-
Notifications
You must be signed in to change notification settings - Fork 5.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Rename master branch to main branch #4945
Comments
For an advanced view into my personal opinion on this matter (which will likely come up in a longer discussion): I'm strongly against renaming things due to current political climate/pressure. What is PC changes on a near daily basis now, and it is way too much cognitive overhead to try to keep up. On this specific matter, the master/slave terminology accurately represents many constructs in software engineering, and just because at some point in history someone used the terms to represent certain relationships between humans doesn't change the meaning of the terms nor degrade their usefulness in expressing relationships between things. Also, and this is much more abstract and philosophical, I feel like in the ethos of anti-censorship, we shouldn't try to censor speech in our repositories (barring extreme cases of disruptive behavior). In this case, defaults have preference barring a good reason beyond the desire to censor certain kinds of speech, and at the moment that means |
This is going to turn into a debate very quickly. I'd also like to apologize in advance if any of this offends anyone. If it does, please do point it out to me, because I'd like to avoid accidentally offending people where possible.
I'd prefer 'current' to be defined. It's coming up on two years since Git has offered this suggestion.
Here's a proposed solution: publicize the EIPIP repo more, and have the community vote. I definitely agree with you that a four-person team of volunteer EIP editors is not sufficient.
I agree that that terminology can be useful for e.g. VMs, but I fail to see the connection for Git repositories. A In addition, other terminology may be far more accurate in expressing how systems work. In a system where the terms "scheduling node" and "executor node" apply, it seems a bit silly to use more ambiguous terms such as a "master node" and a "slave node" - especially when community consensus is against the usage of those terms.
I'm afraid I have to disagree with this point about it not changing the meaning or degrading the usefulness. Look no further than the history of Pepe memes - originally having no connotation, then having a racist connotation, and now a connotation that I'm not sure anyone really knows. But the fact is, people did make racist Pepe memes and changed the meaning of Pepe memes (giving Pepe memes, including ones made before the change a slight racist connotation), and also changed the utility of them (nobody wanted to be called racist for making a Pepe meme).
Firstly, does that extend to, for example, off-topic posts? I feel like there's a line at which it's okay to censor things if there is consensus. Secondly, whose speech would we be censoring?
Reasons have now been provided above as to why using Anyways, I always welcome feedback. These kinds of changes always work best when everyone agrees, so please provide me with any refutations you can think of. |
By this I just mean "the state of political correctness right now". There are some things that are "currently" politically correct and have been for decades or centuries, and there are some things that are "currently" PC that weren't a mere 2 years ago. Essentially, I don't think political pressures should be what drive us to choose our terminology, and definitely shouldn't be what drives us to change our terminology.
I'm pretty strongly against democracy as a solution to problems/disagreements, because it turns into tyranny of the majority (which is often disconnected from what is right/good/correct), and it is heavily subjected to human biases which leads to marketing, brigading, social signalling, etc. I tend to agree that
Curation definitely has value, but when curating I think there is significant value in limiting the curation to the most problematic stuff. It is very easy to take curation too far, to the point where you are excluding people because you don't like the way they communicate, the words they choose, or the thoughts they have. For me, renaming things primarily for the sake of following the latest trends in political correctness is across that line (censorship vs useful curation). |
I support renaming. Using My biggest concern was that it would break external references to the repo, though this appears to be addressed:
What is the cost to perform this rename? I assume person hours to rename the branch and update & test automation, which would be good to quantify. What other costs are there? Which other repos in https://github.com/ethereum use the old default name and could be changed? |
I would also like to understand the benefit to renaming. The only one I am aware of is that the proposed name aligns better with a particular political viewpoint. Is there some technical benefit from renaming? Understanding the benefits will help us weigh them against the costs. If the only benefit is satisfying one particular political viewpoint (while simultaneously offending another), then I think "do nothing" is the most neutral stance we can take (and I think neutrality is good here). |
The social benefit is to remove a name that is offensive to some people. A name change makes a repo more inclusive and welcoming. It's unfortunate that people who are not offended by an existing name would be offended by a name change.
