Remove blacklist terminology · Issue #6475 · osquery/osquery • Hacker News Board

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#11

Post by noble_pleb » Mon, Jun 08, 2020, 1:34 PM UTC

CM30 wrote: post #7

Hmm, wonder what they'd think about some other tech terminology then? https://en.wikipedia.org/wiki/Master/slave_(technology) Or talks of this in typesetting? https://en.wikipedia.org/wiki/Widows_and_orphans But seriously, this is silly. Words and phrases can have multiple meanings, and they can be used for perfectly innocuous reasons in various fields.

Exactly, this is just going to open up a pandora's box. What about black box testing, is that racist too? And english terms like blacksmith and black-belt?

There are terms like "blue-collar vs white-collar" too which (according to some) deems the physical labor class as inferior to the "white" intellectual class. And there are general english phrases too like "things aren't always black and white, there is a lot of grey area in between".

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#12

Post by BlackVanilla » Mon, Jun 08, 2020, 1:35 PM UTC

Allowlist/denylist is clearer and more intuitive. This same pattern (bias free communication resulting in clear communication and less injustice) can be seen in other areas. I wrote this yesterday after Microsoft added a bias free communication document, but it's as applicable here. [1]

At least in the UK, its laws and people would be better served if bias-free communication is rigorously used when lawmaking. The Land Registration Act 2002 an example of statute law which unnecessarily and frequently uses gendered words. [2] Sure, this is written in 2002, times have changed, but this is considered a key statute governing UK land law. Even as recently as 2015, unnecessarily gendered language is used in the Consumer Rights Act 2015, albeit not as consistently, but I'm sure more recent statutes also share this trait. [3] This sloppiness in writing could cause needless issues.

Here's one. The UK's Foreign Office noted that the phrase 'pregnant women' in general comment No. 36 on article 6 of the UN's International Covenant on Civil and Political Rights on the right to life 'may be inadvertently restricting the application of this paragraph to exclude transgender people who have given birth; this has happened in two recent cases in the UK.' [4]

This issue is greater than everyday language in business, but the same type of language used in our laws can potentially cost businesses by poorly-worded contracts or by other means. Legislatures should use gender-neutral language to make their laws clear and reduce potential needless unjustice.

[1] https://news.ycombinator.com/item?id=23439283 / https://docs.microsoft.com/en-us/style-guide/bias-free-commu...

[2] https://www.legislation.gov.uk/ukpga/2002/9/section/24

[3] https://www.legislation.gov.uk/ukpga/2015/15/pdfs/ukpga_2015...

[4] https://www.ohchr.org/Documents/HRBodies/CCPR/GCArticle6/Uni...

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#13

Post by cvburgess » Mon, Jun 08, 2020, 1:35 PM UTC

CM30 wrote: post #7

Hmm, wonder what they'd think about some other tech terminology then? https://en.wikipedia.org/wiki/Master/slave_(technology) Or talks of this in typesetting? https://en.wikipedia.org/wiki/Widows_and_orphans But seriously, this is silly. Words and phrases can have multiple meanings, and they can be used for perfectly innocuous reasons in various fields.

I wouldnt agree there. Words are powerful and just becuase they have multiple meanings does not mean that those meanings are not intertwined.

Regardless, those terms don't translate well and only make sense to people from other languages and cultures if you translate it and add the history to it.

Primary and Replica accomplishes the same goal but don't get weighed down with "what is a slave" or "why is the master in control of the slave".

Yeah, its a little inconvenient, but making software more inclusive is a pretty worthy goal and it wont take long to update the majority of tools and texts.

Staying comfortable because its convenient, while it makes other people uncomfortable, is probably why software is still largely a white male industry.

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#14

Post by jgwil2 » Mon, Jun 08, 2020, 1:49 PM UTC

noble_pleb wrote: post #11

CM30 wrote: post #7

Hmm, wonder what they'd think about some other tech terminology then? https://en.wikipedia.org/wiki/Master/slave_(technology) Or talks of this in typesetting? https://en.wikipedia.org/wiki/Widows_and_orphans But seriously, this is silly. Words and phrases can have multiple meanings, and they can be used for perfectly innocuous reasons in various fields.

Exactly, this is just going to open up a pandora's box. What about black box testing, is that racist too? And english terms like blacksmith and black-belt? There are terms like "blue-collar vs white-collar" too which (according to some) deems the physical labor class as inferior to the "white" intellectual class. And there are general english phrases too like "things aren't always black and white, there is a lot of g…

Those terms don't explicitly attach a moral value to colors in the way that the blacklist/whitelist terminology does.

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#15

Post by haunter » Mon, Jun 08, 2020, 1:55 PM UTC

BlackVanilla wrote: post #12

Allowlist/denylist is clearer and more intuitive. This same pattern (bias free communication resulting in clear communication and less injustice) can be seen in other areas. I wrote this yesterday after Microsoft added a bias free communication document, but it's as applicable here. [1] At least in the UK, its laws and people would be better served if bias-free communication is rigorously used when lawmaking. The Lan…

>Legislatures should use gender-neutral language to make their laws clear and reduce potential needless unjustice

In my mother tongue there are no gender-specific pronouns, there is just no grammatical gender. It doesn't exist so this whole thing so foregin to me.

