Changes

Fixed ResourceRefITest

- see CreateResourceRef, this is why you should not combine assigning a value
  with comparison on a single line. I broke it few commits back because I did
  not notice the assignment.
- but it also led me to a discovery that you usually just create a resource
  on the domain, BUT you can also create a resource reference as a side effect,
  which is a feature, not a bug.

Signed-off-by: David Matějček <david.matejcek@omnifish.ee>

Refactoring: domain and server targets vs. resource references

- I expected then when I call a method called create, something is created,
  which wasn't true -> another time spent by debugging due to code swallowing
  mistakes.
- The method should not be called for domain target. However now it is visible
  and we should consider deletion for target domain would make sense - delete
  all references found in the domain. However I did not implement it as this
  commit is just for refactoring. For now it does the same as previous state
  - doesn't delete references.
- I have created two string constants in CommandTarget, so now it makes more
  sense instead of using SystemPropertyConstants or strings. Other commands
  should be revisited later as this PR is big enough.

Signed-off-by: David Matějček <david.matejcek@omnifish.ee>

Fixed ResourceNamingService + added important logs

Signed-off-by: David Matějček <david.matejcek@omnifish.ee>

Relaxed SimpleJndiName validations as resource JNDI names can't pass them

- This commit is quite painful
- The commit contains many FIXME (antipattern)
- These FIXMEs point out to many issues (antipatterns) which together lead
  to proper functionality.
- Commented out changes fix one thing but break another
- Together it makes a complex view of all related issues which should
  be resolved in a standalone PR soon, which will change internal JNDI names
  of resources to a standard format.
- I tried even to fix those names, but it collided with the way they are
  processed in contexts and always something stopped working
- I will fix that in some of next standalone pull requests as this would be
  rather a "feature change" as it is an old design issue.
- Commented out code is related to the last attempt which cannot pass the
  connector_group_3 tests.

Signed-off-by: David Matějček <david.matejcek@omnifish.ee>

Prepare release 7.0.0

Prepare next development cycle for 7.0.1-SNAPSHOT