> I am uncomfortable with the MIT license's sublicensing grant
I did a little research to try to figure out what you meant; tell me if I have it right:
MIT allows you to redistribute the code under a different license. The problem is that the sublicense might be more permissive than the MIT license, potentially allowing someone to bypass the requirements of the MIT license.
One right that can be passed on is the right to sublicense, which gives the licensee the right to issue new licenses. The licensee cannot provide downstream licenses to rights he or she did not receive. So the license can never be more permissive than the existing one.
The difference is that under the BSD license, the original author grants licenses and requires notice on downstream redistribution. Someone can add code, however, and add restrictions to that code. The MIT license allows something very different, which is the right of a mere licensee to negotiate licenses (provided they are no more permissive than what was granted) on behalf of the licensor.
So you can take MIT licensed code with no changes and relicense it under the GPL. The BSD license requires that you license only your code in it (and the corresponding work as a whole) under the GPL, but the MIT license allows you to add restrictions to code you did not author.
I did a little research to try to figure out what you meant; tell me if I have it right:
MIT allows you to redistribute the code under a different license. The problem is that the sublicense might be more permissive than the MIT license, potentially allowing someone to bypass the requirements of the MIT license.
Is that what you meant?