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#16

Post by polotics » Mon, Jun 08, 2020, 2:08 PM UTC

jgwil2 wrote: post #10

atonse wrote: post #6

I was genuinely curious about the origin of this word based on these various projects changing it. And according to the Wikipedia (and the Oxford English Dictionary) [1], it seems like blacklist came about something like 400 years ago in England. It doesn't seem to be related to race. Is there a movement that is tying those terms to race? Would love to see some links about it. Because master/slave seems like a much m…

Irrespective of origin, the terms seem to equate black to bad and white to good, and as such should probably be retired. As other comments have mentioned, allow and deny are more explicit and thus probably better terms anyway.

A recommended read/watch where white is definitely bad: "a dry white season" by André Brink. Oscar for Marlon Brando in the movie version.

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#17

Post by ThrowawayR2 » Mon, Jun 08, 2020, 3:44 PM UTC

jgwil2 wrote: post #14

noble_pleb wrote: post #11

Earlier quoted context omitted.

Exactly, this is just going to open up a pandora's box. What about black box testing, is that racist too? And english terms like blacksmith and black-belt? There are terms like "blue-collar vs white-collar" too which (according to some) deems the physical labor class as inferior to the "white" intellectual class. And there are general english phrases too like "things aren't always black and white, there is a lot of g…

Those terms don't explicitly attach a moral value to colors in the way that the blacklist/whitelist terminology does.

Then how about "black hat hacker"?

Wonder what the reaction would be if people requested that "brownout" no longer be used because it's insensitive to certain races.

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#18

Post by atonse » Mon, Jun 08, 2020, 9:31 PM UTC

jgwil2 wrote: post #10

atonse wrote: post #6

I was genuinely curious about the origin of this word based on these various projects changing it. And according to the Wikipedia (and the Oxford English Dictionary) [1], it seems like blacklist came about something like 400 years ago in England. It doesn't seem to be related to race. Is there a movement that is tying those terms to race? Would love to see some links about it. Because master/slave seems like a much m…

Irrespective of origin, the terms seem to equate black to bad and white to good, and as such should probably be retired. As other comments have mentioned, allow and deny are more explicit and thus probably better terms anyway.

I agree with you – allow/deny is easier to understand anyway for global non-native speaker (which all open source projects are).

There are many practical reasons to do this, I just wasn't clear on the race-related reasoning for blacklist/whitelist, as these words don't have those etymologies.

The color black is associated with many non-racial things that we do deem negative, things like death with black clothes at funerals, things like the darkness being the absence of light (sunlight is brighter/whiter), being blinded, where we do mostly see black. None of those things are racial, what should we do about all those?

Perhaps a better change is to stop associating our skin color with those words which are clearly inaccurate, and serve absolutely no useful purpose.

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#19

Post by RhysU » Tue, Jun 09, 2020, 12:53 AM UTC

jgwil2 wrote: post #10

atonse wrote: post #6

I was genuinely curious about the origin of this word based on these various projects changing it. And according to the Wikipedia (and the Oxford English Dictionary) [1], it seems like blacklist came about something like 400 years ago in England. It doesn't seem to be related to race. Is there a movement that is tying those terms to race? Would love to see some links about it. Because master/slave seems like a much m…

Irrespective of origin, the terms seem to equate black to bad and white to good, and as such should probably be retired. As other comments have mentioned, allow and deny are more explicit and thus probably better terms anyway.

What if we're whitelisting or blacklisting people for the privilege to attend an organized crime conference? Which is good? Which is bad?

Note that in this unsavory example there is zero ambiguity! The whitelisted will hang with other mobsters, clearly, plotting evil and mayhem all the while.

The terms have bupkis to do with good and bad.

Re: Remove blacklist terminology · Issue #6475 · osquery/osquery

#20

Post by jgwil2 » Tue, Jun 09, 2020, 1:23 AM UTC

atonse wrote: post #18

jgwil2 wrote: post #10

Earlier quoted context omitted.

Irrespective of origin, the terms seem to equate black to bad and white to good, and as such should probably be retired. As other comments have mentioned, allow and deny are more explicit and thus probably better terms anyway.

I agree with you – allow/deny is easier to understand anyway for global non-native speaker (which all open source projects are). There are many practical reasons to do this, I just wasn't clear on the race-related reasoning for blacklist/whitelist, as these words don't have those etymologies. The color black is associated with many non-racial things that we do deem negative, things like death with black clothes at fu…

I agree that'd be great, but between changing the world's use of black and white and changing our use of blacklist and whitelist as pertains to software, one is in our power and the other isn't. Who knows, maybe someone will even feel more welcomed in our profession as a result.