Most active commenters
  • downquarklove(4)
  • type0(3)

34 points atoledano | 40 comments | | HN request time: 1.505s | source | bottom
1. ggggtez ◴[] No.17955083[source]
Complete with someone saying we shouldn't change anything to preserve the history and culture of the older generation.
2. knaik94 ◴[] No.17955115[source]
If they decide to go through with the change, I hope different languages and program agree on the new term.

I personally believe that lack of clarity is the only reason these kinds of changes should be made.

I assumed that master/slave terminology came from the idea of copying from a master record vinyl, am I missing something?

I just wonder if this will set a precedent of controversy about the most innocuous things, just to make sure it's politically correct.

replies(2): >>17955163 #>>17955751 #
3. type0 ◴[] No.17955152[source]
Thou shalt not kill!

I hope that Python community will show some respect to the religious programmers and follows by removing all statements regarding the use of the word "kill" as well /s

4. ggggtez ◴[] No.17955163[source]
Master slave comes from slavery. Generally the master doles out work for the slaves to do.

Specifically, even if the phrase is based on an earlier technology, what is that technology's terminology based on? There is no other answer but actually slavery at the end.

replies(1): >>17957723 #
5. svrtknst ◴[] No.17955269[source]
I like this. It's a pretty small change, for the better. It doesn't (should't...) disturb anyone, but might make things just a little bit better for some others. Remove some slight friction.

A small, positive change.

6. badrabbit ◴[] No.17955320[source]
No problem against this,good if it makes anyone more comfortable in the community.

Master/slave relationship is not uncommon throught history. I think if someone learned history well this wouldn't offend them all that much. But still,this isn't a misunderstanding worth making anyone uncomfortable. When I was a teenager,I remember similarly feeling uncomfortable with IDE interface's master/slave terminology.

replies(3): >>17955500 #>>17956040 #>>17956044 #
7. lurquer ◴[] No.17955354[source]
Not nearly as offensive as ORG in assembly.
8. grep-F ◴[] No.17955424[source]
I don't like that type of changes, it seems very counterproductive. Those are words used in a context and to me it seems that today people lost their minds and trying to keep everybody in perfect hermetic world, where everybody is happy and smiling. On one side we say to not judge anybody or anything, and on the other we do things like that - where if you would get the context everything would be ok. What will be next? Should we change kill command as well?
replies(1): >>17955629 #
9. commandlinefan ◴[] No.17955500[source]
... bad if talented people who could be doing something more important (which includes absolutely anything) are forced to waste their time either doing this or arguing why they shouldn't be.
10. cpburns2009 ◴[] No.17955551[source]
Checks calendar, notices it's early September.

Is contributing to an open source project by any chance an assignment in some college class? Either that or they must be having a really slow couple days over at metaphorical Python HQ where bike-shedding has been taken up to pass the time.

replies(2): >>17955884 #>>17955958 #
11. UncleMeat ◴[] No.17955629[source]
The goal isn't to create a perfect world. The goal is to create a slightly better world at extremely minimal cost, given that there are plenty of workable alternative terms. I agree that this affects a very small number of people. But it does affect them and we have an opportunity to make the world just a tiny bit more kind.
replies(1): >>17955761 #
12. kevin_b_er ◴[] No.17955751[source]
It is a clbuttic problem. Their politics demands extricating the world "master" from speech because of its supposed connection to slavery. Unfortunately things like "master record vinyl" would lead to "main record vinyl" or "parent record vinyl". This makes it a clbuttic problem. And it isn't that the "pair" must exist. Things that are expressed in the "master record" concept will thus need to be clubuttic'ified.

Why it this? Because it is not motivated by code improvement but by political demands. The clarity of meaning will be lost by these changes for future developers, because of the need to censure the word "master" from speech and writing.

And yes, this sets a precedent that all you need to do find more words on the euphemism treadmill that may be labeled offensive to force their replacement. This will result in further loss of clarity in documentation where clarity can be reduced whenever someone demands that "for diversity reasons" words need to be deleted.

13. grep-F ◴[] No.17955761{3}[source]
I don't agree, I think some people are being overprotective.

Let's go the other way around, how many words that might sound bad, in how many languages, exists in all the codebases. And now, how many people feeling bad about this I have to find to make change to some code? I know that sounds ridiculous, but I want to say, hey where's the limit, should we all speak using some machine - neutral words?

And another argument, what if that kind of change could brake my project, do somebody care about me?

replies(2): >>17956673 #>>17956826 #
14. wuunderbar ◴[] No.17955794[source]
Some thoughts on this from a core Redis maintainer:

http://antirez.com/news/122

15. legostormtroopr ◴[] No.17955822[source]
Let’s sidestep the fact that most nations and races have been subjected to slavery, so using the term Slave doesn’t refer to any race - except perhaps the Slavs, who were enslaved so much through history, that’s where the word slave came from. Also, lets forget about the tens of thousands of people who currently live in slavery across the world.