The only technical benefit I can see is consistency with other open source projects. The default in GitHub for new repos is The benefits are social inclusiveness and minor technical consistency vs offence by anti-renamers, costs of renaming + any risks of breaking things. [edit]: It is on the agenda for EIPIP meeting, so that meeting can decide if the status quo is better or if the benefits of renaming out way the costs. |
I'll open an issue on the EIP bot repo, in case this is an issue in that case. |
I am neutral on the decision to rename the
If I called you a "git", would you be offended? Is the term "commit" offensive to people involuntarily committed to mental institutions? |
@SamWilsn Other prominent organizations have performed this rename. The Ethereum community has previously shown that it believes words matter with the rename of an opcode to |
Also, I don't think that this has any risk of breaking any tooling. Renaming to |
Considering the lack of progress, there definitely needs to be a protocol for these types of changes. I would say the EIPIP repo would be ideal, but I think that the purpose of EIPIP meetings is different enough from determining what is or isn't offensive that the task might warrant its own meta-EIP and corresponding protocol. |
I didn't mean to imply that |
A simple approach is to:
The examples you gave of |
I support renaming too. Language and the definition of words in certain context changes in society, it isn't uncommon. As @abcoathup mentioned, EIP-6 has proposed this sort of change before. This change of branch name is simply a change in naming convention to adopt an industry wide standard. This switch wouldn't cause any errors from what I can tell. The term |
Just because someone claims X does not make X true. None the less, I am willing to accept that there exists one or more people who are offended by the name. However, as Sam illustrated, given 8 billion humans in the world you can likely find one or more people offended by just about every word in the dictionary. Presumably, there exists some critical mass of offended people that warrants changing a long held speech pattern, and it is unclear that such a critical mass is reached. We do not have a good credibly neutral and repeatable mechanism (like Sam discussed) to determine whether a critical mass of offended people have been reached.
Follow-the-leader isn't a reasonable strategy IMO. Other prominent organizations also build centralized services, harvest user data, spy on you, etc. (all things I think we agree Ethereum should not do) I think any arguments in favor of this change should rest solely on the merits of the change, not because "other people are doing it".
I wasn't around for that change, but it was in a time when the community was far smaller and the number of people you needed buy-in from was also much smaller. Ethereum in that time also wasn't nearly as credibly neutral as it is now (see The DAO hardfork). FWIW, Uncles was also renamed to Ommers around the same time I believe. |
I - and others here, I believe, have a hard time understanding your argument against the change of branch name. In the last comment, you have used ~250 words to argue the change of one (1) word. It seems trivial, and I feel you're letting your personal, political opinions cloud what is correct in the present social context - in your political view, it is PC (politically correct) to use the word "slave" casually, but that doesn't make it acceptable or correct. You could have put forth your political views in about one line. |
I appreciate that a mechanism for deciding on making such changes is required, especially as potential changes could be contentious. I am also mindful that we have limited people involved in the EIP process, and we want to make the most of this limited resource. The appropriate forum to move this forward appears to be the EIPIP meetings. It was previously parked due to lack of a champion on that call. If someone reading this wants to champion this issue and is available in the meeting timezone, then they can ask to attend an EIPIP meeting to help bring this to a close. |
@abcoathup I'm available during that time and I'd be happy to lead the issue. |
Even though I'm not of African American decent, it's definitely offensive to me whenever I hear it. I get a little sad inside when it comes up since it reminds me of the 400+ years of slavery and wicked oppression even up to this day. Although I fully recognise its not the intention of many, I can't help but think about it. I don't know if this original naming was intentional propaganda or pure ignorance. If the current naming doesn't offend you, it's probably because it doesn't affect you. I really wish tho the black community within the Ethereum community would have more input and say on this topic.. instead of a bunch of whites, Arabs, Latinos, and others discussing it 😁 Linus, creator of git and Linux, said before he always liked The naming is part of a specification and specs evolve. How can Ethereum expect others to follow its specs if it has a hard time following others' specs. Ethereum is a leader and others copy its convention blindly; this change will cascade to other ecosystems that depend and lookup to Ethereum. We're not just a bunch of robots, we have social and human aspects that should be a factor in things. Plus as a bonus, |
