There is no allies to the US anymore, only clients that are expected to pay for protection and to obey.

(hi, from France).

Indeed it is the most obnoxiously hostile act somewhere between bullying and license to kill bounty. I would like to know how the system works because certainly the US is not the only player.

Is it really a case where Trump can willy nilly list whoever he wants, and all allies are obligated to obey without question? Is there no mechanism by which Italy or the UN can say “hold on, we are not going to accept Trump tagging his political adversaries as terrorists and then lick boots unconditionally”? Anti-terror policy is losing legitimacy with every abuse of it, including Trump’s assassination of drug dealers under the abusive claim of “narco-terrorism”.

Designating them as a terrorist group was several steps above mere 'bullying'.

Hence the existence of !email_is_dead@libretechni.ca

The number of self-hosters makes no difference to me. When 90+% of people I need to reach are inside the walled gardens of corps I boycott (Google and MS), then email is just not an option. Self-hosting myself does not by some magic prevent Google from seeing my messages that would traverse Google’s servers when the recipient sits in Google’s technofeudal fiefdom.

I checked date on this. Seems like this happened long ago.

That’s right, although we shouldn’t focus solely on Tesla. There are an increasing number of Chinese car brands (and I’m not even talking about all the Chinese video doorbell models) to which exactly the same applies.

I'm in USa and the router bits sold in stores are wacky expensive compared to Amazon. I also dislike this.

Most legit providers offer filter rules you can configure as you need.

Filters are for email post-delivery. By the time a server is executing your filter instructions, Microsoft’s server has already delivered the message and disconnected. It’s too late at that point. If some rare mail service were to bounce a message based on user-specified filters, it would have to be after the msg has already been delivered. This would risk “backscatter”, whereby the bounced response attempts to connect to the /perceived/ sending server, which could be incorrect. And even if it is correct, the sending server may not accept the bounced msg.

The configuration I describe can only competently be done on the server and it necessarily must be the same behavior for all users on that server because the refusal happens before the sending server even has a chance to send an SMTP “RCPT TO” line.

Unless you are intentionally finding services that do not use M$ infrastructure then you risk losing a lot of access.

It would be senders who lose access. I’m already happy to not connect to surveillance advertisers. The idea is that users on surveillance ad platforms lose access so they cannot deliver to my inbox.

Most legit providers offer filter rules you can configure as you need. Particularly if you're using a paid enterprise service. You could check to see what rules a provider has available and if you can block by MX record or IP. You wouldn't need to control the server, just have a provider that let's you customize your needs.

Unless you are intentionally finding services that do not use M$ infrastructure then you risk losing a lot of access. Flights, trains, hospitals, schools, stores, events may all use M$ as a provider whether or not they are sending from their own domain. Personally I'd rather receive my tickets that I pay for than not.

Any email service could be configured to block Microsoft.

Indeed, whoever controls the server could do that. But who does? I don’t have an always-on WAN-facing machine, so I am interested in finding an email provider who does the blocking because as an end user I don’t have the privs.

I’m not sure why you want to do that at the platform level though. That’s a great way to not have delivered legitimate email traffic.

Microsoft is a surveillance advertiser who I boycott. I do not want to feed MS’s ad machine. For outbound mail, I can do an MX lookup first and only send email to non-MS non-Google recipients, but that’s not reliable because some recipients mask their email providers using a firewall like barracuda networks. So when such users try to email me, I would like their email to be refused.

About blocking “legit” traffic, MS does that anyway. I’ve seen MS reject RFC-compliant non-spam email 1st hand. I am happy to reciprocate.

Any email service could be configured to block Microsoft. I'm not sure why you want to do that at the platform level though. That's a great way to not have delivered legitimate email traffic.

follow the link. See the title. Read the article. It’s about Matrix.

i don't see the connection here. it's not about Matrix.

You don’t. But you don’t get to make that decision for everyone you talk to. People /erroneously/ think they do, so it has become impossible to participate in a public discussion without Cloudflare in the loop.

FOSDEM 2026 demonstrates this embarrassment. Every single FOSDEM talk this year has a link to matrix.to (Cloudflare). Even if your own account is CF-independent, the rest of the crowd who you correspond with is not. You cannot isolate yourself from Cloudflare and also participate in FOSDEM rooms. That’s not ETS.

Why do you need Cloudflare for Matrix?

Delete X

True! that would be v welcome. I did learn recently about GoodRam, a European memory producer

It's an awesome start.

I'm looking forward to the possibility of a new CPU architecture not linked to the US and some EU built chips. Making the software sovereign isn't going to help if they can take away what we run it on, but it's an excellent step in the correct direction.

You’re talking about a workflow for browsing existing content. My OP is about a workflow for finding communities to post new threads to.

For browsing existing content in the decentralised world, I have configured this block list:

That gets the bulk of centralised content w/power imbalances out of my feeds. I believe I could still subscribe to $specific_community@lemmy.world if I wanted, and I suspect it would be an exception to my blocks. But I have no such desire. I subscribe very liberally to a wide range of topics spanning many decentralised instances not under Cloudflare control. My main feed (https://slrpnk.net/) is not a desert (but if it were, I would not care). I’m usually too busy to arbitrarily browse posts (i.e. doom scrolling). I still see posts from users on centralised platforms like lemmy.world /if/ the post is in a decentralised community.

As a digital rights activist who favors power balance, centralised communities are unworthy of my attention.

As for the purpose of this thread, it’s about finding places to post new threads. If I post a new thread in a centralised venue with disproportionate power, that would make me an enabler. I would be part of the problem. Selfish greed for attention causes people to prioritise exposure of their messages above digital ethics. They look for the biggest audience to prioritise number 1. Of course, some people are just naive, not selfish, as they are not consiously aware of the principles of balanced power and the decentralisation concept.

