it looks like jimmy was actually limited by the options I used (-R 4000, and it hit tps = 3999.541557) so I guess it can go faster, I'll re-run it with a higher target
those were using the default scaling factor of 1, so 100,000 rows in the test table. I'm trying again with 1,000,000 rows
jesus2099 has quit
in that test, tps went up for both (I guess increasing the number of rows past some threshold affects how indexes are used and queries are parallelized more efficiently or smth) https://gist.github.com/mwiencek/2cc9681740fd35...
but jimmy still achieves about 3.5x more tps than pink (3.9x in the first test) :)
oops, that test was with 200,000 rows, not 1,000,000
It might seem nitpicky, but details like border weight and colour make all the difference in a finished page/UI :)
If you go to your LB profile > Add listens > Manual addition you can see a search box style that you can possible take from? It’s a really good idea to re-use CSS styles wherever possible