chat-wane / lseqarray Goto Github PK

The identifiers of LSEQ starts from low numbers but can grow very quickly. Unfortunately, when it grows, the number of sources and counters grows too. Still, the number of duplicates in these arrays are high. As a consequence, an order preserving compressing algorithm would drastically improve the space used by these arrays without impacting on the time complexity of algorithm.

It implies modifying: (1) the getSC method to copy only useful information, (2) the comparison function between identifier.

BPlus strategy seems a lot more efficient than BMinus. Due to comparison ?

Currently, the parameters are set in the config.js file. Multiple instances of created class refer to it. Nevertheless, these instances should be factorized in a singleton.

Parameters are: boundary, base, site.

All the distributed nodes must have the same boundary and base values, but different site identifiers.