See crosspost(s) for more discussion:
This is the first big step in the process to develop comprehensive guidelines for the Fedecan non-profit and the various platforms.
While this will mostly involve converting tacit knowledge and experience into an explicit written form, we expect that this process will inevitably bring up some points of disagreement on the best way to deal with different issues. We ask everyone participating in these discussions to please contribute constructively and in good faith. We encourage you to bring up any concerns or issues you have with the proposed structure and drafted guidelines, so that we can work together to fix them early on. However, in order to keep a productive environment for those discussions, we will be pruning any comment chains that devolve into personal attacks, slap fights, etc.
To help ground your feedback, consider these thought experiments when evaluating a potential guideline:
- Veil of ignorance: Would it still feel fair to you if you switched places with someone else on the platform (ex. a new user, a moderator, an admin, a member of a vulnerable group, etc.)?
- Equal Applicability: These rules will be enforced uniformly on everyone. A poorly written rule that helps “your side” today, can easily harm “your side” in the future as circumstances change.
The full guidelines, including governance details like the annual review cycle, can be found on the website: https://fedecan.ca/en/guidelines/
We plan to structure the guidelines as follows:
Tier 1: Fedecan Rules
Internal Conduct
These rules apply to Fedecan team members (directors, officers, admins, and anyone with elevated access). They set expectations for how team members should act.
Universal Rules
These are the baseline rules that apply to every user on every Fedecan platform. They cover the things that are prohibited by Canadian law (threats, hate speech, CSAM, non-consensual intimate imagery) as well as universal policy rules (privacy/doxxing, harassment, fraud, content that could cause harm, labelling of sensitive content, etc.).
Tier 2: Platform-Specific
Each platform has different functionality and norms, so this is where we can be more specific with the rules. The threadiverse platforms (lemmy.ca, piefed.ca, sh.itjust.works) share similar rules around community creation, moderation, vote manipulation, and content labelling. Pixelfed has its own rules tailored to its platform.
Tier 3: Community-Level Rule Templates
These are optional templates that communities can link to, or use as a starting point for their own rules. The idea is that moderators can point users to a clearly written explanation of why a rule exists, and any relevant exceptions, rather than trying to fit everything into the sidebar. Additionally, if many communities are enforcing a particular rule in the same way, then users will have an easier time understanding and following them.
The post title standards template has been drafted, and we plan to add more as the need arises. I have a few others that are in the works, but they have some overlap with the other sections, so I thought that it would be better to let people discuss first.
After reviewing the universal rules, I have two main concerns.
The first is the reliance on Canadian law. I understand that these instances are subject to Canadian law, which makes sense, but I would rather there be explicit rules or values written down, with the preference toward those rules over Canadian law, as laws can change. For example, what would happen if Canada has a MAGA-style wave and removes LGBTQ people from its list of protected classes? I imagine that we would still want to protect such people. Or, perhaps more likely, Canada strengthens its ties with China and regulates certain criticism of China as a result. Canada could also enact laws similar to those of the UK, which classified support of Palestine Action as support for terrorism and outlawed such speech. In summary, I want to make sure that these Lemmy instances protect marginalized groups while also allowing for free discussion and criticism of foreign nations and political groups, regardless of the current sway of Canadian politics and laws.
The second is a more concrete example which reveals an inherent contradiction between universal rules 1 (Violence and Threats) and 2 (Hate and Discrimination). There are currently multiple real-world examples of political groups using bogus victim-hood narratives, astro-turfing, whataboutism, and other means of misinformation to deny or downplay human rights abuses, including genocide. While I am glad to see a rule about misinformation, certain interpretations of universal rules 1 and 2 could shield bad actors from necessary criticism. In other words, should we tolerate the intolerant? Is it OK to say we should punch Nazis? What if I claimed Israel was committing a genocide in Gaza - a fact that many Canadian politicians are unwilling to admit themselves? What about the humans rights abuses in the USA and China? Would it be OK to use these Lemmy instances to stage opposition to such abuses, including sabotage of the encroaching surveillance state? The answers to these questions may seem obvious to some, but I would rather have explicit rules in place than kick the can down the road.
I think this might be a matter of better explaining how our guidelines relate to Canadian laws, rather than needing adjustments to the guidelines themselves. This draft is actually already doing some of what you’ve described. We meet the specific thresholds that Canadian law sets, and then set our bar high enough to account for the areas that the law isn’t covering properly.
In the first example:
For example, what would happen if Canada has a MAGA-style wave and removes LGBTQ people from its list of protected classes? I imagine that we would still want to protect such people.
