Converting Bitcointalk to a decentralized Forum (Non-technical explanation)

[dkbit98]

I personally don't get the obsession with changing everything to decentralisation, since it isn't something that needs to be implemented into everything, in fact I tend to believe in certain situations, decentralisation is a horrible idea.

I agree with you, people wanting to decentralize everything are either ignorant or just a dreamers who don't know how things work in real life.
Decentralized Bitcointalk would be terribly slow and unusable, but I would like to see certain parts of forum using decentralizations, maybe for some kind of voting, or things related with Ban/Unban.
This is not easy to implement, and I am almost certain this could not be implemented correctly in current bitcointalk forum software.

As much as I like the idea, there's a reason I haven't seen any real decentralized forum. The cons don't outweigh the pros.

It's not impossible to make it, maybe based on something similar like Tor, but everyone would probably have to download everything, and I am not sure many people would do that.
I know some decentralized forum attempts, but they never were 100% decentralized and they never got more attention or popularity.

[1715308604]

1715308604

[1715308604]

1715308604

[1715308604]

1715308604

[1715308604]

1715308604

[1715308604]

1715308604

[LoyceV]

It's not impossible to make it, maybe based on something similar like Tor, but everyone would probably have to download everything, and I am not sure many people would do that.

I expect something similar to how Bitcoin is used: some people let Bitcoin Core download the full blockchain, but the majority uses light wallets that rely on a centralized server. Or multiple centralized servers, such as Electrum is using.
When a decentralized forum gets too big, we'll end up with centralized servers again.

[Welsh]

I expect something similar to how Bitcoin is used: some people let Bitcoin Core download the full blockchain, but the majority uses light wallets that rely on a centralized server. Or multiple centralized servers, such as Electrum is using.
When a decentralized forum gets too big, we'll end up with centralized servers again.

There's not a whole lot of incentive to do so, and you're right except I think it'll just be a select few which would host the servers, which is centralised to a point anyway. I guess you could argue that 4/5 users hosting the servers is better than one central figure, but Bitcoin works because there's a common interest in hosting nodes, and it's beneficial to a currency they use.

Most users aren't concerned about the integrity of the forum, at least because they don't think it's being maliciously altered in the first place. So, the motivation or incentive is much lower.

I wonder how many people would store all data knowing bounty report has relative big portion.

I imagine we would have a few people trying to remove the altcoin section completely. It's something that has been talked about for years, and we know from very senior users stating publicly they actively boycott the section. So, if we had a decentralised way of doing it, it might actually be pushed out if the 4/5 people who would be interested in hosting the data agree to remove it. 

The altcoin section definitely does carry a stigma here, which to a point I get, but ultimately the forum wouldn't be as successful as it has been without it. The bounty section is definitely problematic though.

[LoyceV]

I wonder how many people would store all data knowing bounty report has relative big portion.

It's still tiny compared to the blockchain, so I would store it. In fact, I have a copy of all posts already.

people attempt to mitigate it by only store/relay certain data.

That would work too: let the bounty spammers keep their own database.

[NotATether]

Now, given that the name of this site is bitcointalk.org, do you really think that Theymos will move the entire forum on to the blockchain of some shitcoin? There is no dApps support on pure Bitcoin mainnet.

Decentralization is a human right, but when applied unnecessarily such as in places like this, it becomes dumb.

[BlackHatCoiner]

If your concern is archives (in case someone shuts the forum down), you can use various places to read posts, such as loyce.club, ninjastic.space, web.archive.org, archive.ph etc. A decentralized forum doesn't make much sense. To have freedom of speech we need human moderation; I know no way a computer can be programmed to do that.

A truly decentralized forum cannot exist, because there is no way to punish those who'll infringe other people's freedom of speech. There could be bots everywhere. Hell, even in centralized platforms there are.

[KingsDen]

I like that fact that OP sat down and reasoned all these out to share with the forum. I also appreciate the fact that some of his ideas are applicable, even if it will be bogus to implement. But what I don't actually is the compelling reasons to run a decentralized forum. If this was necessary, Satoshi would have implemented it before now.

If securing the forum database is the problem, there are many other ways to do that.

I have developed and discussed a merit-based ranking system that's fraud-proof/fool-proof ... it can only rank up reputable moderators/members. Incase you are interested, let me know and I will share it here.

