dsmiley commented on issue #1351:
URL: https://github.com/apache/lucene-solr/pull/1351#issuecomment-618441539
Okay; I get your point about noise. I think it's also true that smaller
(realistic) data sets may expose how well code can scale down and make
different / better choices than it
dsmiley commented on issue #1351:
URL: https://github.com/apache/lucene-solr/pull/1351#issuecomment-616579840
@mikemccand I'm curious about your statement:
> The wikimedium1m corpus is really too small to draw strong conclusions --
I would use it to run a quick performance test,