Splunk to junk masters and slaves once a committee sanity-checks replacements
(2020/06/16)
- Reference: 1592279166
- News link: https://www.theregister.co.uk/2020/06/16/splunk_changes_language/
- Source link:
Splunk has joined the ranks of software makers pledging to remove objectionable terms like “master” and “slave” from their code.
As [1]detailed in a post today by senior veep for cloud and chief product officer Sendur Sellakumar, a registration-walled [2]post titled “Change the terms ‘Master’ and ‘Slave’ in indexer clustering and licensing” popped up two weeks ago in the Splunk forums.
Just because these terms have been part of our industry vernacular for so long doesn’t make them any less offensive
“This post received upvotes faster than any I can remember, for good reason,” Sellakumar said, adding: “We’re addressing this issue as fast as we can, and will replace the terms with ones that are neutral, inoffensive and, in fact, more descriptive of the meaning.”
“I’m a computer science major, and I can tell you that software engineers and technologists learn these terms in the first year of college,” Sellakumar continued. “They are ingrained in the way we talk about technology, such as in the domains of distributed systems and network security. We can — and should — find better terms that remove unconscious bias from our work. Just because these terms have been part of our industry vernacular for so long doesn’t make them any less offensive.”
Developers renew push to get rid of objectionable code terms to make 'the world a tiny bit more welcoming' [3]READ MORE
Splunk has already started the process of deciding on terms to replace “whitelist” and “blacklist,” and also created “a working group of people from across the organization to develop additional recommendations, guidelines and procedures to identify and replace biased language and to prevent other instances from happening in the future.”
Similar discussions are going on around the industry.
Google has created an " [4]inclusive documentation guide " that suggests replacing words like "crazy outliers" and "crippling" a service with "slows down" and "baffling outliers", because the terms represent "problematic ableist language." Similarly, "sanity check" is objectionable when "a final check for completeness and clarity" is more inclusive.
The web ad giant also wants to replace acronyms such as the cluster-management technique STONITH – Shoot The Other Node In The Head – with less-violent terminology.
An active debate in the OpenSSL community is [5]considering how to do away with “master,” and even whether "whitespace” could be replaced with “blanks.” The Go programming language community is [6]discussing similar issues , and the conversation has pondered whether replacing “blacklist” with “blocklist” might be offensive to those who reside in the former Eastern Bloc of nations.
A final footnote: the forum post that set Splunk thinking about addressing its loaded language may have prompted an unusually swift reaction, but the forums’ most-voted-on recent post is one that calls for the company to develop a Universal Forwarder for 64-bit Armv8. That post has generated 928 votes, compared to 696 for the one on replacing master and slave. ®
Sponsored: [7]Continuous Lifecycle Online, 15 July 2020. Register Now!
[1] https://www.splunk.com/en_us/blog/leadership/biased-language-has-no-place-in-tech.html
[2] https://ideas.splunk.com/ideas/EID-I-325
[3] https://www.theregister.com/2020/06/08/developers_renew_push_to_get/
[4] https://developers.google.com/style/inclusive-documentation
[5] https://github.com/openssl/openssl/pull/12089
[6] https://go-review.googlesource.com/c/go/+/236857/
[7] https://go.theregister.com/tl/1968/shttps://www.bigmarker.com/series/continuous-lifecycle-online/series_summit
As [1]detailed in a post today by senior veep for cloud and chief product officer Sendur Sellakumar, a registration-walled [2]post titled “Change the terms ‘Master’ and ‘Slave’ in indexer clustering and licensing” popped up two weeks ago in the Splunk forums.
Just because these terms have been part of our industry vernacular for so long doesn’t make them any less offensive
“This post received upvotes faster than any I can remember, for good reason,” Sellakumar said, adding: “We’re addressing this issue as fast as we can, and will replace the terms with ones that are neutral, inoffensive and, in fact, more descriptive of the meaning.”