The use of master/slave quickly describes unique a relationship. One orders, the other obeys blindly.

What alternatives are there:

Parent/child - no child obeys willingly, and this implies a hierarchy or inheritance where the parent and child are similar

Leader/follower - this implies the follower will eventually catch-up, great for databases, not so much for describing work allocation.

Aplha/Omega - this was in that comment chain, without going off and learning about wolf social structures, I’d have no clue what this is. It’s just foolish.

Master/slave terminology is used as a metaphor to describe a technical behaviour, it’s not an implicit approval of past human behaviour.

The only people who recommend this change are virtue signalling - they want to look like their are doing a great good, and they pick an easy target. No doubt when confronted with the choice of “change this term, or we call you exclusionary/non-diverse/racist”, people will reluctantly choose the former every time. Which means the signaller gets what they want, change and the ability to say “I did something” not matter how trivial.

replies(4): >>17956661 #>>17956802 #>>17959557 #>>17959595 #
16. nitwit005 ◴[] No.17955884[source]
I'm not sure about contributing, but you can definitely see classes forking repos in github. I'm sure some fraction of them try to get their changes in, even if it's not required.
17. x1798DE ◴[] No.17955958[source]
The OP of that bug is one of the most active core developers of CPython and has been for almost a decade...
18. rtpg ◴[] No.17956040[source]
Not only is the term problematic, in many usages it’s not even really right! In the IDE example the master/slave terminology is used just for coordination of a communication pipeline, and has very little to do with the master determining what the slave is doing. Primary/secondary would actually be better semantically
19. peterhadlaw ◴[] No.17956071[source]
Similar nonsense came up not too long ago so I'm just going to borrow my previous post:

If you want to make progress in tech, don't let "progress" (in quotations) live rent free in your mind. Focus on actually building stuff.

I hate to use oppression Olympics as an argumentative avenue but I'm a Slav, the word slavery comes from the enSLAVement of Slavic people.

It's often an accurate description of the relationship and even if it's just slightly inefficient, still works great. Why pollute your mind and thought stream with such trivial qualms. Build. Learn. Don't sweat dumb crap.

Same with whitelist / blacklist. From an English perspective these and master/slave have been archetypes for as long as the language has existed. Don't make problems where there are none, and don't TRY to pollute my mind by making me think there are problems, where again there are none. Enough people outside of tech are doing that to the populace.

replies(2): >>17957702 #>>17957866 #
20. infiniteseeker ◴[] No.17956584[source]
Next will be male/female (in the context of wires, plugs, sockets etc) and it will be on a "spectrum of gender fluidity".

Political correctness/SJWism has gone insane and cannot be taken seriously.

replies(1): >>17956630 #
21. downquarklove ◴[] No.17956630[source]
Those namings relate to male/female sexes, not genders, so thats not applicable here. And even if it was applicable, if the name is steeped in a context that is oppressive to certain groups and there is a viable alternate name, what's the harm in exploring those alternatives and making conversations comfortable for everyone who might be involved?
replies(2): >>17957073 #>>17960217 #
22. downquarklove ◴[] No.17956661[source]
Don't you think your assumption that "the only people who recommend this change are virtue signalling" is pretty dismissive?

And what about a change of commander/obeyer? Sure it might not roll off the tongue at first, but the concept is clearly expressed and if it makes the community more inclusive then I would argue it's a technically better choice.

replies(1): >>17965221 #
23. downquarklove ◴[] No.17956673{4}[source]
If the code is written in English, and the meaning in English is genuinely offensive to some, then yes I believe it should all be changed.
24. jboles ◴[] No.17956802[source]
If they are going to replace words that some people consider loaded, they could simply replace “slave” with “servant”. Then the previous technical meaning of “master”/“slave” (e.g. Parrallel ATA interface) is preserved, but the meaning of “master” as the original copy (e.g. Git, databases) or as the head personnel (e.g. webmaster) does not have to be changed.
25. UncleMeat ◴[] No.17956826{4}[source]
Sure why not? If a term is truly causing harm then we should seriously consider removing it.

Programs are identical after renaming. How exactly would such a change break your project?

replies(1): >>17957976 #
26. subjectsigma ◴[] No.17957073{3}[source]
This issue has popped up at least once before where someone on Github was complaining about the terminology "killing a child process" because they had recently had a death in the family and were trying to cope. Most agreed that changing it would be a tremendous waste of time (think of all that perfectly good documentation and code which must be rewritten), unnecessarily polarizing (I won't join your project because your dictionary words hurt me!), placing the needs of the few squarely over the needs of the many, and teaching people that "social justice" means "everything should be as I want it."

