while using sqlite as a data format during the initial tracking has too high overhead, we may need to revise the sqlite branch for visualization. i.e. when loading a heaptrack.*.gz file, store the time data into a sqlite DB, then operate on that to build graphs for time slices