This shouldn’t affect our guidelines when it comes to Tier 1 Rule 2, which we wrote as follows:
We do not allow statements that incite or promote hatred against an identifiable group, as defined in the Criminal Code of Canada.
In this section, identifiable group means any section of the public distinguished by colour, race, religion, national or ethnic origin, age, sex, sexual orientation, gender identity or expression, or mental or physical disability.
We extend this definition to include any group that a reasonable person would recognize as a target of hatred, consistent with the intent of the law. Disagreement and sharp criticism of someone’s actions or beliefs are not covered by this rule, as this rule looks for content that targets people based on identity rather than as an individual.
https://fedecan.ca/en/guidelines/1-universal-rules#rule-2-hate-and-discrimination
So even if a group of people was no longer listed as an identifiable group, our rules would still protect them since they are a group that a reasonable person could recognize as being a target of hatred.
The second example is more complicated. This is something that the instances located in the UK and elsewhere are currently dealing with. We are subject to Canadian laws, and while we will definitely advocate for laws that make sense and protect users online, we will have a limited capacity to outright refuse to follow a law after it gets passed. This is something that everyone will need to advocate for BEFORE something becomes a law.
For the second paragraph of the comment, there are some points where we will need to get advice from legal professionals before we decide anything explicitly. Without having talked to the rest of the team yet, I can share my thoughts on a few of these:
Is it OK to say we should punch Nazis?
Under rule 1, we say that “This is based on whether a reasonable person would interpret the statement as a serious expression of intent to cause harm”. So if a reasonable person would see the comment as hyperbole and political speech about opposing violent ideologies, then that wouldn’t be against the guidelines. Meanwhile, if it was a direct call to organize to assault someone, then that would be against the guidelines.
For the next few points:
- Criticizing a government’s actions is not hate speech
- Criticizing a state’s policies is not incitement against a group of people
- Criticizing the actions of a group of people is not hate speech, but it can become hate speech if it is extended to everyone of a similar identity.
Would it be OK to use these Lemmy instances to stage opposition to such abuses
Yes
including sabotage of the encroaching surveillance state
That depends on what this refers to. Again, we’re not going to outright refuse to follow the law, or else we will simply get shut down and the volunteer team members will face legal consequences.
For example, what would happen if Canada has a MAGA-style wave and removes LGBTQ people from its list of protected classes?
Nothing should change because presumably principle 2 covers them.
“Well-being of people who use our platforms and everyone affected by them”
A few discussions:
- “Content that was created with the help of generative AI tools should be labelled.” What if you don’t know for sure, you’re reposting. Can I put “suspected AI?”
- Your Rule 6 in the universal rules is going to be hard to enforce. I like that it’s there, but the fediverse is so small right now, that we rely a lot on user reports. I’m not sure how to fix that. I don’t think some of the users here know they’re even being targeted, and then leave before it’s reported.
- Rule 9 is hard for us Americans. Our own government is doing that. Again, we’re small, so that one probably isn’t a huge issue yet.
“Content that was created with the help of generative AI tools should be labelled.” What if you don’t know for sure, you’re reposting. Can I put “suspected AI?”
I think with a lot of these guidelines, we are relying on people’s best judgment. We’re not going to be banning people for honest mistakes, and we likely won’t be taking down content when it is unclear if it is AI or not. Users can definitely say that they are not sure when they are the ones that are posting it.
I can think of two cases where this would be enforced:
- If a piece of content ends up being dangerous misinformation because it is not labelled as AI-generated. We may remove that content if it can cause harm, instead of waiting for the OP to deal with it.
- If a user has a pattern of posting AI generated content in order to mislead people, or if the account is an OpenClaw bot altogether, we may ban them.
Something I’m realizing, for both of these cases they would be breaking some other rule by not labelling the content. Maybe there is a better way to organize that rule.
Your Rule 6 in the universal rules is going to be hard to enforce. I like that it’s there, but the fediverse is so small right now, that we rely a lot on user reports. I’m not sure how to fix that. I don’t think some of the users here know they’re even being targeted, and then leave before it’s reported.
This is true, and it will apply to a lot of the rules. If we can’t detect something, and it doesn’t get reported, then it likely won’t get actioned. Hopefully once we have some solid rules in place, we can work on tools to make sure we are taking action consistently and fairly.
Rule 9 is hard for us Americans. Our own government is doing that. Again, we’re small, so that one probably isn’t a huge issue yet.
On the bright side, many professional organizations in the US are putting out correct information to counter the disinformation shared by the administration. As we grow in size, we might be able to get more help from experts to identify and deal with misinformation that causes harm. I’m also thinking of putting together another Tier 3 rule template for communities that deal with misinformation often, with resources and guidance on how to approach the gray area topics.
