I hope I don't get accused of super bumping multiple threads tonight (really not so old threads or well not so old last post!)

On first impression, nodeBB was the ticket - without any of the technicals. It's all that matters.

It truly is all that matters. Since the majority of fourm users are not going to be technical and forum have there work cut out to drag people away from the twitters and face books of the world.

A forum can be deployed to any user base so it's application and audience is potentially as wide as you like and varied as you like, that you have to think about simple but not dumb Ui/UX and rely on the beauty of tradition, i.e. established and solid universal conventions that are universal but not afraid to push the envelope to enhance convention to the point it's at least feels innovate if not actually is.

Speed.

It's silky response. The fade out fade in transition was a genius choice, let me type that again GENIUS!

The Composer.

• It looks and is weighed better i.e. it's styled better, simple.
• It's not perfect, but it's one of the best out there.

The composer is the window to all content creation, the crucible of an entire forums everyday lifeforce. It's got to be good, it's got to always be trying to be better, by trying to be great. This is the interface that makes or breaks the user experience and thus a forum.

My UI/UX 2c:

Please consider making the composers upper top line a drag focus draggable area so it is not restricted to the black circle aligned to the left, that should really be centered (it took me too long to figure out) - it breaks a basic UI/UX expectation burned into our retinas at this point!

Updated a nodeBB to 1.13, all seems relatively fine, though, I was not paying attention and I updated to v1.13.4-3. - that's ok for now, seems to work.

However the ns-embd plugin is not embedding tweets, just leaving a blank blockquote. Youtube videos embed perfectly fine, so I suspect the watch + replace needs tweaking?

Watch

(?:<a.*?)?(?:https?:\/\/)?(?:twitter\.com)\/([^\/\"\s]*)\/status\/([^\/\"\s]*)(\/photo\/\d|)(?:.*?\/a>)?

Replace

<blockquote class="twitter-tweet"><a href="https://twitter.com/$1/status/$2?ref_src=twsrc^tfw"></a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>

This is a feature I wouldn't expect in forum software, since traditionally such systems like phpBB don't have it out of the box. It also seems wrong on a base trust level even before we got to the modern era of GDPR etc. etc.

Yet Discourse has this out of the box much to my own surprise and I thought but why would you need it?

Other than tracking abuseive PM's between waring users, which could easily be stopped by the recipient being able to ban PM's form a user. I didn't initially see the valid use case.

The point was made by Atwood, Discourse creator, that for example certain types of nefarious rings might use your PM system to share illicit material undetected. Think the worst kind.

• That is a very valid point and one worth exploring.

However it's still strange being able to read anyones private messages and that users also may simply assume that their messages are private. Few think about databases and that once a person has access to a database they can in theory access any of the info but it's not frictionless access. Discourses implementation is as simple as reading your own PM.

What I felt would be a better solution, more admin friendly and simply effective at addressing the problem, would be to have a system that monitors the PM activity and volume of all user accounts.

For e.g. - A threshold is set with some criteria - then imagine a 0 post or very low post count user has more than 2 pm's a day it's reported and flagged in stats. Imagine another user has 100's of PM's per day or hour, it's flagged with more force, because such activity would be highly suspicious. A whitelist system would handle the good guys and magybe not flag (like mods) but still report their activity too, if required.

So of my 10,000 imaginary forum users how do you find the 4 or 5 who are abusing the system, with little effort but with pinpoint accuracy fast?

Second to that there could also be threshold limits set that users can only read and send x pm's per day until they clock up some rep, which would act as a massive disincentive to abuse the PM system for nefarious and illegal activities. (apologies if such a feature is already there, I'm not 100% familiar yet with all of nodeBB functionality and feature set but getting there).

Open access for admins to PM's does not mean you will easily find and therefore stop the problem. Access only helps you determine if it is a problem on a PM per PM basis, how do you even know which user acc PM's to start looking at - it is still needle in a haystack.

What you need are PM analytics as an admin in the first instance! - therefore it's not an access issue!

I find the Discourse implementation a good example of solving a real niche problem but with the wrong solution applied or at least the 50% wrong.

I would really like to see a PM report/abalytic system in nodeBB natively, as a core security feature to tackle this issue - that can auto lock crazy PM behaviour.

