Re:GitHub to replace “master” with alternative term to avoid...
by athenot on Monday June 15, 2020 @10:09AM
(#23527463)
When I first learned about MySQL replication in 2000, the terminology of Master/Slave seemed really odd and a bit shocking.
20 years later the terminology is still peculiar to me. Even though it's an accurate representation of what slavery is, it is a poignant reminder of a human condition that has sadly not disappeared.
The slave database, as a read-only data store, has no agency and can only follow the master database. Yet it's the workhorse for all heavy queries, especially those part of reporting for various execs. It has no value except as a means to reduce load of the master and to save it's rear end in case of corruption... yes there are many parallels.
But still, after 2 decades, I'm not sure that this terminology is essential to the world of relational databases. "Leader/Follower" or "Primary/Replica" could convey the same data concepts without resorting to a comparison with slavery. Perhaps a wording update might be in order here.
Re:GitHub to replace “master” with alternative term to avoid...
by kerkeslager on Monday June 15, 2020 @10:15AM
(#23527537)
People of color of HN: is this something you care about, or is this just white people virtue signalling? I'm not trying to start a debate here, I'm genuinely asking because I don't know the answer.
Re:GitHub to replace “master” with alternative term to avoid...
by Shihan on Monday June 15, 2020 @10:57AM
(#23527955)
I can tell, it never crossed my mind. Never saw anything related to slavery/race in this and still don't see anything now. In German the word Meister is also used, for example "Meisterberuf", "Lehrmeister", "Bäckermeister", "Weltmeister" - it has nothing to do with slavery, in most of the cases it just means a person who has mastered something. You have this to, e.g. "Master of Arts". So now what, should this also be renamed? "Skillful Person of Arts" maybe? This word bastardization reminds me really of 1984 and I fear it dumbs us all down.
Re:GitHub to replace “master” with alternative term to avoid...
by wreck on Monday June 15, 2020 @10:01AM
(#23527378)
When building my first computer (with help) in the 90s, setting up hard drives in master or slave was the first time I learned the definition of slave. I was 6-7 years old and I'm not American
Re:GitHub to replace “master” with alternative term to avoid...
by EForEndeavour on Monday June 15, 2020 @10:26AM
(#23527664)
Would you agree from personal experience that calling them something like "primary" and "secondary" or "1" and "2" would be less confusing to 6-7 year olds?
Re:GitHub to replace “master” with alternative term to avoid...
by glandium on Tuesday June 16, 2020 @08:19AM
(#23538243)
Except there were 2 IDE buses, a primary and a secondary, each possibly receiving a master and a slave. So at the very least primary/secondary wouldn't have worked.
Re:GitHub to replace “master” with alternative term to avoid...
by _vbdg on Monday June 15, 2020 @09:54AM
(#23527316)
Really? Git doesn't have any actual references to slavery. Even if it did, is everyone now disallowed from using such terminology in a benign manner? I'm not sure calling it "main" makes sense, in any case; anyone who uses git flow knows it's not the "main" branch where most of the day-to-day development takes place. As the article points out, this is a reference to "master" as in a master copy, which is why there's no corresponding reference to slave. If you download the official git book from the website and control-F for "slave", do you know what you'll find? Zero results!
Re:GitHub to replace “master” with alternative term to avoid...
by mmm_grayons on Monday June 15, 2020 @10:14AM
(#23527530)
I suppose that makes more sense than main, but it still begs the question: given that there is no actual slave reference to remove, and discounting the questionable necessity of removing such a reference were one to exist, what's the problem here?
Re:GitHub to replace “master” with alternative term to avoid...
by musingsole on Monday June 15, 2020 @10:46AM
(#23527859)
Do you work with people who are experts in things you are not? When one of them raises a concern from the field they're an expert in, you generally must take it as a given that the issue is real and must be addressed, as the expert raised it and wouldn't have otherwise. Your role might be to point out how the presented fixes to the issue might/will cause issues in your field of expertise, but it'd be pretty unprofessional to question the validity of the other expert's designated problem.
