Improvements in logging

[dmatej]

• I spent four days to fix the last failing tests on Corba refactoring, ORB updated to 5.0.1 glassfish#25804 comparing logs of original state and my PR in a hope I will find the last side effect I could not find.
• I also noticed few other issues I have seen in GF in last weeks.
  • Local host name resolution was strict, localhost was not accepted as local.
  • ignorance of InterruptedException

The initialization is heavily dependent on the order of operations, part is done from GlassFish, another is sparked from setParameters method call and finished by PEORBConfigurator. I would like to make it more robust in the future, but I need to know perfectly the required order of operations.

The #288 should be merged first!

[dmatej]

@pzygielo Could you please review this? I would then continue with some automated cleanup steps in another PRs:

• Code formatting
  • Running script just like on other projects, removing trailing spaces and tabs from sources.
  • Adding @Override annotations (make refactoring a lot safer, breaks the build when we forget about children)
  • Formatting imports (just like on GlassFish)
  • I would probably stop here, the rest must be done with manual changes. Not sure with that, some parts are really hard to read as whitespaces and new lines are nearly random in some files. Maybe it would be better to do that before we make other important changes.

Then ...

• Then we would release 5.1.0
• Then we would remove Applet references as they are not available in Java 26
• Then we would integrate 5.1.0 to GlassFish 8 and GlassFish 7.1.1
• Then we would start 6.0.0-SNAPSHOT
• Then we would remove SecurityManager references as we already removed them in GF8
  • That would probably break TCK for GlassFish 7 so GF7 will stay with 5.x.x forever.

GF8+ will also do more iterations in the direction of eclipse-ee4j/glassfish#25804 to simplify the initialization of ORB. At this moment I don't know if it will need more changes in this project, but it is possible.

Finally this project is not in so bad shape as I originally thought, but ORB initialization is quite complex and was made even more complex and obfuscated in GlassFish. The cause is that components spark initialization of other components and they depend on each other, on properties set for each other, so the sequence must be perfectly controlled. Some runtime components are in this library, some are in GlassFish, and the hell is open.

[pzygielo]

I've not looked at ORBImpl yet.