Live data from Hacker News

On Redis master-slave terminology (2018)

antirez.com

81–90 of 162 posts

Re: On Redis master-slave terminology (2018)

#81
post #9

In the meantime, the Slave terminology is on its way out in Redis too. Here are a few commits: https://github.com/antirez/redis/search?q=%22slave+removal%2... Here's the issue where some of the discussion happened: https://github.com/antirez/redis/issues/5335

It's paradoxical that they want these changes merged to "master".

They could always rename the "master" btanch [0] to something else, I suppose.

---

[0]: https://news.ycombinator.com/item?id=23473387

Re: On Redis master-slave terminology (2018)

#82

I suspect this is one of those topics that at some point, everyone will have just switched terms because it's not worth the hassle to fight these fights over and over again. Kind of a "I'm not wrong, but is it a good use of 20 hours a year to argue about this?"

Which means that the perpetually offended will dictate name changes every other year. If somebody argues that they cannot use $software because it contains $word as a variable in its source code somewhere, it's certainly not a good idea to argue with them. It also not a good idea to appease them by doing what they ask. Maybe developers should steer clear of Twitter so they don't get exposed to the problem. The only w…

Yeah? I don't see evidence that every two years a new word gets brigaded in source code. Do you have examples of projects this has been happening to? If not, this seems like a slippery slope fallacy.

Re: On Redis master-slave terminology (2018)

#84
post #41

Earlier quoted context omitted.

Yikes. "Blacklist" has nothing to do with skin colour. In a different context, "being in the black" is a positive situation and contrasted with "being in the red". We start down that path and we'll end up banning "Paint it black" as a racist song.

A blacklist may not have anything to do with skin color, but being on a blacklist usually has negative connotations, not positive. The related term is "whitelist," not "redlist." So I can understand why people might believe the pair has racial overtones.

Well, the function of a blacklist tend to have negative connotations, so that those connotations match how they are actually used seems like a feature, not like something to avoid.

Re: On Redis master-slave terminology (2018)

#85

The project maintainer absolutely has the right to choose not to remove offensive language. Although, I find his position to be unnecessarily defensive. What is he really fighting for? The language is outdated and stems from a practice that is universally shunned. There is work involved in changing the language yet there is no harm in it. People absolutely have the right to be offended. Feeling offended is just that…

> The project maintainer absolutely has the right to choose not to remove offensive language.

What offensive language has been communicated in this case?

> What is he really fighting for?

Wasn't that clearly explained in the post?

> The language is outdated and stems from a practice

Outdated? Has the 'practice' you're referring to dissappeared?

> from a practice that is universally shunned.

What to think of people practicing BDSM? Or bioligists describing insect behaviour? Is their language offensive?

> There is work involved in changing the language yet there is no harm in it.

It is harmful when it deflects from the actual problems we have in our society these days. Just because something is immediately actionable doesn't mean it is effective or even helpful.

> People absolutely have the right to be offended.

Of course people may feel offended, they just shouldn't automatically get additional rights by being offended.

> Changing the language would be more welcoming towards people

I agree with that. But I'm dissapointed that people are content with it. IMO it distracts from more pressing issues that need to be addressed.

Re: On Redis master-slave terminology (2018)

#87

The project maintainer absolutely has the right to choose not to remove offensive language. Although, I find his position to be unnecessarily defensive. What is he really fighting for? The language is outdated and stems from a practice that is universally shunned. There is work involved in changing the language yet there is no harm in it. People absolutely have the right to be offended. Feeling offended is just that…

I would say the opposite. Yes, people have the right to be offended.

But people don’t have a right to NOT be offended. That’s the right that’s being asserted here. You should change something so that I am no longer offended. Well, the fact that one is offended is a struggle that’s personal to them.

Re: On Redis master-slave terminology (2018)

#88

Here we go again. But seriously, what's wrong with using "primary" and "replica" in a database context? They're generally clearer terms anyway. Obviously we'll be stuck talking about "master" and "slave" IDE connectors and the like, but other than that it seems to be falling out of use pretty much everywhere.

> But seriously, what's wrong with using "primary" and "replica" in a database context

Those terms are fine.

The practical argument to change the terms because you think they're clearer isn't a compelling reason to change the terms. But, also, a practical argument doesn't really mix well with the moral argument. (If "master"/"slave" were better understood than alternatives, but agreed to be offensive, would it be better to use those terms?).

I don't see the argument as to being about which terms are best, though. The contention is 'the terms "master"/"slave" are offensive, and should be changed'.

In terms of "what's wrong": this discussion has enough comments in favour of / against the idea to get a sample of what people are disagreeing about. If I understand correctly: "it's a small change to make, that will reduce the exclusion some people feel", and "the terms aren't used in an offensive way, it's not of substantial worth to change it".

Re: On Redis master-slave terminology (2018)

#90
post #82

Earlier quoted context omitted.

Which means that the perpetually offended will dictate name changes every other year. If somebody argues that they cannot use $software because it contains $word as a variable in its source code somewhere, it's certainly not a good idea to argue with them. It also not a good idea to appease them by doing what they ask. Maybe developers should steer clear of Twitter so they don't get exposed to the problem. The only w…

Yeah? I don't see evidence that every two years a new word gets brigaded in source code. Do you have examples of projects this has been happening to? If not, this seems like a slippery slope fallacy.

You're right, master/slave and the new blacklist/whitelist demands did not have two years between them, not even this post is two years old.
Post reply on HN