Re:GitHub to replace “master” with alternative term to avoid...
by mmm_grayons on Monday June 15, 2020 @06:59PM
(#23534061)
Sorry, who exactly is the expert here that shouldn't be questioned? While I generally respect experts' opinions, I also find it reasonable to question them when they don't pass the smell test.
Re:GitHub to replace “master” with alternative term to avoid...
by musingsole on Tuesday June 16, 2020 @01:41AM
(#23536248)
I can respect your smell test if you can respect that it's highly susceptible to Dunning-Kruger and the various mishaps of not knowing what you don't know.
Re:GitHub to replace “master” with alternative term to avoid...
by _vbdg on Tuesday June 16, 2020 @09:02AM
(#23538638)
You still didn't tell me who these "experts" are, though it seemed you were referencing several in a prior comment, so I'm unsure how to address that point. I'd of course respect the opinion of an expert, and I'd recognize that I have a very good chance of being wrong, but that makes it even more important to question things. If I'm right, we come to a better conclusion; if I'm wrong, I learn something.
Re:GitHub to replace “master” with alternative term to avoid...
by _vbdg on Tuesday June 16, 2020 @11:24AM
(#23540226)
HN doesn't permit users to downvote replies to their comments, so you can be sure that wasn't me. Maybe it's obvious to you, but I'm not sure how one qualifies as an expert in the necessity (or not) of removing the word "master" from github branches. Would you please elaborate?
Re:GitHub to replace “master” with alternative term to avoid...
by IshKebab on Monday June 15, 2020 @11:18AM
(#23528178)
That would be very confusing considering git repos are DAGs, and `master` is normally exactly the opposite of the root node. Look up `git rebase --root`.
Re:GitHub to replace “master” with alternative term to avoid...
by gnusty_gnurc on Monday June 15, 2020 @11:51AM
(#23528714)
But in all seriousness, if seeing it on github is damaging to the mental health of descendants of slavery in america, why do we allow the word to exist at all and where is it acceptable to use? Should words like this be put in something akin to a biohazard lab?
Re:GitHub to replace “master” with alternative term to avoid...
by BigJ1211 on Monday June 15, 2020 @10:16AM
(#23527564)
The team I work in is rather diverse, (not just black/white). This never crosses anyone's mind here.
It also reeks far more of virtue signalling than it does of actually taking action against racism or racial bias. This screams loudly "Look at all we're doing to stop 'racism'" without actually addressing the real harmful racist issues that exist. (Slavery wasn't just against black people, though it's the most recent or well known form of slavery. It sadly also still exists)
Like people of minorities being paid less or people of minorities havin to go to significantly more interviews. This isn't limited to black people, this is for anyone who has a non-white sounding name. (at least in Western civilisation).
I don't really care if they do change it personally, but I'd much prefer actual action against racial bias.
Previous discussion.
This one won't be any different.
https://news.ycombinator.com/item?id=23500093 (138 points/224 comments)
https://news.ycombinator.com/item?id=23519813 (44 points/84 comments)
https://news.ycombinator.com/item?id=23522859 (15 points/5 comments)
20 years later the terminology is still peculiar to me. Even though it's an accurate representation of what slavery is, it is a poignant reminder of a human condition that has sadly not disappeared.
The slave database, as a read-only data store, has no agency and can only follow the master database. Yet it's the workhorse for all heavy queries, especially those part of reporting for various execs. It has no value except as a means to reduce load of the master and to save it's rear end in case of corruption... yes there are many parallels.
But still, after 2 decades, I'm not sure that this terminology is essential to the world of relational databases. "Leader/Follower" or "Primary/Replica" could convey the same data concepts without resorting to a comparison with slavery. Perhaps a wording update might be in order here.
It also reeks far more of virtue signalling than it does of actually taking action against racism or racial bias. This screams loudly "Look at all we're doing to stop 'racism'" without actually addressing the real harmful racist issues that exist. (Slavery wasn't just against black people, though it's the most recent or well known form of slavery. It sadly also still exists)
Like people of minorities being paid less or people of minorities havin to go to significantly more interviews. This isn't limited to black people, this is for anyone who has a non-white sounding name. (at least in Western civilisation).
I don't really care if they do change it personally, but I'd much prefer actual action against racial bias.