Hacker News Reader

Comment by oneeyedpigeon | Hacker News Reader

Did anyone ever make a claim that master —> main was going to have a significant, direct impact on the fight against racism? Obviously, it's a step, but it seemed like an entirely positive thing to me (shorter, more semantic) with essentially zero downside.

View on HN

> Did anyone ever make a claim that master —> main was going to have a significant, direct impact on the fight against racism? Obviously, it's a step,

It's not obvious at all. The music industry has been using the term "master" for ages: master bus, master copy, mastering engineer. (It's worth noting this is the exact etymology of the git term, too, as opposed to master in the context of master and slave.) The Black community has overall had zero problem with this, and it hasn't stopped many Black musicians and artists from being successful. The music industry also had none of the "we have to rename the master bus" nonsense that came from the tech industry, despite, or more likely because, a significantly larger proportion of BIPOC and LGBT minorities being involved in music.

I don't care about the master vs main name in the abstract. Either branch name is a fine choice. It just completely doesn't matter and pretending that it has any positive impact or meaningful change on the racism and discrimination faced by Black people in America is insulting. It's purely driven by self-indulgent white people who don't want to make material changes to their own extremely comfortable lives while pretending they're fighting the good fight.

> Either branch name is a fine choice. It just completely doesn't matter and pretending that it has any positive impact or meaningful change on the racism and discrimination faced by Black people in America is insulting. It's purely driven by self-indulgent white people who don't want to make material changes to their own extremely comfortable lives while pretending they're fighting the good fight.

Exactly, and I agree with this completely as a Slavic person (from which the word slave is derived). I frankly consider this insulting as well as having a great grandparent used for forced labour in the Ukrainian Canadian interment camps during WWI, and a grandparent in the German forced labour camps during WWII.

If only there was no downside, but you can't change 15+ years of convention and not break something in the process. For example, I've had Homebrew upgrade fail because someone thought the "master" branch must not only be renamed but also permanently eradicated from a cask repo, Yocto had the same issue [1], etc.

[1] https://github.com/ostreedev/ostree/issues/2360

Reddit died for a number of hours because of this, too.

https://www.reddit.com/r/RedditEng/comments/11xx5o0/you_brok...

> The nodeSelector and peerSelector for the route reflectors target the label `node-role.kubernetes.io/master`. In the 1.20 series, Kubernetes changed its terminology from “master” to “control-plane.” And in 1.24, they removed references to “master,” even from running clusters. This is the cause of our outage. Kubernetes node labels.

Yep, this has been a constant pain point for me nearly every day. Having dozens of repos I work in across multiple organizations, where half follow the old convention (master) and the other half follow new (main), when I want to switch branches I usually have to do it twice. First `git co main` and if that doesn't work, `git co master`. Vice versa doesn't work because older repos that have been converted have a master branch! It's usually way behind. There are also tons of scripts and CI/CD yaml everywhere that has to be modified for main vs. master.

Main is a better name IMHO (increased clarity and brevity) but it is far from "zero downside."

It should have been built into the client so that if a co of main fails, it tries master. But that would be too easy.

> Obviously, it's a step

Not sure if it is at all obvious.

Reply on news.ycombinator.com