For issues like this I personally would prefer to keep the discussion out of the call until we decide it is time/necessary to make a final decision. Unfortunately, political topics tend to consume huge amounts of time in calls and are rarely productive, which is why All Core Devs calls (and many other technical open community calls) have gone to great lengths to avoid/minimize them. Do people here feel that there is nothing more to be gained from asynchronous discussion in this thread? |
I don't think anyone has a problem with this change, and at the same time everyone is concerned with the side effects. If @corbanvilla or someone who has enough context to make this change can write up the steps involved, I think the concerns would melt away and we can move this to a synchronous discussions for when to officially pull the trigger. Knowing what's involved is key and hoping we can materialize that here. I've made this rename change on some GitHub repo's before and was super simple, and had the team update their clients and was barely a 5 minutes change and disruption:
I totally realize and appreciate there's much more moving pieces here and fully share the concerns, but with the automatic aliasing in place it can be a slow and gradual migration. |
I'm keen to see a quick resolution to avoid it becoming politically heated & drawn out. Even if the decision is to keep the status quo for now. Hence thinking a timeboxed synchronous discussion could be the best way to get to a resolution. Given the number of repos in the EF GitHub org what is the best forum for a name change discussion? I don't know what autonomy each repo has. Finally, whilst I think changing to more inclusive language is important, I don't want it to delay the merge (switching off PoW). |
That definitely makes sense, I'll continue to monitor this channel for additional input in the coming days. If we conclude most arguments have been voiced by early next week we can make a decision.
There's definitely a large number of repositories that fall under this. Though most people in favor of this would probably prefer a change in all Ethereum repositories at once, we could also consider a gradual phase change. Fedora project did something like this. It also appears that there's an API endpoint to enable this change. I would be more than happy to write a script that would apply this change globally. I appreciate how busy everyone here is and want to minimize the cost on everyone's time and energy. I also know this change would make a difference for a significant number of people and continue bringing more talent to the Ethereum space. A gradual phase out would look something like:
|
I am much more strongly against the Ethereum Foundation setting rules on language usage that are applied across all repositories in the Ethereum GitHub organization. That being said, I highly doubt they would do such a thing as they have historically taken very back seat (credibly neutral) approach to anything that may be considered stearing the community. Each GitHub repository has its own set of owners who each have (as individual groups) a lot of autonomy over their respective repositories, and this is a very good thing IMO. I don't think any of us should be getting involved with how go-ethereum names their branches, as none of us are active go-ethereum developers (the same for all of the other repositories). |
I quite strongly favor no-change if the discussion is cut artificially short, especially for contentious/political topics where being credibly neutral is a desirable feature. |
That's fair, you definitely have a point there. Taking that into consideration, I think this conversation should stay pertaining to EIPs. I could however see a potential benefit in official guidelines published by the Ethereum foundation, like a Code of Conduct which outlines recommended inclusive language. Then it's up to the autonomy of each repository how those guidelines are used, interpreted and enforced. Here's an example. I acknowledge that this is probably not the forum for that conversation, though. |
Such an overarching policy could be an Informational EIP, and then its discussion thread might be a good forum. |
There has been no activity on this issue for six months. It will be closed in a week if no further activity occurs. If you would like to move this EIP forward, please respond to any outstanding feedback or add a comment indicating that you have addressed all required feedback and are ready for a review. |
There has been no activity on this issue for 1 week. It will be closed after 3 months of inactivity. |
@Pandapip1 should this issue remain open or have the editors decided not to do it at this stage? |
This is just stagnant. It might be re-opened later. I'll change it to "not planned." |
See: https://www.theserverside.com/feature/Why-GitHub-renamed-its-master-branch-to-main
Issue on EIPIP repo: https://github.com/ethereum-cat-herders/EIPIP/issues/122
So that the EIP editors can accurately gauge if the community desires this, please vote with 👍 or 👎. (I've reacted with both so that both reaction buttons appear, so subtract 1 from both counts).
List of EIP-related repos with
master
as the primary branch name: #4945 (comment)The text was updated successfully, but these errors were encountered: