RFC 9454: Update to OSPF Terminology, Changed Master/Slave to Leader/Follower | Hacker News Reader

Follower

rfc-editor.org

18 pointsSea-n70 comments

Thinking that changing words changes the material reality for anyone is peak Californian ideology.

I'm sure all the slaves being sold in markets in Libya are ecstatic about this change.

Anytime anyone cites "Californication" it's just a clear indicator that they don't live here, and know nothing about the place.

California is NOT what people think in Texas and Arkansas.

Isn't it the US south that's banning certain words and discussions?

That said, as a person living in California, I totally agree this change in nomenclature is pointless posturing.

In my opinion, being a "follower" is even more disparaging than being a "slave"

Slaves don't really have a choice, they're forced to labor, whereas followers are volunteering their subservience.

Sad thing is it's not pointless. It's very much got a point like all the other apparently pointless word fights these people engage in. It's a way to separate and identify ideological fellow travelers (one might say followers) from the rest, so when they do something more consequential they know who their allies will be. A lot of so called virtue signalling is like this. People often label something that without asking, signalling to whom? Same thing with pronouns and other highly visible loyalty pledges to the ideology. Nobody does them because they literally think they're useful, they're doing it whether they realized it or not to signal that they're one of the tribe and can relied on when a more important fight comes up.

Of course this is a sword that cuts in both directions. Smart leaders who realize this now know who the troublemakers are and can find ways to get rid of them. Unfortunately most don't take it seriously enough to do so(yet).

As somebody who lived in California, I also was startled how stupid some TV shows I enjoyed while living in LA became when I moved to another state. It's not some California magic though, but the lack of a different perspective. When I moved back to California I had no problem observing the general Californian madness with a little effort of telling myself "what would a normal person think about that?".

And our infrastructure becomes a little bit less BDSM.

Petition to change terminology to top and bottom.

I could work with strange and charmed.

This is what happens when radical political ideology clashes with technical rationality. If only Orwell were alive today...

Why is it clashing? There’s nothing more “rational” about one set of terms than the other.

The whole thing is a bit silly IMHO, but … meh?

If Orwell were alive he’d be worried about far more important things than what terms some of the worlds wealthiest professionals are using to describe their tools

...then as a lifelong leftist and champion of precision and clarity in writing, he would perhaps have applauded the choice of language which is both inclusive and technically exact?

I don’t really understand the divisiveness of changes like this at all.

I already say “allowlist” and “denylist” habitually. Whether or not that matches my sociopolitical ideology, that’s just where the industry is at this point, so I’m there too.

Everything in technology is constantly changing, and that is neither good nor bad in my view. It seems strange to be outraged by a change in vernacular when that’s amongst the easiest changes we face in industry generally. I say, save the anger for when SAP stops investing in on-prem, when Azure signing keys are compromised, when your storage vendor loses your data, or when your upstream vendor changes licenses on you, etc.

If this terminology was more explicitly sexually tinted instead, the American right would have no problem with changing it “to protect the children.”

But because it’s associated with a legacy of systemic violence, suddenly changing it is an offense to civilization that would have Orwell turning in his grave.

You see this same bizarre logic in the new education laws in Florida. It’s bad to tell children about human sexuality. It’s fine to tell them about human violence. But it’s very bad to tell them about society’s role in violence.

> If this terminology was more explicitly sexually tinted instead, the American right would have no problem with changing it “to protect the children.”

If that was true wouldn't they have objected to male, female, and hermaphrodite connectors?

If some tech company introduced a product today that used words like “hermaphrodite” and “trans” to describe anything, it would soon be featured on Fox News as evidence of subversive leftist corruption.

Still sounds much better than back history changes.

You really think if entrenched terminology had potential sexual connotations, people wouldn’t make a fuss for it being changed?

You must not have met humans.

If the original terms had been “cock/pussy”, I don’t think many people would be up in arms about changing it.

And if you think that’s ridiculous because nobody would ever have named database components something so obviously offensive, that’s the whole point. Norms change. There was a time just 50-60 years ago when sexual harassment in workplaces was much more accepted than today. Similarly non-inclusive language was much more accepted a few decades ago. In both cases many men have trouble adapting and think it’s a sign of decline of society that their assumptions are being challenged.

So we're not allowed to label connectors as "male/female" anymore, because that's transphobic.

It seems like peak science denialism to me, and it's ironic that we were already, 30 years ago, using terms like "gender-changer" when discussing these connectors too.

Frankly it seems to me that most modern connectors are trans already: is USB-C male or female? Both?

> is USB-C male or female? Both?

I’m not sure what you mean. Like all of its predecessors, USB-C has both a male and female connector. Your phone has the female port, your charger has the male one.

No, USB-C has one type of plug and one type of port. The port is what you're calling "female" except look closer. Look closer at both ends. You'll see that the plug has a female void in the center, and the port has a male tongue in the center to fit that void. Therefore, it is hermaphrodite on both ends.

It's basically the same with USB-A.

Hermaphroditic connectors are a thing, and it’s not what usb c is.

Wikipedia has a good article. https://en.m.wikipedia.org/wiki/Gender_of_connectors_and_fas...

