This PR bumps the version of spine-logging to the latest one, in which the published library is operational again.
Also, it substitutes a runtime dependency on Flogger's JUL backend with Spine Logging. In general, specifying of a backend makes no sense unless a facade is used. I think we can go with the following options:
Configure Spine logging by default with our facade and JUL backend. This option is employed in this PR.
Doesn't pre-configure logging at all. Remove specifying of Flogger's backend.
We potentially can detect which facade is used, and then specify the corresponding backend in jvm-module. But this will have to be done with project.afterEvaluate { ... } closure. Because if jvm-module doesn't dictate the used facade, then a concrete module is meant to dictate it. A script plugin is applied first, it can't obtain module-configured dependencies in advance. Thus, I would rather avoid this sort of tricks that tampers with Gradle configuration phases.
The version of org.ow2.asm:asm transitive dependency has been forced to prevent conflicts.
This PR bumps the version of
spine-logging
to the latest one, in which the published library is operational again.Also, it substitutes a runtime dependency on Flogger's JUL backend with Spine Logging. In general, specifying of a backend makes no sense unless a facade is used. I think we can go with the following options:
We potentially can detect which facade is used, and then specify the corresponding backend in
jvm-module
. But this will have to be done withproject.afterEvaluate { ... }
closure. Because ifjvm-module
doesn't dictate the used facade, then a concrete module is meant to dictate it. A script plugin is applied first, it can't obtain module-configured dependencies in advance. Thus, I would rather avoid this sort of tricks that tampers with Gradle configuration phases.The version of
org.ow2.asm:asm
transitive dependency has been forced to prevent conflicts.