Hacker News

points by goodmythical 3 days ago | 0 comments

Genuinely confused here...release notes say improved performance, so kitten _benchmark__, right?

kitty 0.48.2 provided by fedora:

Only ASCII chars : 1.25s @ 159.8 MB/s

  Unicode chars                        : 1.05s      @ 172.8   MB/s

  Unique multi-codepoint Unicode cells : 5.17s      @ 33.9    MB/s

  CSI codes with few chars             : 1.39s      @ 71.9    MB/s

  Long escape codes                    : 1.42s      @ 551.3   MB/s

  Images                               : 1.02s      @ 524.4   MB/s

kitty 0.50 provided by install.sh:

  Only ASCII chars                     : 4.72s      @ 211.8   MB/s

  Unicode chars                        : 4.32s      @ 208.1   MB/s

  Unique multi-codepoint Unicode cells : 24.9s      @ 35.1    MB/s

  CSI codes with few chars             : 5.08s      @ 98.4    MB/s

  Long escape codes                    : 7s         @ 560.1   MB/s

  Images                               : 4.88s      @ 546.0   MB/s

I've run each multiple times and the results are similar...the new version seems to have higher throughput and yet takes longer to finish the tasks? Like, a lot longer?

[–] goodmythical 3 days ago

Looks a little bit like my misunderstanding as there is some suggestion that the new kitten __benchmark__ (25.4MB in the 0.48 install, 30.1MB in the 0.50 install) includes more data to more thoroughly benchmark but I can't find any citation for that and they're binaries so that's beyond my scope to inspect.

edit: Does kitten __benchmark__ rely on benchmark.py? if so comparing them on github does reveal lots of changes but I don't know where in ~300 lines changed that it indicates any of the benchmarks should be taking ~5x to complete.