News for hackers. Stuff that matters.
Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by tryptophan on Friday April 22, 2022 @01:12PM (#31124657)
Its worse because suddenly 20 years worth of tutorials are slightly 'wrong' and will confuse new users even more for no reason.

"Why does my git say main and not master? Did I break it?"

"Why cant I push to master like the 100x tutorials show? I get errors!"

Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by TrevorJ on Friday April 22, 2022 @02:51PM (#31126338)
Chesterton's Offence: Before we change a word, we should first understand why it was there in the first place. :P
Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by akomtu on Friday April 22, 2022 @01:59PM (#31125537)
This is when Ministry of Truth comes into play. The main character's job in 1984 was literally this: rewriting history when it came into conflict with the updated "truth". Orwell didn't foresee that in a world of computers, such updates are trivially made: no need to reprint newspapers and books, since all of them are virtual.
Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by saalweachter on Friday April 22, 2022 @09:49PM (#31130612)
Master character.
Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by shadowgovt on Friday April 22, 2022 @01:32PM (#31125033)
This is tech. 20 years of tutorials are always becoming slightly wrong.

Don't tell me you're still using bare pointers, `new` and `delete` in your C++ classes instead of using smart_ptr fields, or explicitly declaring local variable types instead of using `auto`...

Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by rurp on Friday April 22, 2022 @06:50PM (#31129200)
They were responding to the claim that the master -> main change has "no cost", which is clearly untrue.
Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by shadowgovt on Friday April 22, 2022 @07:18PM (#31129444)
It's an amortized zero cost because training material is continuously out of date and churning anyway. Amortized in the sense that the cost would be paid one way or the other, because people would still be updating their documentation.
Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by Shaanie on Friday April 22, 2022 @07:38PM (#31129622)
You can make the same argument about breaking changes. Code is constantly changing and needs to be updated, so breaking backwards compatibility is zero-cost?

In reality each change that requires updates to documentation (or code) is of course not actually zero cost.

Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by shadowgovt on Friday April 22, 2022 @08:33PM (#31130049)
Not only can you, we often do in the circles I've run in.

Adapt or die is the motto out here.

Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by google234123 on Saturday April 23, 2022 @01:48AM (#31131752)
> Adapt or die is the motto out here

This is pretty much the view of the left now :p

Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by Pxtl on Friday April 22, 2022 @03:46PM (#31127102)
Refactor mercilessly.
Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by valeness on Friday April 22, 2022 @01:32PM (#31125025)
Does this mean we should never make progress or change anything?

Also if your tutorials are using a base repository to work from, then you can still have the branch "master" it's just not default. So your existing repos should still work. And if you changed your repo then you should be responsible for updating your documentation to reflect that. It's just good practice.

Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by thegrimmest on Friday April 22, 2022 @03:20PM (#31126770)
I fail to see how changing the names from "master" and "slave" makes any "progress" at all. What is the most concise way to express the idea that one entity is totally subordinate to another, and must comply with every request the other sends?

I struggle to come up with any two terms that make this more clear than "master" and "slave". Just because we've abolished chattel slavery, doesn't mean we should avoid the very words themselves when they are appropriate. Destruction of meaning is far worse than some abstract offense that doesn't seem reasonable to take on behalf of a computer process.

Re:Google’s AI-powered ‘inclusive warnings’ feature is very ... by tiborsaas on Friday April 22, 2022 @01:44PM (#31125264)
What about giving master a new meaning? Actively banning it conservers its original meaning. Probably nobody today thinks about woman sitting in a room doing calculations when we speak about computers.