For those who are informed and also non-selfish, with a sound understanding of the problem, a good strategy is to first post in the most relevant decentralised community -- the smallest of them. Then to get more exposure on it (if needed), it can be cross-posted to other larger related communities-- but still decentralised of course. Those readers will see the source venue, become aware of it, and subscribe. So if you are clever, you can subtly promote the more decentralised communities through a post that is not directly promoting them. If you get desperate for even more engagement, you can cross-post to $whereever@lemmy.world such that those subscribers see the original community and possibly subscribe. But to date, I have never reached such despiration to compromise my values. If I did, I might take a Cliff’s Notes approach, and only put a summary of the content in LW and write that the full version is in the decentralised parent post.

Sorry for the delay; I wanted to do this on a desktop.

I want to be clear that I wasn't trying to be critical, although it sounds like it even to me. I just don't understand it. Let me try to clarify what is confusing me, and my predicates:

1. Lemmy is best used by reading "subscribed." Viewing "world" is handy, occasionally, but as Lemmy gets more and more Reddit refugees, the content quality drops and "world" becomes more aggrivating. For example (and this is not because of Reddit folks), when I view "world" the feed is frequently a stream of furry quasi-porn. I could block entire communities, and sometimes I do, but I try to wield that hammer lightly. Rather, I curate my subscription list and spend most of my time browsing that; I assume most people do, for similar reasons.
2. If a user subscribes to only sparsely populated communities, they will see almost no content in their "subscribed" feed. Normally, I would suspect folks who are doing this are doing it selectively to associate with certain groups of people.
3. A scheme that merely finds the least used communities would perforce populate a "subscribed" feed of almost no content, with random user associations. This works against #2, in which users are trying to associate with specific, known users; and it works against #1, where a user wants to actually see content, but filtered by topic.

So I don't understand this idea of finding the most sparsely populated communities and building a subscription list from it. It seems only like a recipe for wandering in a content desert.

While your mission statement says you want to increase decentralization, functionally you’re maximizing subscription to communities with the smallest audiences and least amount of content.

Indeed, the logic is sequitur. But to be clear, it’s not exactly about subscriptions. You can post without a subscription.

If you assume everyone in the FediVerse is browsing World

We need not make assumptions. The data is public.

but this approach makes topic-focused subscription browsing a wasteland. What am I missing?

Subscription has nothing to do with this. The script does not look for subscriptions. It simply looks for communities that are not part of a centralised concentration of disproportionate power.

Subscription counts are in the dataset. So it would be trivial to adapt the script to express those numbers and sort on them. But the fedi is small. There are usually not so many competing results to make that interesting. After shortlisting a community, before posting I dump the figures to the terminal so I know to what extent my target forum is a ghost town.

Why is this good?

Oof. This thread is not the place to explain the value of decentralised digital communities. The premise is that the reader already accepts that. There’s probably a blog somewhere that details the purpose of the fedi and the harms of centralisation. I’ll write something up in a separate post.

(edit) The philosophy is covered here:
https://slrpnk.net/post/24598576

I will say, if support for meta-filtering was broader - browsing not by subscription, but by key phrase match in either community or post tags, this would be a good approach.

Lemmy communities don’t have tags AFAICT. There is a description, part of which is in the lemmyverse dataset (which my script searches if specified). M/Kbin mags also lack tags, so you’ll have to clarify. Do you mean hashtags?

When you say post tags, that sounds like a huge dataset. It would be useful but the usefulness is at odds with resource efficiency.

Cloudflare avoidance is also a mark of AI training scrapers

Yikes. This is like saying opposition to mass surveillance is the mark of a criminal. I could just as well say Cloudflare advocacy is the mark of a Cloudflare shareholder.

What are you doing in ETS with this attitude? This community /theoretically/ advocates European Tech Sovereignty¹, and you are saying avoiding a US tech giant is somehow nefarious. WTF?

And again, your assumption exposes a failure to read the post. Obviously if I am separating Cloudflare-centralised listings from free-world listings, I would not even be interested in a dataset that contains CF venues in the first place. I don’t care if the dataset has CF’d communities or not -- they are useless to fedi users who keep to the decentralised areas.

so it’s reasonable to ask what the dataset is to be used for.

It is not. It’s rediculous. It’s like assuming someone looking to buy a car must be looking to do a drive-by shooting, unless they can justify otherwise. But it’s even less reasonable than that, because someone looking to attack the fedi is already coding a scraper bot anyway to interact with targets.

¹ OTOH, reactions to this thread show the contrary. Opponents of ETS are reading this community in higher numbers than advocates.

While your mission statement says you want to increase decentralization, functionally you're maximizing subscription to communities with the smallest audiences and least amount of content. If you assume everyone in the FediVerse is browsing World, maybe that's Ok, but this approach makes topic-focused subscription browsing a wasteland.

What am I missing? Why is this good?

I will say, if support for meta-filtering was broader - browsing not by subscription, but by key phrase match in either community or post tags, this would be a good approach. It'd balance the FediVerse tree. But no client I know of supports this model, and subscribing using the approach you describe seems as if it would only make Lemmy appear dead.

Please re-read my opening statement:

What I need most is an open dataset of communities other than #Lemmyverse, which is incomplete.

A dataset is typically created by crawling (scraping) by whoever produces the dataset. If I were scraping, I would not be asking for a dataset.

As for “why”, please re-read the whole post. The whole thing makes it quite clear why the data is useful and thus why someone would scrape the data, if needed. The most relevant reason to ETS is Cloudflare avoidance.

By the way, I see you are using lemmynsfw.com. What are you doing in ETS if you are happy to use a centralised US instance?