Huh, thanks for the article: I've learned a bunch of new terms that we still can't use anymore because they refer to gender and they're transphobic. The guidance said to use terms like "plug" and "port" and "jack" and stuff. Although isn't the latter rather sexual???

Sexual doesn't have harmful or derogatory overtones, it's neutral.

Yeah, well, you know, that's just like, uh, your opinion, man. https://youtu.be/Z-xI1384Ry4

The guidance says that referring to "gender" is offensive, and so the guidance lists preferred terms that replace the "gender-based" offensive ones. That's the opinion of the folks who wrote the guidance.

What exactly is the technical reality that ties us to the terms "master" and "slave", or means that we can't change them if enough people think that's good?

You mean other than the decades of connotation lost by switching terminology and the innevitable miscommunication that will ensue?

There's nothing wrong with switching the term "red" for the term "carmine", "scarlet" or "ruby" either. They are synonymous on paper, but the latter simply doesn't evoke the same recall in people's heads.

You are welcome to run the experiment yourself. Go and ask people if they normally stop at (carmine|scarlet|ruby) traffic lights and observe the ensuing confusion at your question.

Then there is the emergent discrepancy in terminology between legacy systems and current RFCs.

I don't find this at all convincing. The terms in question here don't need to evoke connotation, they are labels in a technical system. It is very straight-forward to re-label things. Documentation can easily cover the change.

And I don't find it convincing that you don't find it convincing.

All terms evoke connotation. That's how and why language works. Which is why I proposed the social experiment for you.

Go ahead and replace the word "red" with the word "scarlet" in your day-to-day vocabulary and see if communication with other humans breaks down.

Update the documentation as necessary.

You're generalising far too much. Of course replacing "red" with "scarlet" generally would make a difference in understanding, not least because "scarlet" is more specific than "red". But that doesn't at all speak to what is going on in this RFC, which is already technical documentation and very specific. If you want to talk about connotation, the entire point here is to get rid of terms that connote horrific crimes. Maybe not to you, but clearly to enough people that this change has been made.

Can you explain what you mean?

Extremists have taken accepted standard terminology completely out of context and interpreted it in the worst possible way, using it as a weapon to turn STEM into an environment where (often fabricated!) emotions triumph over rational thought.

One might say that this "inclusiveness" has also lead to the inclusion of idiots.

It really hurts me as an educator, because of course we need to adopt the updated wording because we are sensitive to the changing times and morals, but we also need to make sure that students are aware of the old terms so that they are not unduly confused when they get into the real world, because out there, you're going to find a lot of legacy code and legacy documentation, and most importantly, legacy people like me. At fifty years of age and never much for PC terminology, I am having difficulty adjusting my vocabulary to the "new normal" particularly some things which had come so naturally, such as pronouning people I don't know. I am just thankful that I don't work over real-time video, so I can adjust and correct my writing if I slip up.

Master and slave are just words, sequences of letters, the meaning is in peoples heads and may be context dependent. Never in my life have I thought of slavery when I had to deal with those words in a technical context. I mean, I get the intention, if we would coin those terms today, then we would surely pick terms without this historical burden. But now that those terms have been widely used for a long time, trying to change language in countless standards, books, code bases, peoples heads, ... that just does not seem quite reasonable.

Here's how I think of it: would it be acceptable today, in 2023, to write a book or produce a film that depicts slavery, and has characters who are slave masters and slaves? I would hope it's acceptable, stipulating a sensitive depiction, because masters and slaves are part of reality, they are part of history (and they are still current and happening in our world today).

So if fictional masters and slaves are okay, then what are routers running OSPF? They're not human, and I hope nobody wants to anthropomorphize them so much that they could feel pain, or know injustice. So a router's master/slave relationship is strictly a technical fiction. Is it the best description of the relationship? Is it better than "leader/follower"? Perhaps; I know several other systems that qualify.

Does this trigger African-Americans today who never were slaves? Perhaps.

I'm just not sure of the wisdom to paper over history, to paper over reality, by saying these words are taboo because of what they represent, indeed, in a certain human context of usage.

I could argue that if we're prohibited from depicting masters and slaves in books, films, and OSPF, then people will more easily forget what it means to be a slave, and thereby be condemned to repeat the past which they do not remember. Perhaps it would be more useful to retain "master/slave" relationships among computers, so that we can demonstrate how that works to our children, and show them why it is detestable to human beings.

> But now that those terms have been widely used for a long time, trying to change language in countless standards, books, code bases, peoples heads, ... that just does not seem quite reasonable

You know that there are other terms which have been widely used for a long time that we have successfully changed (most) people's minds about using? It is entirely reasonable to do.

Not only that, but there are a number of terms which are "updated" on a regular basis because they inevitably take on a derogatory connotation and must be euphemized repeatedly. I feel that this is often unreasonable political correctness.

In fact, it's a non-Inclusive practice. One effect of updating the language is that you exclude people of a certain generation. So you can identify and out and then segregate people based on the terminology they use. Are you familiar with "shibboleths"? Happens all the time.

What the IETF is doing here is creating shibboleths for the next generation of IT professionals to separate themselves from the previous generation, who may not be careful enough to update our language.

