FTR the main support based on maven historical configuration is trivial and fast to do with a postprocessing, the release flag sounds great but until i missed a trick to not have it moving files/abusing IO it is way slower than validate so if abused (lot of files) it will be a perf killer (even for 10 entries it is slow compared to not using it), not sure about the debugging, previous code was trivial, new one is easy even if some will dislike the partly function style, just habit? multimodules can be handled with previous code so not sure the issue
don't get me wrong, i'm not against the move and i'd be 1000% for not having a wrapper at maven and reuse what the JDK provides, but there perfs are sometimes crazy, for 10k classes with --release you are 4 times slower for ex. Romain Manni-Bucau @rmannibucau <https://x.com/rmannibucau> | .NET Blog <https://dotnetbirdie.github.io/> | Blog <https://rmannibucau.github.io/> | Old Blog <http://rmannibucau.wordpress.com> | Github <https://github.com/rmannibucau> | LinkedIn <https://www.linkedin.com/in/rmannibucau> | Book <https://www.packtpub.com/en-us/product/java-ee-8-high-performance-9781788473064> Javaccino <https://javaccino.dev/> founder (Java/.NET service - contact via linkedin) Le mer. 26 août 2026 à 22:34, Martin Desruisseaux via dev < [email protected]> a écrit : > Le 26/08/2026 à 22:05, Romain Manni-Bucau a écrit : > > > oki, let's disable validate by default then but it also makes hard to > > justify the drop of the old impl IMHO, we might want to "milestone" it > > right in terms of comm no? > > We still have the validation when using the --release options. In that > case, the validation is done at JAR creation time and does not require a > second pass. We also have the --main-class option that I mentioned in > previous email. In a modular JAR, it is not just an attribute in > MANIFEST.MF. We still have the easier debugging. And finally, this pull > request is critical for multi-modules. Without this pull request, all > the work for Module Source Hierarchy support is lost. > > Martin > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > >
