io7m opened a new issue, #53:
URL: https://github.com/apache/felix-atomos/issues/53

   Somewhat related to #52, it seems that it's not really possible to write a 
straightforward "host" program that's JPMS modularized. There are conflicts 
between the various OSGi artifacts that cause a number of issues.
   
   I don't doubt that it would be possible to work around them using `bnd` and 
various class and module-rewriting magic, but I feel like it really should be 
possible to at least get a program to build and run without them.
   
   I've put together a trivial example here that simply tries to start up an 
essentially empty framework from a `main` method without any tooling:
   
   https://github.com/io7m/atomos20220505
   
   Note that in order to compile the code, there needs to be a dependency on 
the `osgi.core` module (due to dependencies on the core classes by the 
Framework API). If the `osgi.core` module isn't on the module path, the build 
obviously fails. If the `osgi.core` module _is_ on the module path, the build 
succeeds but the JVM then refuses to start due to multiple modules having the 
same packages.
   
   I suspect I might be in the minority with trying to start up an OSGi 
container from a JPMS program, but... It's pretty bad that it's apparently not 
possible.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to