WARNING when a module tries to export to another module but the other module is not yet compiled

[Robert Scholte]

I wanted to figure out what's happening here, because I didn't recognize 
any related message as created by the maven-compiler-plugin nor by one 
of its dependencies.
A small example showed me this:
[WARNING] ..../src/main/java/module-info.java:[2,50] module not found: 
x.y.z
And indeed, the message is actually coming from the javac tool, exposed 
via Maven.

With maven-compiler-plugin 3.11.0 there was an update of one of 
libraries that changed the output a little bit.
The reason behind this is: 
https://github.com/codehaus-plexus/plexus-compiler/issues/183

Related to this there's an open issue at OpenJDK as well:
https://bugs.openjdk.org/browse/JDK-8292634

Not sure if this answers the questions, but as Alex already pointed out: 
your question is mainly about Maven.
I suggest to ask your question via the user mailinglist of Maven.
See https://maven.apache.org/mailing-lists.html for details

regards,
Robert Scholte

------ Oorspronkelijke bericht ------
Van "Alex Buckley" <alex.buckley at oracle.com>
Aan jigsaw-dev at openjdk.org
Datum 19-10-2023 02:30:37
Onderwerp Re: WARNING when a module tries to export to another module 
but the other module is not yet compiled

>This list is not able to help with Maven configuration issues.
>
>However, it's actually allowed for module x to say `exports foo to y;` when module y doesn't exist.
>
>(JLS 7.7.2: "It is permitted for the to clause of an exports or opens directive to specify a module which is not observable (§7.7.6).")
>
>Alex
>
>On 10/18/2023 4:36 PM, Siddharth Jain wrote:
>>I am using JPMS with Maven compiler plugin. say i have:
>>
>>module x {
>>exports foo to y
>>}
>>
>>when i do a clean install I get a WARNING from maven that y does not exist. Its true, the dependency tree has to be built from bottom to top so x will be built first and only then y can be built.
>>
>>i also see strange errors while executing tests where maven clean install will fail with some errors saying type X could not be found but if i re-run the command it passes.
>>
>>i think this error happens because at first y does not exist so JPMS does not know about it. but when i re-run y does exist. is that right?
>>
>>i don't mind the WARNING my question is about the weird errors. has anyone run into same issue?