Discussion on potentially offensive term "black box"

[Karthikeyan Govindaraj]

Hi Folx,

Initiating a thread for discussing the potentially offensive term "black box". Meanwhile, IETF also kind of says `black box` is not an offensive word. (ref: https://tools.ietf.org/id/draft-knodel-terminology-02.html#rfc.section.2.2). Let's start the discussion and arrive at a conclusion for this term.

Potential Offensive Term:

black box

Possible alternatives: [1]

Closed box

Opaque box

mystery box

Ref:

- https://www.duncannisbet.co.uk/removing-harmful-language-from-my-lexicon 

- https://community.infosecinstitute.com/discussion/137715/new-inclusive-terminology-on-comptia-exams 

- https://github.com/spring-projects/spring-framework/commit/9acb5b1a4aba8b8f8cdd070cccfbf43e2e9c5666 

Findings: [2]

Hound: 

- https://cs.k8s.io/?q=blackbox&i=nope&files=&repos=

- https://cs.k8s.io/?q=black%20box&i=nope&files=&repos=

GitHub:

- https://github.com/search?q=org%3Akubernetes+%22black+box%22&type=Code

- https://github.com/search?q=org%3Akubernetes+%22blackbox%22&type=Code

---

Many Thanks,

Karthik

Karthikeyan Govindaraj (he/him)

DevOps Architect | OSS Enthusiast | Blogger

Know Me! | @gkarthics

[1] The possible alternatives are not limited to this list. Please add any additional terminologies as well while we discuss them.

[2] Hound search is only for Codebase, GitHub search included the issues, PRs, etc.

[Jordan Liggitt]

I don't see evidence the term is offensive, though we could recommend alternatives that would likely be clearer. New terms people are not familiar with are potentially confusing, and I don't think it's important to maintain the "box" terminology here.

Hoisting from my response on the earlier thread:

"Black box" is usually used in these contexts:

• "Black box testing", in which tests only observe external behavior, and not the internal implementation. I'd suggest "behavioral testing" as an alternative here.
• A system being a "black box", or unobservable. "Opaque" might be a better alternative here.

--
You received this message because you are subscribed to the Google Groups "Kubernetes WG Naming" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kubernetes-wg-na...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-wg-naming/CAG2XbAu8e--k3pcY6YbiKScXkoB5SpTr4wqRzn-Nc83LQ-a-PQ%40mail.gmail.com.

[Karthikeyan Govindaraj]

I agree with you on the part that it is not offensive. But I still went ahead with this because there are folks in the community who have moved away from this term either they thought as you mentioned it's unclear or 'potentially' offensive. So looking for an async consensus on this.

And yes in our context we don't need to maintain the `box` terminology. 

I vote for removing this term as `confusing` and moving towards "behavioral testing". Looking forward to others' opinions.

[Aaron Crickenberger]

Agree with Jordan's suggestions.

One exemption I can think of is when referring to something outside our project that already has the term in its name, e.g. prometheus' blackbox prober/exporter (https://github.com/prometheus/blackbox_exporter) as we do in test-infra (ref: https://github.com/kubernetes/test-infra/search?q=blackbox)

- aaron

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-wg-naming/CAG2XbAteXMp9TYb7kM4r93KhaJsFeAL1gJwLzBZMXG_c%2BB8SMQ%40mail.gmail.com.

[Karthikeyan Govindaraj]

Ping to @Celeste Horgan to bring to the top of the inbox.

[Karthikeyan Govindaraj]

Hey Folx,

Do you think we can close this thread by agreeing on Jordan's recommendation?

[Celeste Horgan]

Hi all,

I think this is good by me to close.

Celeste