Revision history for Meow
0.01 2025-05-26
- Experimental release
0.08 2025-05-27
- Fix pod encoding issue
- Attempt to fix issues on windows
0.09 2025-05-27
- Change to use any_ptr instead of a hash variable
0.10 2025-05-27
- dTHX for malloc
0.11 2025-05-27
- Add ro keyword to generate readonly attributes.
- Add Default keyword so you can set default for attributes.
0.12 2025-05-27
- Add an extends keyword for multiple inheritance
0.13 2025-05-27
- Add Coerce keyword for attribute coercions
- Add Trigger keyword for attribute triggers
- Add Builder keyword for lazy building attributes.
- update POD
0.14 2025-05-27
- Update benchmark to show Mouse is faster
- Improve test coverage
0.15 2025-05-27
- remove sv_sethek usage
0.16 2025-12-20
- Update the benchmark to include Marlin.. it's closer than Meow..
0.17 2026-01-23
- Meow is now faster than Moo and Moose but you should probably all be using Marlin...
- Meow 2222222/s 153% 36% -- -69%
- Marlin 7142857/s 714% 336% 221% --
0.18 2026-01-23
- Switch to a just in time approach via Inline::C
- Updated benchmarks but limit is now hardware
- Meow: 0 wallclock secs ( 1.10 usr + 0.00 sys = 1.10 CPU) @ 4545454.55/s (n=5000000)
0.19 2026-01-24
- XS::JIT brings the unnecessary need for speed..
- Meow: 1 wallclock secs ( 0.96 usr + 0.00 sys = 0.96 CPU) @ 5208333.33/s (n=5000000)
0.20 2026-01-24
- Move to slots which reduces overhead... overhead maybe in XS callstack itself now unsure :)
- Meow: 5 wallclock secs ( 5.30 usr + 0.01 sys = 5.31 CPU) @ 6277761.96/s (n=33334916)
0.21 2026-01-24
- Expose slot variables and then show accessing by slot index is actually faster than pure python..
- limit with accessors is in perl stack call not sure how to get around that one yet..
- Meow (direct): 5 wallclock secs ( 5.00 usr + 0.02 sys = 5.02 CPU) @ 7230451.20/s (n=36296865)
0.22 2026-01-24
- The unholy op's...
- Meow (op): 5 wallclock secs ( 5.16 usr + 0.01 sys = 5.17 CPU) @ 7394453.19/s (n=38229323)