Warning + notices to users easily administered by default, that PM's are not publicly viewable but admin viewable if x. y and z occurs.

The crazy thing is Discourse team originally implemented this and gave moderators access.

That's just insane IMHO, even as an Admin I am not really comfortable with that once click power - it's clearly open to I would say more abuse that what it tries to solve.

@julian Cool thanks, yea poor ns-embed is not happy in 1.12.2 - I and others already flagged in that thread, still, A Great plugin!

Ok SMTP/EMAIL is fixed!

Some deeper DNS/MX work seems to have made the difference and using "encrypted" setting - initially I was using StartTLS as per email provider SMTP guide, which specified TLS - tested it again and StartTLS broke it.

Encrypted works.

This is a feature in Discourse, and when i first saw it years ago I thought, COOL!

Many years later, when actually interacting with the feature in Discourse. I realised it was generally a bad system because it was really a bad idea.

What problem was it actually solving?

Users not being able to moderator other users comments????

That is Chaos!

Why invite Chaos?

The problem you see it as is lack of staff to police the conversation. This is a problem for all busy forums. Overall auto features are too crude to deal with rising noise IMHO.

One very obvious downside of "social justice" which these days is IMHO another term or mask for mob rule, users quickly learn the weakness of the system, and actively suppress other users in a targeted fashion, so what might seem to solve the moderation burden, could end up being weaponised against ordinary users and cause serious community problems, the kind you may not recover from, as recriminations and all kinds of things could explode form a few bad turns.

The most important thing for a user first off, is to be able to report or flag a post easily, that see as problematic for whatever reason and then for that to appear quickly and very visibly to mods and admins alike.

At least a Troll is very easy to see, users who are abusing a system like this, how visible are they? You need to back up the user accounting side too. This is starting to seem like a headache.

Traditional moderation with a good team is the solution, that is sustained and enabled by a great admin/backend that make moderation of problem users fast and efficient. The better this system is the less moderation users you might need to assist in day to day moderation.

Get that right and you are winning.

Naturally moderation teams have there problem and this is often due to progress of time in my experience, so to add a crude feature like
mob down voting that kills post, does nothing to address team work and good admin policy, vision and direction.

A good report/flag system with SPAM and some other options where a user report may temporarily hide a post, is a good start - SPAM might be hidden automatically, (yes that could be abused) but since it is granular more likely it will be used in good faith most of the time (but I have seen people go absolutely nuts with report features too which is just silly), options for off-topic or disruptive might good (it might be goo to allow an admin set extra report flags and also control the action after the report, doe sit hide the post or not. Things like that, which aids the internal operation more than handing over executive powers to the mob!

Setting a downvote threshold to remove a post is basically giving your users based an achievable target to counter disruptive use or exact revenge! Trolls might also use this feature to counter also and then you have all out war.

Leave all the power in the hands of Admins, Mods and give them good control to track, control and censure problematic posts and users, that is hight visible quickly, fast and able to batch operate in comfort and then users will understand that the odd troll might get through or a bit of spam may appear but since you do such a great job of keeping the house clean, they really appreciate your efforts.

No point in making yourself obsolete now is there?

This one ain't too bad either, you can create palettes and is very hands on too

+you can apply various type of colour blindness filters to the palette to simulate how it might be perceived.

To abide by my own call to keep things simple.

I would suggest to base such a feature on report/flag action. rather than the downvote.

Downvotes are visble to users.

Reports/flags are not as a rule - this should be the trigger.

This goes some way to addressing the weakness of user mob behaviour to game the vote system to knock people out.

Yes!

As outlined in the second part to this post here - I have created this dedicated topic for the purposes of pitching this new feature and functionality for composer.

A composer that can align in one click to the right of the screen - creating a split view.

I guess if people want left they can have it but the concept is being able to easily reference posts on the right and compose on the left, keeping an unobstructed view of the entire flow of the topic posts without having to minimize the compose editor.

Meet the new contextual button that would toggle between positions.

Click to toggle horizontal position:
Click to toggle right-aligned position (split view):

As an evolution of composer core feature & function, I believe the case for, is incredibly self evident.

