bereng commented on code in PR #1891:
URL: https://github.com/apache/cassandra/pull/1891#discussion_r1129010090
##########
src/java/org/apache/cassandra/db/RangeTombstoneList.java:
##########
@@ -58,12 +58,12 @@ public class RangeTombstoneList implements
Iterable<RangeTombstone>, IMeasurable
private ClusteringBound<?>[] starts;
private ClusteringBound<?>[] ends;
private long[] markedAts;
- private int[] delTimes;
+ private long[] delTimes;
Review Comment:
Sure let's leave this one open.
Iirc having a int and a boolean pair is going to be the same length (4+4) as
having long (8). But we could have an int array of the invalid ldts indices in
the first array. As this will be 99.9% of the time just a null, we shouldn't
have invalids, so it seems like a possible solution.
I think I'll revert given we're entering the early optimizations world and
at cost & complexity. The underlying idea here, imo, is that if this ever
becomes a problem (or some other area of the code) we seem to be able to come
up with reasonable alternatives.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
