🎓 Dr. Freemo 🇳🇱 @freemo@social.cleverlibre.org

@oliphant so in terms of receipts you have two categories here...

One is not receipts at all, like the attached screenshot, it is literally someone making accusations and a screenshot of that. That is not a receipt.

The joepie "receipt" is closer to being a receipt, its a screenshot of something i said at least. But it was also exposed as being a lie as it intentionally left out content to manipulate my response, which i replied to that link with thr full receipt. Lets break down that receipt shall we:

I woukd ask if i woukd defederate with child porn. I said no one had ever reported child porn so it wasnt an issue that had came up before... i immediately brought it up to the moderators at qoto and within 30 minutes got a preemptive rulling ton child porn and within moments of being asked the question we made a rulling and updated our ToS to address child porn and prohibit it. So within minutes in that same conversation i updated them as to our policy... that was of course left out of the screenshots andni have provided them.

Now ask yourself why did a server provide receipts that intentionally left out such an important response in the thread, a response that completely changes the receipt.

Do you really think receipts so blatantly manipulated and dishonest without showing important details and rejecting them when presented has any place on a publictly curated and adopted list? Does that really even count as a receipt when its so blatantly manipulated out of context?

@oliphant

Would you want to work together in expanding tbis system and try to address some of the concerns. I already have tools ive been developing on my own to address some of these problems and i think youve got a good head on you to do it

@oliphant we are mostly in agreement about needing a good block list for most new servers..

What gets me is the very low quality choices on the council, nonrequirement for receipts or any way for users to effectively research blocks, and the fact that notorious instances like rage.love known to lie and manipulate to encoirage their block list on others, makes for a bad aggregate list.

Getting 50 servers on the council could potentially address the server, but if a list of 50 servers has the same poor quality members as the 10 you have now then you will have the same problem.

I think if you want to solve the problem then the solution is not to simply have more servers on the council but rather have criteria for what servers can be on the council. Id rather have 3 server council that has well maintained receipts and transparency around their receipt process than 100 servers all hand picked to look like rage.love with few receipts and a bad reputation for abusing their block list.

The best solution is to build a receipts database and ensure any participating server must have receipts on a block throuvh the receipts server in order for the bloxk to be calculated at all. At least then users of thr blocklist will have the tools to do their own due dilligance

@oliphant fair concerns and yes if those instances make the sort of posts you claim yea id ubderstand why they are listed.

That said i woukd not be suorised if i went through the list in earnest id find many false positives when you have instances like rage.love that have been caught lying and manipulating the receipts around their blocks.

You picked some of the worst most abusive blocklists known on the fedi to be on your "council".. bad data i and bad out. Qoto is a prime example of that, a highly moderated pro-lgbtq instance with absolutely no record or receipts of wrong doing.. i have no doubt others made it to that list.

@oliphant 80%, no 60% is enough for your largest T0 list, and T0 is recommended as minimal "start here" type lists. Plus out of, what was it like 10 hand picked authority servers, and of that very small number you pick instances like mastoart which is outright abusive with its list (citing reaasons like an instance even **federates** with another instance they block), rage.love also has a history of draconian and even arbitrary blocking criteria.

so yea when there is a selection of 10 servers, some of whom are well known to be abusive with how excessive their block list is... then numbers like "80% agree" doesnt mean much.

As for instances that have no place on a T0 "start here" sort of list but seem to be on one of your T0 lists (good-actors):

* berserker.town
* qoto.org
* detroitriotcity.com (I may be misremembering these guys though i need to check if they are the server i am remembering)
* scots.network

These were the four that caught my eye.

@oliphant

If you want my opinion... instead of blindly basing blocks on how popular a block is, start with receipts... **require** any block on any list to have a receipt.. start a universal receipts system (I wrote one already, havent released, happy to team up).

That alone will eliminate the vast majority of the issue.. users can, as you say, actually validate against receipts, and the accused should (if the receipt system is open) be guaranteed the ability to provide counter-receipts, true transparency

@oliphant The whole point of using something like this, a T0 list, is that they are picking a "minimal" set, very few if anyone is going to start with a minimal set, especially with such a long list as that, and review it one by one.. and without good receipts they cant really evaluate it well either...

Id argue starting with T0, considering the main T0 list linked is riddled with good-actors, is far FAR worse than starting with 0 and building up.

@oliphant Yea I get that, and it is good you provided a shorter list... but having seridy in a T0 list, any T0 list just isnt a good look IMO. Its being recommended as a choice for a starter instance and as a "minimal" block list, despite being riddled with good-actor instances that have no place on a "minimal" block list is all im saying.

@oliphant

I am not talking about the size of the blocklist, im talking about the size of the instances that do the blocking.

As you probably know there is a list out there that lists how many other instances block an instance. This in turn is used to populate many block lists. In turn this appears to have been gamed by many using mostly one-user instances toblock an instanc een masse to make it look like they are more blocked then they are.

It looks that many of your "trusted sources" use this list directory or indirectly to create their own list, largely because it appears that quite a few instances with little or no justification appear on your tier 0 list.

You yourself federate with us (QOTO) and im sure are well aware we are strictly moderated and dont engage in any activity that could possibly justify a minimal list like a tier0 list, yet there we aare along side many other good-actors.

I think int he end the fact that the tier0 list is huge and has so many good-actors on it discredits the list out of hand. I am glad the process is open, but I'd argue the issue isnt with the process itself but with the premise.