I will like to understand your theory of the merit system that will replace the human conscience with machine codes. I highly doubt there's a chance of a better merit system than what we have now.

[LoyceV]

Allow me to quote theymos here:

Decentralized forum software has existed since before Bitcoin. The most successful example that comes to my mind is Freenet's FMS. However, since decentralized forums can't have normal moderation, using them generally requires more responsibility and work from readers. The vast majority of people don't want to do this, which is one big reason why decentralized forums are very obscure. (Another reason is that there are very very few developers working in this space, so the tools are often not so great, both from a usability standpoint and in general. IMO Freenet is very unlikely to actually be secure in the face of serious attack, for example.)

Increasing fault-tolerance is a long-term goal, but (re)creating a truly decentralized and uncensorable forum is outside of bitcointalk.org's scope.

They've existed on Freenet since before Bitcoin existed, and they function pretty well. (Usenet is arguably also a very old form of decentralized forum.) The problem is that:
 - People are too lazy to install software to use a forum.
 - People don't want to be responsible for their own moderation. There are effective ways to do it which aren't manipulable and don't require wading through endless spam, but it's still work which every user must do for themselves. 99.9% of people simply won't do it. (Freenet's FMS inspired bitcointalk.org's trust system. Imagine the forum's trust system, but for moderation instead of trust, and no DefaultTrust.)

So there is essentially no demand. If you built it on the "new hotness" ipfs (which is Freenet but much worse... and I'm not a fan of Freenet's design), then maybe you'd get a bit of attention for a while, but it'd eventually fade.

If you wanted to implement Merit in a decentralized forum (ie. one in the vein of Freenet's Frost or FMS), you could do it in this way:
 - Everyone can, from their own perspective, give unlimited merit to posts, and these merit transactions are put into files which each user publishes via the decentralized system. (Like a merit.txt.xz which every user publishes.) Unlike on bitcointalk.org, you can also give people merit without an associated post.
 - For everyone who has merit, you download their merit-transactions-list, but scale down/up all of the numbers so that the total merit that they send is equal to the actual sMerit that they own. It might or might not be useful to do this via some sliding time frame scheme so that merit transaction amounts aren't just continually diminished over time as they increase in quantity.
 - Apply the above step recursively, creating a web-of-trust-style merit network

Then every user has a subjective merit score for each post (sort of like the bitcointalk.org trust system, which was inspired by FMS). And if you wish, you can assign people to be merit sources from your perspective by sending them large amounts of merit directly; these might or might not appear in the merit-transactions-list which you publish.

The more I read about it, the more I want to see it in action

Now, given that the name of this site is bitcointalk.org, do you really think that Theymos will move the entire forum on to the blockchain of some shitcoin?

The point of being decentralized is of course that theymos doesn't have to do it I can imagine using the Torrent protocol to share all posts, and using Bitcoin blocks as timestamps. But don't ask me how to build it

A decentralized forum doesn't make much sense. To have freedom of speech we need human moderation; I know no way a computer can be programmed to do that.

See theymos's posts above: every user can moderate everything from their own perspective, and every user can choose who's moderation they accept.

A truly decentralized forum cannot exist, because there is no way to punish those who'll infringe other people's freedom of speech.

They'll basically be shadow banned from the perspective of other users.

[Welsh]

A truly decentralized forum cannot exist, because there is no way to punish those who'll infringe other people's freedom of speech.

They'll basically be shadow banned from the perspective of other users.

Which you'd basically have a voting system for users to vote on users that should be banned, which sounds absolutely brilliant in practice. However, when it comes to reality, it'll be a shower of shit (for a lack of a better term). Considering how many disagreements users generally have on a forum, mix that in when there's financial gain to be made, you'd have a disaster in the making. You'd not only have disagreements which would result in feuds in terms of moderation, but also you'd have users looking to censor certain users to benefit from it financially. I hate that we always seem to talk about signature campaigns, but they're usually the best example.

For example, if you had someone that is in a high paying campaign, they could be subject to report abuse by using alt accounts. At least, with a centralised figure you have some sort of quality control. That also eliminates alt accounts abusing the moderation system.

Basically, for moderation to work on a decentralised forum with as big of a community as we have, it would have to be implemented absolutely perfectly, and even then most users don't want to contribute to moderation, and just want to browse, and discuss, which is absolutely fine. However, when it comes to relying on the community to moderate it'll likely be misused more than it's used for legitimate reasons.

