Hi! I'd like to gather your opinion on replacing Guava with the phoenix-thirdparty shaded version on the 4.x branch.
The change was not applied on 4.x, as the incompatibilities that the pre-shaded Guava version solves are not present on Hadoop 2/Hbase 1. However, the difference between the branches causes extra work, and frequent problems when porting code between the branches. Applying PHOENIX-6010 to the 4.x branch would make our lives easier, while having no negative effects that I know of (we already pull in phoenix-thirdparty via Omid 1.0.2) looking forward to hearing your opinion Istvan