Skip to content

Add a pure Ruby implementation#21

Draft
tenderlove wants to merge 1 commit intoGargron:masterfrom
tenderlove:pure-ruby
Draft

Add a pure Ruby implementation#21
tenderlove wants to merge 1 commit intoGargron:masterfrom
tenderlove:pure-ruby

Conversation

@tenderlove
Copy link
Contributor

Hi,

I ported the C implementation to pure Ruby (the code is not pretty since it's pretty much a straight port from the C code). I don't know how fast this is compared to the C implementation. I did this because I think it would make a good JIT benchmark, especially since we can compare against the C implementation.

The commit doesn't replace the existing C implementation, it just adds an encode_rb function. I don't expect this PR to be merged necessarily, I just wanted to send it and say "hey, there's a pure Ruby version if you want it! 😄"

@tenderlove tenderlove force-pushed the pure-ruby branch 2 times, most recently from d463669 to 8883456 Compare January 6, 2023 23:41
We should try using this for benchmarks
end

def set(y, x, z, val)
i = z + (x * @z) + (y * @z * @x)
Copy link

@Nakilon Nakilon Jan 9, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi. I'm just randomly subscribed to this repo and was interested by seeing such pull request.
Just wanted to say that AFAIK this won't do any profit in Ruby, you better just have real 3d array. Not sure if there are benchmarks here but I bet on this.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't really want to rewrite this to use a "real 3d array". You're welcome to do it and test the performance. My guess is that they'd be pretty much the same, or this version is faster. Why?

  1. A real 3d array will have many more objects, causing more time in GC
  2. A real 3d array will require more pointer chasing (we would have to dereference more objects)

In the presence of a JIT compiler, this version will no doubt be faster (math is much cheaper than reading memory). Again, you're welcome to test!

@sandstrom
Copy link

@Gargron Friendly ping 😄

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants

Comments