> One effect of updating the language is that you exclude people of a certain generation.

Your argument is that making language more inclusive is, in fact, exclusionary because people of certain generations are, presumably, incapable of updating their language and thus will continue to go around saying the old, exclusionary things and thus being shunned by the younger generations?

Do you not want to think about that for a bit and consider how utterly ludicrous in sounds in the face of, y'know, reality?

Excerpts from Rules for Radicals by Saul Alinsky:

3. "Whenever possible go outside the expertise of the enemy"

4. "Make the enemy live up to its own book of rules."

5. "Ridicule is man's most potent weapon. There is no defense. It is almost impossible to counterattack ridicule. Also it infuriates the opposition, who then react to your advantage."

8. "Keep the pressure on."

11. "If you push a negative hard and deep enough it will break through into its counterside; this is based on the principle that every positive has its negative."

13. "Pick the target, freeze it, personalize it, and polarize it."

If we have to change people anyway, why not change them to not associate those terms with slavery in a technical context - which I think most people already do anyway - instead of making them use different words? Same result and we do not have to update any existing technical texts.

Orwell the famously committed democratic socialist who sought to promote a more egalitarian and fairer society?

That Orwell?

I imagine he'd be writing an essay asking why this hadn't happened sooner and whether a better pairing of words than Leader/Follower could have been found.

""Guidance for NIST Staff on Using Inclusive Language in Documentary Standards" by the US National Institute of Standards and Technology (NIST) for its inclusive language guidelines"

Finest american garbage ideology

Seems like a reasonable change.

It's corporate window dressing. It doesn't address any real issue. At best it does nothing, at worst gives a corporation something to point to and say "look! we're doing something!"

I will, of course, go along with it precisely because of its inconsequentiality. It's not a hill anyone should be dying on.

It's actually a great way for corporations to enforce ageism and political correctness.

Just bring the person in for an interview, ask him/her to describe OSPF router relationships, and as soon as he/she says "master/slave" you just tick the box that reads "not a good culture fit" and show him/her the door. You didn't even need to inquire about age or political beliefs.

It does seem like the division of a single group into an in-group and out-group, with a shibboleth to distinguish those groups, is the most tangible effect of these kinds of wording changes.

It's interesting, because the same thing is done in the Catholic Church with the liturgy.

The liturgy was very stable for about 500 years, but that's not to say that there weren't revisions. Typically there were not a lot of opportunities for the people themselves to really respond. Of course it was all in Latin, and could be deployed anywhere in the world intact without change.

Now when the liturgy was translated into the vernacular, it presented the opportunity for a vast machinery of translations. There are hundreds of Catholic Masses in all languages, all translated from Latin, and kept updated on a regular basis. This enables musicians to freshly copyright new tunes at every update and rake in cash at parish expense!

So the effect of a regularly updated, vernacular Mass is that sometimes they change the parts that the assembly says. The most drastic and noticeable changes were implemented in 2011, and there have been minor "corrections" since then.

As a result, if you go to a wedding or funeral, you can just listen critically and immediately pick out anyone who hasn't been to Mass in N years because they are automatically reciting the old words by rote!

Likewise, you can catch a priest who's rusty and not in the habit of using the Missal, because they've made really subtle but critical changes to his words in a few places. (Priests are always supposed to read out of the Missal verbatim.)

I don't think there's a subculture that doesn't use jargon/shibboleths as a gatekeeper, and that jargon has to have turnover, otherwise it ceases to be useful (when was the last time you heard "helicopter view"?) Viewed from an anthropological perspective, keeping up with the jargon is just an expected cost of belonging to a subculture, and as an industry we're more guilty than most (look at the sheer amount of new language we coined for Software Patterns, or Agile).

I've never really understood why the internet at large uses BGP instead of more advanced protocols like OSPF.

OSPF is an interior routing protocol that transmits link state. Link state based routing algorithms aren't a good fit for interconnected networks

Every now and then a major ISP accidentally redistributes a million border routes into their IGP with catastrophic results, as a reminder for the rest of us that OSPF et al aren’t designed to handle that many LSAs.

Apples and Oranges: OSPF is an interior gateway protocol for routing within an autonomous system, and it's not designed for exterior routing, which is how eBGP holds the fabric of the Internet together.

Leader in German is "Führer". Isn't that offensive too?

Just to be clear, I'm not even against ditching the master/slave terminology, I'm a privileged white cis man and I think we have to listen to all the oppressed groups.

It's offensive to Anglophones because we only know it in one context!

I would hope that Germans are capable of nuanced understanding of one of their words and not allow it to be hijacked by one rather bad apple.

It'd be like if we couldn't call anyone "President" anymore because Nixon did such a bad job at it.

You don't have to translate literally. For example common polish names for master/slave replication translated back literally would be main/secondary. Each language users group gets to find their own consensus.

Fine by me

Leader/follower is only inclusive language if you’ve never heard of anarchism.

Words get more soft as our industry gets more despotic.

Interesting..

seems like the industry's opinion is it's OK to exploit the common as long as you do it politely and you have marginalised people in your crew.

Reply on news.ycombinator.com