Near matches maybe slowing/breaking matching multiple strings

[AttaInsularis]

Generated quite a lot with 0.5.4 and while running the same set of -n=name, name2, name3 did observe that have almost no results and that is weird because gaining a lot from that machine for certain time span.

Also should it be more in fact because of that near match, so there are some omitted in older ver.

So trying less and less names to be found, but having now no more than 15 strings - most 4 char long, few 5 and one 6.

With this set up:

For each run, (2 machines, Linux/Win), 10x some string: rana -n=str0, .. ,str -g 12 | tee -a output.txt

0.5.5 - for all runs: quickly there are 2, rarely 3 matches, within a minute usually - then silence (24h run, just add only 2 new entries).
Multiple runs, dozen or so, no more than 3 matches in output file per each new run.
All found are after the start. Usually the last one is exact, if 3 matches than they are to be present with similarity first or second. Not last.

0.5.4 - literally dozens, up to hundred entries in output file after same time. Can work with at least 15 entries if maxing out.

If that being because the wast space of sha that can be generated - but that should be the same for both versions, and so far it is not to my experience.

It just does not add up only 0.5.5 stuck outside of the searched strings while 0.5.4 does not.
Few times maybe, but this is hours and hours.

Does anyone having same issues maybe?

Not sure if that is just slow down or it completely stuck somehow - because no results at all after a while for 0.5.5.

The CPU is 100% all the time while rana running - or according to percentage by the -c flag.