With the success of the Kurds in taking control of their hunk of Iraq, partition has started to become hip to discuss again. The idea is that if ethnic groups are in conflict, why not just divide them and give each a hunk of territory whether it is called a country or not. When folks remind the partition advocates that things have not worked out that well for Kashmir or for Israel/Palestine, the response is usually: partition works great if done right.
And my take on that is that there are reasons why partition is not done right, and those reasons are inherent in any partition process. That ethnic cleaning and forced migration tend to brew more problems rather than leave everyone tranquil, that the lines that are drawn are especially messy, and that the process of partition is just a wee bit political.
That is, the process of partition tends not to be drawn by technocrats but by politicians with stakes in the game. So, things tend to be violent and leave hard issues unresolved.
If you advocate partition, just be a bit mindful of the realities of past partitions and that those are not so much accidents of history but the products of dynamics built into the process.
[There is much scholarly literature on it, but I am kind of lazy this late afternoon, so scholar google partition and if that does not work throw in Sambanis or Kaufmann to get started]