Reworking Memory Management in Cruby

Posted3 months agoActive3 months ago

ksec

38 points

8 comments

railsatscale.comTechstory

calmpositive

Debate

20/100

RubyMemory ManagementPerformance Optimization

Key topics

Ruby

Memory Management

Performance Optimization

The article discusses improvements to CRuby's memory management, and the HN discussion explores the technical details and implications of these changes.

Snapshot generated from the HN discussion

Discussion Activity

Light discussion

First comment

Peak period

108-120h

Avg / period

Key moments

01Story posted
Oct 12, 2025 at 8:18 AM EDT
3 months ago
Step 01
02First comment
Oct 17, 2025 at 5:49 AM EDT
5d after posting
Step 02
03Peak activity
4 comments in 108-120h
Hottest window of the conversation
Step 03
04Latest activity
Oct 17, 2025 at 2:52 PM EDT
3 months ago
Step 04

Generating AI Summary...

Analyzing up to 500 comments to identify key contributors and discussion patterns

Discussion (8 comments)

Showing 8 comments

Alifatisk

3 months ago

1 reply

> One of the superpowers of MMTk is that it supports parallelism in the garbage collector. Unlike Ruby’s default garbage collector

Wow

sebiw

3 months ago

2 replies

"Wow" as in "why doesn't Ruby have that already" or as in "parallelism will be great"? :-)

Lio

3 months ago

When we say "Ruby" doesn't have parallel GC we're really just talking about MRI cRuby.

Both the JRuby and TruffleRuby implemenations have had parellel GC for a very long time.

I think Peter Zhu work on cRuby has been brilliant here and is really appreciated.

Thev00d00

3 months ago

Why not both :D

schneems

3 months ago

1 reply

It’s not even remotely close to the main point of the article but: I really like the Chris Seaton callout. I talked about this initiative with him the RubyConf before he passed.

Open Source contributors and maintainers feel that there is a constant pressure to be producing. That somehow our worth is somehow only as good as our recent patches. That note says to me: that’s not the case. That even after the contributions stop, or we do, that our efforts and impacts are felt and remembered.

ksecAuthor

3 months ago

1 reply

>I really like the Chris Seaton callout

I miss him. And I really hope someday TruffleRuby will succeed and become mainstream or at least 2nd only to Cruby.

schneems

3 months ago

I miss him too. We knew each other online for some time, but that was the first time I ever met him in person. I came back and couldn’t stop talking about him to my wife when she asked how the conf went. Truly an inspiring experience.

I hope truffle ruby does well, too. But even if it doesn’t, he’s still made an impact on me, and on the community. I told him “thank you” to his face when we met. I also wish I could have also somehow let him know my thanks and appreciation was unconditional.

pjmlp

3 months ago

As historical background information, MMTk was initially implemented in Java, and used on the JikesRVM VM, one of the very first bootstraped Java implementations.

https://www.jikesrvm.org/UserGuide/MMTk/index.html

Eventually they rewrote it in Rust,

https://www.mmtk.io/

Thus it is much more than the 5 years mentioned on the post, that only covers the rewrite.

View full discussion on Hacker News

ID: 45557669Type: storyLast synced: 11/20/2025, 2:24:16 PM

Want the full context?

Jump to the original sources

Read the primary article or dive into the live Hacker News thread when you're ready.

Open link View on HN