Something I’m realizing, for both of these cases they would be breaking some other rule by not labelling the content. Maybe there is a better way to organize that rule.
I don’t know how you’d do that, but that sounds great too. When you say labelled, are you saying in the title or in the body? IMO, I think the body is best case because so many AI mentions in the title would get weird and seem like we’re promoting AI.
Hopefully once we have some solid rules in place, we can work on tools to make sure we are taking action consistently and fairly.
That’s a great way to look at it and I agree.
On the bright side, many professional organizations in the US are putting out correct information
I’ll keep a look out for them as well as I go through the news.
Thanks for putting a lot of work into this and opening it up for discussion. You’re doing a great job.
Your Rule 6 in the universal rules is going to be hard to enforce.
I had someone create my user on 4 separate instances to downvote me. You can take the redditor out of reddit…
Lol, I’m always surprised to see the trolls here. I don’t really see the point since we’re pretty small.
Thank you for organising and sharing your rules and processes transparently. Scanning quickly, there’s lots here that must have taken a lot of work. The various hierarchies for users and rules seem sound. The rules themselves line up with what I’d expect and hope for as a user, particularly one based in Canada. I’ll check back to see some of the developing sections, like the AI guidelines, once theyre done. Thank you for including a changelog, and a commitment to annual review and updates, and what seems like a commitment to including community feedback.
Thank you :) I’ve been collecting notes from when users made suggestions, and that helped to get everything organized. It took a lot longer than I was expecting because there are so many resources out there to read and get ideas from.
I’ll check back to see some of the developing sections, like the AI guidelines, once theyre done.
Do you have any quick thoughts on what you want to see for that section, or any concerns that you want to see addressed?
Eh, I’d like to interact with as little AI as possible, recognising that that’s increasingly unlikely and realistically altogether impossible… but I’m from the before times so I’ll hold out hope. I imagine you’ll all be able to come up with reasonable, actionable guidelines that are more realistic than “No AI at all” and that you’ll sort through the challenges as they come up. Thanks again.
That sounds good, I’ve noted that down. Cheers :)
I think rule 5 should be amended to allow posting of PII for public figures with publicly available information however disallow any private information (like home address or personal cellhpone) of these public figures.
The current wording would mean you can get banned if you give out information on how to contact your MP (for posts about bad bills for example).
That’s good feedback, we can add a few lines to clarify that you’re allowed to share publicly available information. Maybe anything that is already shared publicly by the organization in question, such as information on the contact page of a company? I see some more feedback around rule 5 in the comments, so I’ll wait a bit before drafting up a new version.
I have made a note of your suggestion, thank you! Also happy cake day :)
Not to just make things harder but just about everyone’s address and phone number is publicly available, so maybe some language specifying public figures.
That’s true, we’ll need to refine this one quite a bit. My gut says that anything that has been shared by the party in question, would be ok to share for the purpose that it is being used for.
So if a company shares contact information, or a public figure has contact information shared, then that would be ok to share that.
If a private individual shares a contact method for a specific purpose (ex. organizing an event, or they are selling something that they make), then it would be ok to share that contact method with people who want to use it for that purpose. However, it would not be ok to share that contact method for some other purpose, such as harassment or spam.
Meanwhile, personal information from data mining companies would not fit in to this?
Based on the Canadian charter? As a native American man, that doesnt fill me with a lot of confidence.
That is fair feedback. I was actually intentional with how I wrote that section:
We make up one part of the Fediverse, and we intend to have a positive influence on the rest of the network. Living in Canada, we’ve been shaped by frameworks such as the Canadian Charter of Rights and Freedoms, and its values naturally shape how we approach the decisions we make about our platforms.
My intent with that paragraph was to point out that our guidelines and enforcement may differ from how other jurisdictions handle similar content. The best way I thought to describe that was that as a Canadian team, those decisions will be guided by the social/cultural/legal frameworks that we are most familiar with, one of which is the Charter.
While doing the research for this, I came across the guidelines of some other platforms that explicitly tie their rules to the Charter. However, I didn’t want to do that. While the Charter does offer protections that other parts of the world don’t have, there is also plenty of academic discussion and criticism about the Charter and its limitations. It is a product of the political moment when it was created, and while it can shape some aspects of how a modern online platform should be run, there are going to be gaps that we need to address ourselves.
A related example is our Tier 1 Rule 2. We start with how the Criminal Code defines an “identifiable group”, but extend it to any group that a reasonable person would recognize as a target of hatred. The list on the Criminal Code goes back decades, and it takes a lot of time and political will to update it. An online platform meanwhile will need to address issues around hate quickly, and so we will need to use our best judgment on when we need to act.