• I'm waiting for someone to tell me this feature is already available...
• I will edit an addition this OP with summary points and extra info if required.

Please Upvote - If you want to see this as a core feature in standard nodeBB deployment.

Ok if your moderation people are along a few hours later or no longer that 12 hours. Then using the report/flag function is perfect.

Not everyone who votes might be a good reporter and vice versa, as the two things are cognitively different actions IMHO.

I think someone who uses a report/flag feature, is able (*or we hope, but not always) to separate their personal interaction with posts versus helping keep the neighbour hood clean - maybe the way to see it is like this:

(voting) personal action vs (reporting) operational action.

Reporting/flag is a very clear functional action, because it implies that the report will be seen directly, and mod actions are expected to follow soon after or more swiftly.

Adding in extra purpose or breaking out to the voting system, into a secondary or sub role that it was not designed for, is I think, as outlined previously fraught with other issues.

Using reporting/flagging system, the Admin might benefit from be easier reports generated from reports/flags, and can plot over time... using the voting system to extra a report like this I would imagine is more complex or at least needing a lot filtering and re-stating to be useable.

User with most flags
Flags by day, month week... etc. etc.

In the admin there is already "Most Flagged" in the user, so let's go with what is already baked in and expand it out for best performance and visibility, and possibly allow combination of flags + user states to filtering user lists.

Then also if Admin can tailor flags auto-actions based on criteria, that would be sweet:

Flag action (post): HIDES POST - when user x 3> (w/100> posts) flag post - post is hidden, pending further approval/attention + user is frozen from all activity other than reading, until unfrozen, banned or whatever by mod/admin.

It would be useful say for a mod who has to freeze multiple user to be able to drill down a list of FROZEN users, to then investigate further for whatever reason or group unfreeze or unfreeze some and not the others.

User accounts could do with extra User States might be an idea with default but customisable privileges for admins to tailor, for example as above "Frozen".

I've never really seen or understood the true value of a shout box but I'm open to other views - to me it seems to be an ambiguous duplication of forum function and often as an interface element in a forum, it always feels squeezed in.

That's my experience on other platforms.

Obviously, if nodeBB has a shout box it will be the best in the world!

@gotwf said in automatic temporary ban of a troll user with minus X votes:

First off: Lot of good discourse in this thread. Lame pun intended.

@omega said in automatic temporary ban of a troll user with minus X votes:

Should reputation be broken out to include more metrics, that can be used for greater granularity of other admin controls. Probably.

This would allow user contribute up into such

I guess we can think about moderation in two ways.

Passive Mod Role:

Facility to allow users, access to features based on a criteria matrix to support active moderation actions.

Active Mods Role:

Traditional power users who have been granted greater privileges by admins for direct control over user activity and forum content.

In summary, what you need is to enhance the right mix of passive controls, functions and features that will support the active moderation.

More work and thought on this would benefit every nodeBB install out there.

Disagree w/above though. Too much added complexity. KISS.

Tough crowd, always with the tomatoes... try this one:

Let the Mods be Mods, let the Users be Users and Admins be King & Queen, forum developers could be Heros!

Granted it's more BOWIE than KISS but you get the point right?

Oh and can we have a better grammar-typo bots for my posts...

Ha, I was not entirely sure if reputation was calculated as net total of votes, it appears it is, I found this video from 2016:

NodeBB reputation basics

Should reputation be broken out to include more metrics, that can be used for greater granularity of other admin controls. Probably.

This would allow user contribute up into such

I guess we can think about moderation in two ways.

Passive Mod Role:

Facility to allow users, access to features based on a criteria matrix to support active moderation actions.

Active Mods Role:

Traditional power users who have been granted greater privileges by admins for direct control over user activity and forum content.

In summary, what you need is to enhance the right mix of passive controls, functions and features that will support the active moderation.

More work and thought on this would benefit every nodeBB install out there.

I think someone who uses a report/flag feature, is able (*or we hope, but not always) to separate their personal interaction with posts versus helping keep the neighbour hood clean - maybe the way to see it is like this:

Since I wrote the above let me be the first to say I am being a bit dumb or naive here.

