Chris Burroughs

This is the current "scrub" job ``` echo "Scrubbing jobs for user $USER_TO_SCRUB" mls /$USER_TO_SCRUB/jobs | xargs -n1 -P4 -I{} bash -c "mrm -r /$USER_TO_SCRUB/jobs/{} && echo deleted /$USER_TO_SCRUB/jobs/{}" echo...

A user might assume that anything implementing `SettableConfigContext` already is a mutable `ConfigContext`. Maybe the javadoc should could that? `ConfigContextBuilder` might have been a name more in line with intended...

> I would assume that anything that is public has an assumption of stability from an API contract standpoint As a consumer of libraries that is my expectation as well....

https://docs.google.com/spreadsheets/d/1lly891zfckkYJfV6nfbBvC2flFsUu9Nya2bqF7F07zw/edit# I went class by class and took a stab at what this might look like. I think my largest take away was just how broad the api surface area...

> Is the primary issue here that there are classes which should be package-private but need to be public to work across package boundaries? Yes I think that's a good...

Is this (and #155) excluded from the 3.0 milestone intentionally? (Is CTR-only a minimal viable product?)

(I should also point out that I'm unsure if this is a "real" problem in the final jars or a strange corner case.)

Possible related, I see the following warning when doing `mvn -DskipTests=true package` while shading the benchmark jar: ``` [WARNING] java-manta-client-3.1.3-SNAPSHOT.jar, fast-md5-2.7.1.jar define 3 overlapping classes: [WARNING] - com.twmacinta.util.MD5State [WARNING] -...

In general we shade and relocate dependencies for the primary java-manta-client jar. So `com.foo.Foo` becomes `com.joyent.manta.com.foo.Foo`. This eliminates conflicts between the library versions java-manta uses, and what downstream applications are...

So ``mvn integration-test site` does not induce an error on your workstation? Does the behavior change if you `clean` and `rm -r ~/.m2/repository/com/joyent/`? My understanding is that we can't `enforcer:enforce`...