I have made a decision to alter and/or remove various restrictions on Voat. I’ve thought a lot about this and it’s something both @Atko and I believe needs to be reevaluated.
Voat has always had a problem with spam. @Amalek would spam posts and hijack the new queue making it unusable. MH101 and then later @SaneGoatiSwear would hijack comment pages making them unusable. The rules Voat uses were put in place in to combat this behavior. They are old rules, mostly remaining unchanged from the initial versions of this site. Most, if not all, of the rules were in direct response to spam attacks. It was never Voat’s intention to limit non-spam accounts, but this is what has happened as an indirect result of these rules.
Voat will not keep in place a system that permanently limits a segment of users from debating and conversing. This isn’t Free Speech as I see it or as I want it.
Voat will shortly be going live with a new code base, and I want to have a new system designed and ready for when this happens, so I am posting this announcement to get feedback from the community.
The main areas of concern:
- Commenting restrictions on negative CCP accounts that aren't spamming their comments
- Limiting any account that spam comments
TL;DR
We need to allow unpopular opinions while preventing comment spam.
How do we do it?
All options are on the table
https://voat.co/v/announcements/1330806
view the rest of the comments →
10246470? ago
:D
Spam is an issue and we don't want it overrunning the website. But at the same time you're right, these restrictions have been inhibiting people who have done nothing wrong but share too many unpopular opinions, and it isn't in the spirit of Voat.
We should consider what tools we have available. The /v/ReportSpammers community is very hard-working and dedicated to keeping Voat free of spam, and it is a community very capable of growing. Spam is against Voat's rules; accounts that spam get permanently banned from the website. We determine that accounts are spamming by responding to user reports against specific accounts, evaluating their comments / submissions, and then deciding if they have indeed spammed. If they have, you eventually ban them. I think that's the basic process.
Waiting for a spammer to accrue negative CCP is actually relatively slow. What we could do instead is this: if an account receives spam reports, and one of the trusted community members in /v/ReportSpammers marks the report as actual spam, then upon that marking the account could be restricted until such time as you or someone else is able to review the reports and ban the guilty users.
As far as I am aware this follows the same process as right now, except it will not restrict any account's commenting ability based on CCP, only on confirmed spam reports. As I understand it this should restrict guilty accounts much faster than negative CCP would have, without restricting non-spam accounts. All we require is a sufficiently large and trusted report marker section of the community, and then the awareness of the Voat community at large to place spam reports instead of downvotes in the first place.
The community at large can vote on who they want / trust to mark reports as actual spam, and we can keep those who have been doing a perfect job already (@Cynabuns namely. I'm sure @NeedleStack would do well also).
I can adjust anything I've written above for feasibility reasons but I think some interpretation of this will work for Voat well without punishing the innocent.
guinness2 ago
I like this idea, but the current CCP restrictions works well against bots... I'm not so sure your idea would survive malicious script kiddy attacks that create new accounts for immediate farming and spamming or flood /v/ReportSpammers with random reports just to keep the mods overworked and unable to keep up with legitimate reports.
10252237? ago
Now that is a compelling argument. At the very least the current system does not give bots a reason to generate spammy reports. Doing so after implementing my suggestion could effectively derail the system. Some CCP requirement would likely be necessary just to submit reports (just like we had the 100 CCP to downvote) in order to prevent this. Maybe a higher requirement level, 1000 CCP or something. Only one report is necessary, perhaps, to draw someone's attention to the spam account in question, so fewer community members are required to have the ability to report. And anyone with less than the threshold could still contact someone above the threshold and ask them to file a report, whereas that was not possible or meaningful with downvotes.
There is certainly a great deal to consider.
guinness2 ago
When trying to solve any problem, I think it's always worth asking "Have I solve this problem or just moved it?".
But without the existing voting and posting CCP restrictions, farming with bots and new accounts would be even easier and the extra CCP constrains on the submit reports function could be quickly overcome.
The problem of CCP restrictions on voting and posting hasn't been solved, it's just been moved to the Report button.
'It could be really nasty!
10252365? ago
Vote manipulation is still vote manipulation, and Putt can detect it easily. If so the guilty accounts would be banned.
The thing is, by only assigning comment restrictions to people who have been successfully reported for spam, only actual spammers will ever face the restrictions, which was the intention from the beginning. In this case "moving the problem" has some justification -- because it is the location that is the problem.
guinness2 ago
I disagree because in both cases you are using human intervention to fight bots... and that's not a winnable strategy.
The malicious purpose of bots will be to create so much work for Putt and the Report Spammer mods that they can't keep up with deleting farming alts and dismissing fake reports... while spamming or brigading is taking place.
Currently, a malicious user is required to manually create and tend to a set of alts in preparation for a single brigading or spamming event, which takes human intervention and time and effort... whereas bots without posting / comment / voting CCP restrictions can tend to themselves at lightning speeds to rapidly gain enough CCP to begin fake Report attacks.
10252546? ago
This is the fundamental issue with my suggestions, I agree.
Could not CCP restrictions similar to those in place for downvoting stymie this enough, though? I know others have suggested ways of limiting the formation of reports -- a CCP requirement plus ten different reports on one comment for a report to actually be generated, in combination could probably keep bots away, somehow, maybe?
guinness2 ago
If a new alt can up-vote comments, then no... because then bots have control over farming each other's CCP... and they wouldn't need to be blatant about it either because each alt would only need to up-vote another once and then it's served it's purpose. I wonder if Putt's farming / brigading logic could cope with that: a bot that creates a disposable alt that only votes once?
No, alts are disposable and creating one that posts one fake report is as easy as creating ten that each post one fake report.
All solutions would need to place up-vote and down-vote restrictions based on CCP and then you've recreated the existing solution, but with extra steps.
Think of how the poor Voat database would feel after such abuse!
10252638? ago
I think if they are either coming from the same IP or the same machine he gets alerted by that, in which case he would be able to ban those accounts before they started flooding RS with reports.
It's not the 100 CCP restrictions that currently exist for downvoting that we are considering changing, though. Those work and would remain. It is the negative CCP restrictions that would be moved to have a different cause, potentially. If accounts are required to have a certain amount of CCP to submit reports, then it seems we would have as much issue with bot reports as we have with bot brigaders now -- in short, not too much of an issue, less so if we up the requirement of CCP.
guinness2 ago
I wonder how that works with anonymous proxies?
Fair enough... but I suspect the lack of bot attacks thus-far is no guarantee the future won't be dominated by bot attacks, especially when the other vulnerabilities are mitigated.