes time to establish, time the French appear to lack at the moment.
While the current French strategy of trying to raise the transaction costs in the informal economy is probably part of the solution it cannot solve this conflict by itself. Among other things, the price of detaining thousands of rioters is simply to high. Lessons can be learned here from the American and Afghan "war on drugs". Despite detaining thousands of people the US hasn't even made a dent in the drugs economy. The suppliers of the stuff simply have to big an economic incentive, in a large part because they lack other economic opportunities. Most poppy and coca farmers would rather have legitimate produce, if only it could support their families.
An offensive approach must be accompanied by a (long overdue) push to crack down on discrimination and include potential rioters (or drug dealers, etc.) in the formal economy by opening up economic opportunities. Such an approach could be inspired by successful development strategies in the third world, since it must deal with a lot of the same problems. And owing to the limited timeframe, a distributed, inclusive, "open source" strategy seems the only way to acquire the leverage needed to counter the positive feedback loop the rioters are trying to establish. If successful in disrupting the French economy the rioters are feeding their own recruiting grounds of unemployed youth. The challenge for the French government is to have jobs waiting in the morning for their disenfranchised youth, to prevent them rioting at night.

"Progress lies not in enhancing what is, but in advancing toward what will be."
Kahlil Gibrand

While we overemphasize and overestimate the short term possibilities of a plan, we tend to underestimate the long term potential of a paradigm shift. And though paradigm shifts seem hard to bring about, history is riddled with them. Just ponder for a moment the aggregate effect on behaviour and decision making of that famous dream I quoted. Sometimes, envisioning the way things ought to be can inspire people to work at making it that way.

Our civilisation is in a state of flux. During such times, uncertainty is high and control is weakened, thus forming a friendly environment for reorganisation and innovation. It is at times like this that individual cells, individual people have the greatest chance of influencing events. At times like this, dreams can have their greatest impact. I hope this site will be a place for dreams. A place where dreamers learn to dream effectively, a place to explore the power of dreams and to dream up the future. So I invite you to share your dreams and nightmares, to envision the world you want to live in in a couple of decades time and share that vision. Together we can learn to wield the power of dreams and who knows, maybe we will even change the world…

"I often wonder, when reading descriptions of the scientific process by sociologists, if this is how an atom would feel if it could read a quantum mechanics textbook."

"We propose a new model of opinion formation, the Opinion Changing Rate (OCR) model. Instead of investigating the conditions that allow consensus in a world of agents with different opinions, we study under which conditions a group of agents with a different natural tendency (rate) to change opinion can find agreement. The OCR is a modified version of the Kuramoto model, one of the simplest models for synchronization in biological systems, here adapted to a social context. By means of several numerical simulations we illustrate the richness of the OCR model dynamics and its social implications."