I would love to clarify the wording around the Charter (and any other sections) to address the concerns, if you have a moment to share more thoughts on what to work on :)
I like what you’ve said, I guess I was just off put by the language. “Shaped by” the charter is very different than “inspired by” the charter. My worry is that if you tie yourself too strongly to an established legal framework, you bind yourself to the biases and loopholes of that legal system. You lose the flexibility and adaptability that I feel is more important in this particular format and forum. Say the charter is altered or expanded upon by conservatives in a way that discriminates against or targets lgbtq lemmys, tankies, or any other minority group. Would you be then forced to follow suit? Maybe rewrite your rules to exclude that portion of law? Thats a slippery slope all on its own. Section 35 of the charter only came about ~40 years ago. Its not a perfect document.
I would much rather read about policies and practices inspired by the intention and language of the Canadian charter than ones shaped by the realities of its current day application.
Case in point. Abuses still happen today.
I agree, I will come back to this comment when I make those changes. Our intent for the guidelines is in line with what you are saying, so it is a matter of clearing up the wording.
I did discuss some of those other points in another comment chain. We can definitely make it so that our guidelines offer more protections than the law requires. However, we also will have a limited ability to outright refuse something that becomes law, else we would just get shut down: https://lemmy.ca/post/67952523/24386875
I like this line in particular, I think we can rework that section around this line:
I would much rather read about policies and practices inspired by the intention and language of the Canadian charter than ones shaped by the realities of its current day application.
Thank you for the screenshot by the way. I went looking for articles about what the Namaygoosisagagun community is going through, and I agree with what Avi Lewis posted. If anyone else comes across this comment chain, this article has a good introduction: https://www.cbc.ca/news/canada/thunder-bay/namaygoosisagagun-first-nation-wildfire-9.7272350
You mean the charter (section 35)where indigenous treaty rights are enshrined? See any of that in the US constitution?
Tell me where section 35 punished anyone for the atrocities committed against my people? Settlement is not punishment. When are the innocent children of white settlers scheduled to be taken from their homes and forced to learn indigenous customs and language? Whipped and beaten and raped for talking back? How about a fat cheque for me in compensation for the french last name i had my entire life?
Settlement is forgiveness, and Canada is not forgiven.
Looking forward to diving into this.
Please do :)
I haven’t had the time to give it the proper attention I wanted to yet but I assume it will continue to be open to feedback?
Yes, I don’t think we will be finalizing the guidelines anytime soon, and not for another month or two at a minimum.
They will always be open to feedback even after we finalize the guidelines, but non-urgent changes will happen on a predictable schedule so that users can keep up
Please keep the rules in the sidebar so people on mobile devices can read them without leaving their client.
Definitely!
My plan was to have a short version of the rules in the sidebar of each site, with links to the relevant section of the full guidelines when people want to read more. That way the user can skim through the rules that don’t apply to them, and read more for the ones that they are curious about. We could also put the entire rules in the sidebar, but I think that might be overwhelming and result in users not reading them.
Honest input to your thoughts - as a registered/logged-in user, that Lemmy home page infobox is a UI/UX nightmare - it blocks my quick access to the second box below that has my subscribed communities. Upon visiting, I would immediately collapse that top box without ever looking at it; I now have a simple uBlock Origin rule that hides the top infobox on the home page for me automatically. On old.reddit, the list of subscribed things is a dropdown top left so they avoid this UI/UX problem with stacked infoboxes.
Yea that’s a good point. I know the Lemmy devs are working on a new frontend, and it would be nice to be able to customize some of these things.
Have you tried https://photon.lemmy.ca/ ? It is more customizable and it puts the subscriptions and sidebar on different sides
we have a few other frontends: https://fedecan.ca/en/guide/threadiverse/alternative-uis
(just talking shop) I’ve poked at them, you’d think the old.lemmy.ca would be appealing but not for me - I rather like this Lemmy webUI overall as a design. Good font sizes, line spacing and so forth - it has a well formed comfortable feel which I can tell has an experienced eye to it (minor details matter). Conversely, when I look at, say piefed or old.lemmy I just got “ugh, those fonts! that line spacing!”. The default UI is actually a well crafted piece of usability kit - the infobox is an Achilles heel.
Random: I’ve used the Lemmy webUI details to craft similar feeling looks to HackerNews and (old) Reddit using nothing more than uBlock filters. Nothing grandiose but the fine tuning makes them sorta feel like the Lemmy UI (but worse, holy cow is there so much hard coded HTML in these old sites - lipstick on a pig but it helps).
