This sounds to me like something that is already happening. For instance over time we added the "pulsing" optimization to better handle PK lookups, added FST-based suggesters and recently added support for BKD trees in the index formats. I'm pretty sure this will keep happening.
Le dim. 25 sept. 2016 à 14:52, Cristian Lorenzetto < cristian.lorenze...@gmail.com> a écrit : > Lucene changed web applications in the last years. I created a bit > revolution in many real commercial services. > A lot of applications, services, servers based on lucene are born in the > last years. The reason is simple. The modularity is the key for the > computer science fast evolution. It permitted to developers to concentrate > themselves to develop their business logic without to lose years for > getting stable the core of one of the most important functions for their > business. > The necessity to search different ways to store unstructured data was and > is onother focus and success point for lucene. In the future this necessity > could be extended for new needs more complex. > Based on this simple analisys , i suggest to Apache/Lucene to plan in the > future to extend it with other engines implementing other algorithms(b+ > plus,point quad,g-tree,trie,string B) for other application contexts. > It would give a boot to technological development. > > Cristian >
