Ok, I guess you can ignore this question because I think the conclusion is
that bootstrap cannot be used in JPMS at all.

I have figured out a boot process-ish and I can see things a little more
clearly now.

- Ray

On Sat, Jun 13, 2020 at 9:12 PM Raymond Auge <raymond.a...@liferay.com>

> Forgive me if this is a well worn topic.
> - Ray
> On Sat, Jun 13, 2020 at 8:58 PM Raymond Auge <raymond.a...@liferay.com>
> wrote:
>> Hello Tomcat Devs,
>> Is it expected that when Bootstrap.init() invokes
>> "org.apache.catalina.startup.Catalina"'s constructor (which it finds in
>> catalina.jar) via reflection through catalinaLoader that some of the
>> classes in the package org.apache.catalina.startup, originating in
>> bootstrap.jar, are still used by classes loaded by the new classloader?
>> The result is that there are now two jars in runtime loaded by two
>> separate containers sharing classes from the same package.
>> I just want to make sure this is by choice and not accidental because
>> this is going to cause problems for both JPMS and OSGi and I need to figure
>> out how to deal with it. The first thing that might need to happen is that
>> bootstrap may need to export some of it's packages which will certainly
>> result in a split package scenario (since other modules already provide
>> those same packages). As you can guess in both module systems split
>> packages are a challenge.
>> Thx
>> --
>> *Raymond Augé* <http://www.liferay.com/web/raymond.auge/profile>
>>  (@rotty3000)
>> Senior Software Architect *Liferay, Inc.* <http://www.liferay.com>
>>  (@Liferay)
> --
> *Raymond Augé* <http://www.liferay.com/web/raymond.auge/profile>
>  (@rotty3000)
> Senior Software Architect *Liferay, Inc.* <http://www.liferay.com>
>  (@Liferay)

*Raymond Augé* <http://www.liferay.com/web/raymond.auge/profile>
Senior Software Architect *Liferay, Inc.* <http://www.liferay.com>

Reply via email to