The report function can absolutely be used in an angry and abusive way, but only admin/mods will see it - sometimes it is a users whose head has gotten too hot in the heat of battle or is disgruntled contributor feeling a lack of attention to previous flags or other complaints is not being taken seriously, can go on a reporting button rampage - they might hunt down multiple posts from their target user (or posts in a target thread they do not like) and go ballistic reporting tens and tens of posts until they lose interest or cool down.

Some kind of limit/gate for flags need to kick in at that point, but again because it is a flag feature, it is easier to gate the flagging than abusive downvoting and as stated, only backroom ops will see it, also the flags are attached to the the user doing the flagging marking them out - so you can detect the nutburgers who abuse any report or flag system and have auto feature kick in and gate abuse here also. This is where some circuit breaker rules might help.

Perhaps a simple daily limit per flag per user might be enough to handle it, leave it to the admin to decide what is best for their acitivyt levels.

Also allowing only users with a higher contribution metrics to the forum by way of post + time also helps reduce abuse of flag system to minimum.

A new user should not be able to flag off the bat, or use a flag that will have a follow on action like hiding post + freeze user.

That is too much power for a newly registered account.

The thing I like most here is using reporting overall gives a far great ability to create and extend a more accurate and insightful audit trail of user flag/reporting activities.

As I type I also think there might be a good case for a moderation panels or moderation dashboard to aid efficient moderation in busy forums.

Ok if your moderation people are along a few hours later or no longer that 12 hours. Then using the report/flag function is perfect.

Not everyone who votes might be a good reporter and vice versa, as the two things are cognitively different actions IMHO.

I think someone who uses a report/flag feature, is able (*or we hope, but not always) to separate their personal interaction with posts versus helping keep the neighbour hood clean - maybe the way to see it is like this:

(voting) personal action vs (reporting) operational action.

Reporting/flag is a very clear functional action, because it implies that the report will be seen directly, and mod actions are expected to follow soon after or more swiftly.

Adding in extra purpose or breaking out to the voting system, into a secondary or sub role that it was not designed for, is I think, as outlined previously fraught with other issues.

Using reporting/flagging system, the Admin might benefit from be easier reports generated from reports/flags, and can plot over time... using the voting system to extra a report like this I would imagine is more complex or at least needing a lot filtering and re-stating to be useable.

User with most flags
Flags by day, month week... etc. etc.

In the admin there is already "Most Flagged" in the user, so let's go with what is already baked in and expand it out for best performance and visibility, and possibly allow combination of flags + user states to filtering user lists.

Then also if Admin can tailor flags auto-actions based on criteria, that would be sweet:

Flag action (post): HIDES POST - when user x 3> (w/100> posts) flag post - post is hidden, pending further approval/attention + user is frozen from all activity other than reading, until unfrozen, banned or whatever by mod/admin.

It would be useful say for a mod who has to freeze multiple user to be able to drill down a list of FROZEN users, to then investigate further for whatever reason or group unfreeze or unfreeze some and not the others.

User accounts could do with extra User States might be an idea with default but customisable privileges for admins to tailor, for example as above "Frozen".

To abide by my own call to keep things simple.

I would suggest to base such a feature on report/flag action. rather than the downvote.

Downvotes are visble to users.

Reports/flags are not as a rule - this should be the trigger.

This goes some way to addressing the weakness of user mob behaviour to game the vote system to knock people out.

I was thinking and writing in terms of discourse from experience. This is kind of what happens in discourse too.

I like to keep things simple.

Do you want the system to provide technical backup to users to perform moderation duties through their mass Actions?

Personally. I do not like this from an admins moderation point of view.

However I totally see the reasoning and need for support. It is a compelling case. It comes down to trust. There is a trust system in discourse too with user levels. I do not particularly like it but it could be done in a fashion that it does help to reduce moderation burden and overhead on a busy growing forum.

Maybe think of it more like village militia.

I preferred earned posting access through tiers.

Tiered moderator powers would also be good too or users after for eg 100 contributions get such an ability. Not full mod but partial mid abilities like this.

It’s about bringing people into the group so they can be part of the maintenance. There is a lot of this thinking in Discourse but I don’t fully like it in real use. It’s about implementation and the root behind the feature.

What is to stop a bot squad Or user working in hit squads on a forum from disable targets?

