explore feels slow in big tables #26

Activity

1. amtoine

   added

   bugSomething isn't workingSomething isn't working

   performance

   on Sep 4, 2023
2. amtoine commented on Sep 4, 2023

   

   amtoine

   on Sep 4, 2023

   Last edited by amtoine

   OwnerAuthor

   More actions

   if the history is big enough, one can compare

   open $nu.history-path | query db "SELECT * from history" | first 20000 | explore

   and

   open $nu.history-path | query db "SELECT * from history" | first 200 | explore

   alternately, one can use the following command to generate a dummy table of a certain length

   def t [n: int] {
       let ids = seq 1 $n | each { random uuid } | wrap id
       let names = seq 1 $n | each { random chars --length 64 } | wrap name
       let vals = seq 1 $n | each { random int } | wrap val
   
       $ids | merge $names | merge $vals
   }

   it starts to feel slightly slow with t 10_000, and it gets worse with t 20_000 and t 30_000 freezes.

   Note

   also works with seq 1 $n with $n = 10^{5}$