“I’m a computer science major, and I can tell you that software engineers and technologists learn these terms in the first year of college,” Sellakumar continued. “They are ingrained in the way we talk about technology, such as in the domains of distributed systems and network security. We can — and should — find better terms that remove unconscious bias from our work. Just because these terms have been part of our industry vernacular for so long doesn’t make them any less offensive.”
Developers renew push to get rid of objectionable code terms to make 'the world a tiny bit more welcoming' [3]READ MORE
Splunk has already started the process of deciding on terms to replace “whitelist” and “blacklist,” and also created “a working group of people from across the organization to develop additional recommendations, guidelines and procedures to identify and replace biased language and to prevent other instances from happening in the future.”
Similar discussions are going on around the industry.
Google has created an " [4]inclusive documentation guide " that suggests replacing words like "crazy outliers" and "crippling" a service with "slows down" and "baffling outliers", because the terms represent "problematic ableist language." Similarly, "sanity check" is objectionable when "a final check for completeness and clarity" is more inclusive.
The web ad giant also wants to replace acronyms such as the cluster-management technique STONITH – Shoot The Other Node In The Head – with less-violent terminology.
An active debate in the OpenSSL community is [5]considering how to do away with “master,” and even whether "whitespace” could be replaced with “blanks.” The Go programming language community is [6]discussing similar issues , and the conversation has pondered whether replacing “blacklist” with “blocklist” might be offensive to those who reside in the former Eastern Bloc of nations.
A final footnote: the forum post that set Splunk thinking about addressing its loaded language may have prompted an unusually swift reaction, but the forums’ most-voted-on recent post is one that calls for the company to develop a Universal Forwarder for 64-bit Armv8. That post has generated 928 votes, compared to 696 for the one on replacing master and slave. ®
Sponsored: [7]Continuous Lifecycle Online, 15 July 2020. Register Now!
[1] https://www.splunk.com/en_us/blog/leadership/biased-language-has-no-place-in-tech.html
[2] https://ideas.splunk.com/ideas/EID-I-325
[3] https://www.theregister.com/2020/06/08/developers_renew_push_to_get/
[4] https://developers.google.com/style/inclusive-documentation
[5] https://github.com/openssl/openssl/pull/12089
[6] https://go-review.googlesource.com/c/go/+/236857/
[7] https://go.theregister.com/tl/1968/shttps://www.bigmarker.com/series/continuous-lifecycle-online/series_summit
Where will this end....
BigE
What next, is someone going to be holding something your you know what when you to the toiler? oh I am sorry, you don't have one of those, so that makes me a misogynist. You must feel excluded.
Ok....but whitespace?
I am David Jones
The word has no positive or negative connotations (that I know of), it is purely descriptive.
Erasing the word “white” from our vocabulary would seem to be a step too far...
When STONITH falls
Replace all instances of [something] that may offend [somebody]. Let us also do away with humor and sarcasm while we're at it. One reader's merry world is another reader's nightmare. That has always been the case. When will the madness stop? Do we really need to define all possible variations of "unwords" and replace them with new expressions that will be considered "unwords" by another generation? When STONITH falls, which is humorously and actually quite appropriately called STONITH, then nothing is safe. Can you "kill" a computer or are metaphors no longer an acceptable language construct to convey a specific meaning?
Really? Remove "crazy" and "baffling" outliers? But, the checks are there to handle crazy and baffling outliers. Remove "sanity check"? Any programmer knows that the user is not to be trusted to input anything valid. The user is, by definition, not sane, from a programmer's perspective. A user is (probably) human and therefore a non-logical entity. The computer is nothing but logical. The programmer is the intermediate between the logical and the illogical. That perspective is both crazy and baffling.
Fast-forward a few generations... the "blocklist" is not an acceptable term. It is discriminatory and therefore unacceptable. Terminology like "primary" and "secondary" is no longer acceptable because it denotes unacceptable preference and denies the egalitarian view. Any "word" will, at some stage, have multiple meanings in which some may be offended. Maybe we should do an introspective and think about what that says about us?