On 8/2/23 13:30, David Smiley wrote:
I wonder if it's worth it porting the 8x codeline into our build infrastructure for 9x? i.e. use our existing Gradle build instead of Ant. This way subsequent releases would be easier (if needed), would be easier for anyone maintaining a fork of 8x, and we needn't increment Lucene releases just for Solr's needs.
How difficult would it be to actually reach a workable gradle build for 8.x? I suspect it would not be anything close to trivial, and would probably take a long time to fully troubleshoot. Whoever was the major architect of the current gradle build should be able to tell us whether that's a massive undertaking or not. I wasn't involved in any of that work.
The ant build isn't broken, and will not be needed for very many more 8.x releases, so I think it's probably best let it continue as it is. That effort is probably better spent getting 10.0 in shape for eventual release.