It really depends on what you mean by "a context that is oppresive to certain groups", but I think this is roughly the same. Nobody should have to endure insults, bullying, harrasment, sexual innuendos, etc. in open source projects, especially maintainers and contributors. However it sounds to me like hurting over words such as master/slave, male/female, kill/fork/dongle/etc is an incredibly small group of people actively looking to cause problems. If you absolutely can't accept the terminology "killing a process" then perhaps the problem lies with you, not 40-odd years of convention.

replies(1): >>17957146 #
27. downquarklove ◴[] No.17957146{4}[source]
What about phasing out the usage of offensive terms in the same way we do with non-programming languages? We don't go back and correct all of our books, but when reading old books the offensive terms that were common in previous times stick out like sore thumbs. And I know that I have at least a few phrases that I used to use without a thought that I conciouslly chose to drop from my speech now that I see the wrong in them. I think we can choose to change as a community with little overhead to the "many" but provide a great benefit to the "few".
replies(1): >>17960331 #
28. sitkack ◴[] No.17957702[source]
> whitelist / blacklist

I tried really hard to find a race basis for this but could not.

29. nocigar ◴[] No.17957723{3}[source]
Wait until you hear where the word "server" comes from...
30. type0 ◴[] No.17957866[source]
This topic is very US centric and shows how deep the disregard is for people of other ethnicities.

> The word slovo ("word") and the related slava ("glory, fame") ... [0]

The word is indeed derived from "slav" ethnonym and it's unbelievable that no one actually cares. The whole thing smells not just "word policing" but "history revisionism".

Jewish people were enslaved for so much of their history (by babylonians, egyptians etc) should we simply forget that because some people feel uneasy hearing the word slave and imagining this inhumane practice? Also no one forces you to antropomorfize technical concepts.

It strikes me how fast the discussion went off topic to the "subconscious racism" of black people in the previous threads about this. [1]

As mentioned in other discussion, this is a very slippery slope to go. Next time I won't be able to use the technical term "torture testing" to refer to server workloads because torture sounds aggressive and inhumane. What more combination of words I won't be able to use? Will I be able to write about "cookie stealing", it's illegal to steal so we surely need to ban that word. There are after all people that have their whole life savings stolen from them by the crooked investment managers so the "stealing" concept will remind them about that and might hurt their feelings.

[0] https://en.wikipedia.org/wiki/Slavs_(ethnonym) [1] https://news.ycombinator.com/item?id=17930979

31. type0 ◴[] No.17957976{5}[source]
How can technical terminology cause harm? This is simply "history revisionist" attempt to erase the concept from the public consciousness. The usage of the "slave" word in software is so abstract that removing creates much more harm because it will produce miscommunication and result in faulty programs where there shouldn't .

What word will one replace "slave" with? Should we go to the origins of the word and just use ethnonym "slav" instead? After all it has only positive connotations {The word slovo ("word") and the related slava ("glory, fame"...}. And sure slavic people won't mind right?! /s

32. knaik94 ◴[] No.17959557[source]
I would be fine with Master/Clone or Master/Replica. I feel like master is needed to convey the meaning clearly, but slave isn't. Master as in master vinyl record.
replies(1): >>17963441 #
33. kruczek ◴[] No.17959595[source]
Primary/Secondary is what comes to mind.

But yea, such "political correctness" make people fear and despise words, instead of concepts behind them.

34. sjapkee ◴[] No.17960217{3}[source]
>relate to male/female sexes, not genders

It is the same.

replies(1): >>17985943 #
35. subjectsigma ◴[] No.17960331{5}[source]
That's a reasonable suggestion that's hard to argue with - I still don't see it happening. There's no consensus on what words are bad (what makes "kill" worse than "dongle", etc.) or how these should be phased out. Does that mean at the next major version number? By 2025? Will the documentation be internally inconsistent during those years? If some projects agree and others don't there would be cumbersome fragmentation. And... for what? So a couple dozen people in the entire community can avoid the concepts of murder or sex or slavery, as if they didn't exist?

I guess it all comes down to context, really... It's easy for me to sit here and say "This change is meaningless" but I guess if I had lots of influence over Python and I knew people who were actually suffering because of this, I may approach it differently.

36. legostormtroopr ◴[] No.17963441{3}[source]
A slave process or drive is not a clone of the master though. They might be entirely different code.
37. tooltalk ◴[] No.17965221{3}[source]
why is master/slave not inclusive? Who are we "excluding" with those terms?
38. aphelion ◴[] No.17981081[source]
In the real world, what's going to make Python less accessible to PoC new to programming(and new programmers generally) is the arbitrary change in well-defined industry-wide terminology. Working from documentation and learning resources that employ master/slave when those terms have been memory-holed in the newest versions of the language is the sort of seemingly minor change that trips up neophytes.

I'm not totally opposed to changing master/slave. But in light of the real cost of this sort of breaking change I'd want to see some evidence of actual harm, rather than some guy(who, I feel compelled to point out, is a white man) insisting that we change it for what are, at best, totally theoretical and speculative harms.

39. yellowapple ◴[] No.17984336[source]
If I'm into BDSM, can I claim offense with this ticket and resulting PR?