It's like the saying; "absolute power corrupts absolutely" - John Dalberg-Acton, now that isn't always the case for everyone. However, when everyone has access to power, which could be artificially increased via alt accounts, you'll have a problem on your hands which wouldn't be easy to rectify.

It would be interesting to see it implemented just to see what sort of carnage pursue, but in terms of usability in the long term, it probably wouldn't work. At least, for a community of this size, and diversity. Then obviously, you mix in the financial side of things which always tends to cloud users judgement in the first place.

On a side note; I'd like to see shadow bans be introduced to the forum anyway. There's some situations where I think it could be beneficial. Specifically new users, but then that does increase the workload on moderators as posts would need to be reviewed in a pretty quick time scale in my opinion for it not to become too detrimental to new users.

[LoyceV]

Which you'd basically have a voting system for users to vote on users that should be banned, which sounds absolutely brilliant in practice. However, when it comes to reality, it'll be a shower of shit (for a lack of a better term). Considering how many disagreements users generally have on a forum, mix that in when there's financial gain to be made, you'd have a disaster in the making.

That's probably why it never really took off. If the users of this forum would switch to a decentralized forum, I thought to include most of the current Moderators as Moderators. I would assume most of the Moderators would include the users who made good Reports, and it would be pretty good from there.

For example, if you had someone that is in a high paying campaign, they could be subject to report abuse by using alt accounts. At least, with a centralised figure you have some sort of quality control. That also eliminates alt accounts abusing the moderation system.

The alt accounts won't have a say in this, just like non-DT users are quite meaningless to the Trust system now.

However, when it comes to relying on the community to moderate it'll likely be misused more than it's used for legitimate reasons.

Maybe. Or a user could choose not to follow the moderation coming from abusers and still see the posts.

Now that I think about it: if different users all see different posts in a thread because they have different moderation-
settings, the thread can become messy. I'm also unsure how new users would have to decide who's moderation to follow.
Someone should build it to answer all those questions

[LoyceV]

But i expect older post on bounty section cannot be obtained easily. It could be because node which store it currently offline or the software unable to found out which node has the data.

Even better: if nobody thinks the posts are worth keeping, it gets forgotten. That means the decentralized software should be able to handle parts of the data being offline, but also be able to handle it in case it comes back online (years) later.
Kinda like a pruned node

If all software by default use same moderation configuration, i expect it'll be less messy.

Who decides the default moderators in a decentralized system?

[Welsh]

Even if you took the current moderators, and used them to sort of define a moderation standard, you'd still be introducing centralisation there, since users would have to opt in on who they trust to moderate, the majority would pick who's the most popular out of them, retrospective to their actual moderation work.

Basically, for it to work you'd need a ton of statistics, a transparent way of showing what was moderated, and why, and then you can't bring in the current moderators as moderators for the decentralised forum, since that's introducing centralisation to a decentralised system, even if you get to choose who to follow for moderation.

Also, I think selecting multiple moderators to follow, and then seeing or not seeing content based on that would be confusing for everyone. Most new users struggle with the layout, and functionality of Simple Machines forums since they aren't used to it, let alone a system which would be confusing even for the most senior of users of that system.

Don't get me wrong, I'd like to see a limited trial (albeit I wouldn't want to participate, just watch) run just for the carnage as I previously stated, but realistically I don't think it would work very well. I guess for a community that talks about how much they love cars, it works better. However, Bitcointalk in particular is quite a large, and complex set of users, that basically attracts people from all walks of life. That in itself adds complexity.

[actmyname]

I guess for a community that talks about how much they love cars, it works better.

Have you seen the bloodthirsty animals that plague the forum? Ever hear the mantra, "Jaywalkers deserve death," echoed by much of the anti-zebra coalition?Unlike the analog universe, where you can simply choose to exit a conversation with a dedicated idiot or shill, would you not have to choose between sorting through enormous bloat (without fees) or extreme new/unknown-user segregation? A concession for newbies is multiplied by several magnitudes for malicious spammers. The value-based filters will target both bloat and unrecognized users: if you want to see or merit new-user activity, you will need to tackle an ungodly amount of spam, effectively becoming a voluntary moderator (but only for what sections you deem important).

It's an interesting premise but isn't the capacity of spam too high?