Lack of market for such a product. Now maybe but later on? Who knows.

All features depend on the personality and culture of a forum and tbh this can also change over time.

I guess we are starting to get busier on your forum eh!

Ok this would be a useful feature when you need to calm and maybe do some moderation clean-up. It is better than locking a thread, because it neutralising the problem user, while causing the least amount of disruption to the overall forum.

Next is locking a thread, which is often required if you have a mega multi-user spat. Unless it is easy to group mute, then this is essential.

I would favour a FREEZE topic, that is more attuned to dealing with eruptions of chaos, moderation is happening and is a temp process, the LOCK TOPIC is very definitive and possibly Admin only option - leaving Freeze to moderators and higher - but maybe I am splitting hairs here but I think it has some merit. Something to expand on later.

I see some users also want "ignore" features.

I am not sure an ignore features is a great idea tbh. It goes against the spirit of an open forum and also is again a crude feature to try and address what moderation shortcomings exist.

Basically it is a feature that is impolite by by stealth.

+1 for a Mute feature.

This is a feature in Discourse, and when i first saw it years ago I thought, COOL!

Many years later, when actually interacting with the feature in Discourse. I realised it was generally a bad system because it was really a bad idea.

What problem was it actually solving?

Users not being able to moderator other users comments????

That is Chaos!

Why invite Chaos?

The problem you see it as is lack of staff to police the conversation. This is a problem for all busy forums. Overall auto features are too crude to deal with rising noise IMHO.

One very obvious downside of "social justice" which these days is IMHO another term or mask for mob rule, users quickly learn the weakness of the system, and actively suppress other users in a targeted fashion, so what might seem to solve the moderation burden, could end up being weaponised against ordinary users and cause serious community problems, the kind you may not recover from, as recriminations and all kinds of things could explode form a few bad turns.

The most important thing for a user first off, is to be able to report or flag a post easily, that see as problematic for whatever reason and then for that to appear quickly and very visibly to mods and admins alike.

At least a Troll is very easy to see, users who are abusing a system like this, how visible are they? You need to back up the user accounting side too. This is starting to seem like a headache.

Traditional moderation with a good team is the solution, that is sustained and enabled by a great admin/backend that make moderation of problem users fast and efficient. The better this system is the less moderation users you might need to assist in day to day moderation.

Get that right and you are winning.

Naturally moderation teams have there problem and this is often due to progress of time in my experience, so to add a crude feature like
mob down voting that kills post, does nothing to address team work and good admin policy, vision and direction.

A good report/flag system with SPAM and some other options where a user report may temporarily hide a post, is a good start - SPAM might be hidden automatically, (yes that could be abused) but since it is granular more likely it will be used in good faith most of the time (but I have seen people go absolutely nuts with report features too which is just silly), options for off-topic or disruptive might good (it might be goo to allow an admin set extra report flags and also control the action after the report, doe sit hide the post or not. Things like that, which aids the internal operation more than handing over executive powers to the mob!

Setting a downvote threshold to remove a post is basically giving your users based an achievable target to counter disruptive use or exact revenge! Trolls might also use this feature to counter also and then you have all out war.

Leave all the power in the hands of Admins, Mods and give them good control to track, control and censure problematic posts and users, that is hight visible quickly, fast and able to batch operate in comfort and then users will understand that the odd troll might get through or a bit of spam may appear but since you do such a great job of keeping the house clean, they really appreciate your efforts.

No point in making yourself obsolete now is there?

@crazycells Ok cool tx!

Updated a nodeBB to 1.13, all seems relatively fine, though, I was not paying attention and I updated to v1.13.4-3. - that's ok for now, seems to work.

However the ns-embd plugin is not embedding tweets, just leaving a blank blockquote. Youtube videos embed perfectly fine, so I suspect the watch + replace needs tweaking?

Watch

(?:<a.*?)?(?:https?:\/\/)?(?:twitter\.com)\/([^\/\"\s]*)\/status\/([^\/\"\s]*)(\/photo\/\d|)(?:.*?\/a>)?

Replace

<blockquote class="twitter-tweet"><a href="https://twitter.com/$1/status/$2?ref_src=twsrc^tfw"></a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>