Most active commenters
  • simonw(6)

←back to thread

Ruby Central's Attack on RubyGems [pdf]

(pup-e.com)
659 points jolux | 12 comments | 19 Sep 25 08:09 UTC | HN request time: 1.074s | source | bottom
Show context
krmbzds ◴[19 Sep 25 13:05 UTC] No.45301255[source]▶
The recent actions taken by Ruby Central - removing long-time RubyGems and Bundler maintainers without warning, seizing administrative access, and consolidating control under a small, centralized group - represent a serious breach of trust within the Ruby ecosystem.

This was not a misunderstanding. It was a hostile takeover of key infrastructure, undermining both the long-standing maintainers and the broader community that relies on RubyGems and Bundler every day.

The Ruby ecosystem thrives on collaboration, openness, and mutual respect. What we've witnessed over the past week violates those principles. Ruby Central's actions - unilateral access revocations, exclusion of experienced volunteers, and refusal to engage in transparent dialogue - are not just organizational missteps. They're a threat to the decentralized and community-driven spirit that has sustained Ruby for decades.

I oppose this power grab.

Even more concerning is the idea that contributor access could become contingent on employment status or ideological alignment. Whether someone is employed by Ruby Central - or holds left-leaning, right-leaning, or apolitical views - should have no bearing on their ability to contribute to open source. Merit, dedication, and community trust must remain the foundation.

If Ruby Central is serious about supporting the Ruby community, they must:

- Immediately restore access to all maintainers removed during this incident.

- Publicly commit to a transparent, community-driven governance model, similar to what the RubyGems team had begun drafting.

- Respect the autonomy of open source maintainers, regardless of whether they are employed by Ruby Central.

- Acknowledge the harm caused by these actions and engage in meaningful dialogue to rebuild trust.

The Ruby community has always been about people - diverse, passionate, and united by a love for a beautiful language. It's time we demand that the institutions claiming to represent us act accordingly.

And if Ruby Central does not do this we must pressure sponsors to stop funding Ruby Central and ultimately; if all else fails, we must build and maintain our own infrastructure unencumbered by these shenanigans. Also, in order to re-establish trust in the community; the people responsible for causing this ruckus should be fired.

Ruby-Level Sponsors (Top Tier): Alpha Omega, Shopify, Sidekiq

Gold-Level Sponsor Flagrant

Silver-Level Sponsors: Cedarcode, DNSimple, Fastly, Gusto, Honeybadger, Sentry

replies(6): >>45301272 #>>45301522 #>>45301533 #>>45301996 #>>45302057 #>>45302537 #
1. simonw ◴[19 Sep 25 14:21 UTC] No.45301996[source]▶
Why did you include that list of sponsors at the bottom of your post?

What's with the "contingent on employment status or ideological alignment" bit about? That's not been mentioned anywhere else so far.

Were those parts (or indeed your entire comment) written with the help of an LLM?

replies(3): >>45302162 #>>45302331 #>>45306847 #
2. UseofWeapons1 ◴[19 Sep 25 14:37 UTC] No.45302162[source]▶
The post is quite clear? They call on the sponsors to stop funding ruby central, and the employment status bit is a clear concern extending from ruby central’s supposed takeover.

Read the post more clearly before accusing someone of LLM usage. And even if it is, they are still valid points to be discussed, as opposed to trying to bury it with an LLM accusation.

replies(1): >>45302333 #
3. dragonwriter ◴[19 Sep 25 14:53 UTC] No.45302331[source]▶
> Why did you include that list of sponsors at the bottom of your post?

Clearly, that was because this information directly supports readers following through on the call to action: “And if Ruby Central does not do this we must pressure sponsors to stop funding Ruby Central”. That’s obvious.

> What's with the "contingent on employment status or ideological alignment" bit about? That's not been mentioned anywhere else so far.

Yes, both the original pdf and the RubyCentral statement edplicitly refer to admin status being made contingent on being full-time employee of RubyCentral. If you just mean no one has explicitly brought upthe ideological angle, well, that’s a fairly easy concer to reach wrih something being contingent on employment at a particular nonprofit, so it would be weird to interogate like this even if you had clearly focussed on kn just that point.

replies(1): >>45302372 #
4. simonw ◴[19 Sep 25 14:53 UTC] No.45302333[source]▶
I brought up LLM usage precisely because the two things I called out here are weird - the kind of details an LLM might add.

If that's what happened then it's bad because it leaves people who read the comment confused - hence my questions asking about those.

If the author confirms that those pieces I asked about serve an intentional purpose then I don't care if they used an LLM or not.

My problem isn't with using LLMs to help write comments - there are plenty of reasonable reasons for doing that (like English as a second language). My problem is letting an LLM invent content that doesn't accurately represent the situation or reflect the LLM user's own position.

(The author could also say "I didn't use an LLM", which notably they haven't done elsewhere on this thread yet.)

replies(1): >>45302560 #
5. simonw ◴[19 Sep 25 14:57 UTC] No.45302372[source]▶
Where did the ideological alignment piece come from then?
replies(1): >>45302705 #
6. cxr ◴[19 Sep 25 15:14 UTC] No.45302560{3}[source]▶
What content has been invented?
replies(1): >>45304823 #
7. krmbzds ◴[19 Sep 25 15:27 UTC] No.45302705{3}[source]▶
You can read it here: https://world.hey.com/dhh/no-railsconf-faa7935e

The cancellation of DHH's keynote was purely political. At that time, RubyCentral's response was similarly uncommunicative and their explanation was BS.

This is not the first strike.

8. simonw ◴[19 Sep 25 18:26 UTC] No.45304823{4}[source]▶
Maybe none? That's why I asked.
9. angoragoats ◴[19 Sep 25 21:27 UTC] No.45306847[source]▶
> Why did you include that list of sponsors at the bottom of your post?

The paragraph immediately preceding the list begins with a sentence mentioning the sponsors. How did you not see this?

> What's with the "contingent on employment status or ideological alignment" bit about? That's not been mentioned anywhere else so far.

“not been mentioned anywhere else” is false. If you click on the PDF linked to in this very post it mentions that only full time employees of RubyCentral maintained access to their GitHub account.

I find it ironic that you’re so quick to question whether something is LLM-authored given that you write so much about using LLMs.

replies(1): >>45307912 #
10. simonw ◴[19 Sep 25 23:14 UTC] No.45307912[source]▶
I'm quick to question precisely because I can see LLM telltale hints in the text - in this case the "not just X but Y" pattern.

I don't mind if it's LLM-assisted text if everything in it is a reviewed and accurate representation of the point the author is trying to make.

But if the LLM throws in extra junk tha distracts from the conversation and the author fails to catch that in review, that's bad.

I think it's likely I was mistaken here - that the author either didn't use an LLM or used it for minor style tweaks but ensured that it was making the points they wanted to make.

replies(1): >>45312197 #
11. angoragoats ◴[20 Sep 25 10:52 UTC] No.45312197{3}[source]▶
There is no “extra junk that distracts from the conversation” in the post you’re responding to, and I and others are trying to point that out to you. You don’t seem to be responding to those points much if at all.
replies(1): >>45313512 #
12. simonw ◴[20 Sep 25 14:02 UTC] No.45313512{4}[source]▶
I said "I think it's likely I was mistaken here", what more do you want?

(Personally I'd still like to see the author clarify if they used an LLM or not, but that's more for my own personal curiosity at this point, to check if my radar needs adjusting.)