I'm just seeing a ..."Your message has been saved, and will be posted after being **approved** by a moderator". This doesn't make sense.

1. what criteria decides if my comment deserves approval or not?
2. Is there a full-time moderator available to ensure there no bureaucracy/delay?
3. Is it not a much more better approach to delete when its reported as inappropriate by the public? (White-listing vs Blacklisting,...which one permits a democratic and free speech environment?)

On Saturday, 22 September 2018 at 14:58:58 UTC, aberba wrote:
> I'm just seeing a ..."Your message has been saved, and will be posted after being **approved** by a moderator". This doesn't make sense.
>...

This happens only for new topic?

S.G.

On Saturday, 22 September 2018 at 16:45:15 UTC, SashaGreat wrote:
> On Saturday, 22 September 2018 at 14:58:58 UTC, aberba wrote:
>> I'm just seeing a ..."Your message has been saved, and will be posted after being **approved** by a moderator". This doesn't make sense.
>>...
>
> This happens only for new topic?
>
> S.G.

I'll not create a topic to check this behavior, but this message doesn't show up when replying inside a topic.

PS: By the way the CAPTCHA is awful, look what they throw to us:

int v()
{
  return 26 % 3
    ? 13 / 3
    : 42 % 5;
}

I mean

S.G.

On Saturday, 22 September 2018 at 14:58:58 UTC, aberba wrote:
> I'm just seeing a ..."Your message has been saved, and will be posted after being **approved** by a moderator". This doesn't make sense.

Your post was flagged by the spam filter. It was a false positive, which sometimes occurs with very short posts, as was yours. It was approved a few minutes after it was submitted.

> 1. what criteria decides if my comment deserves approval or not?

Your post is not spam, an attack on other forum members, or egregiously inflammatory / off-topic.

> 2. Is there a full-time moderator available to ensure there no bureaucracy/delay?

There are several persons who receive moderation notices and can act on them.

I generally receive email throughout the day, so you could at least count me as a full-time moderator.

> 3. Is it not a much more better approach to delete when its reported as inappropriate by the public?

No. We did this until this year. This resulted in:

- Lots of spam (despite the spam filter and CAPTCHA). Web forums attract a LOT of spambots (and humans paid to post spam).

- Spam in mailing list users' inboxes, even after a moderator deleted it off the forum, since you can't unsend an email.

The new method catches a lot of spam that would otherwise get through. You don't see it, but the moderators do.

On Saturday, 22 September 2018 at 16:48:35 UTC, SashaGreat wrote:
> PS: By the way the CAPTCHA is awful, look what they throw to us:

If you have a better idea of a CAPTCHA that would be easy for D programmers but hard for spammers, please submit a pull request:

https://github.com/CyberShadow/dcaptcha

On Saturday, 22 September 2018 at 17:13:18 UTC, Vladimir Panteleev wrote:
> On Saturday, 22 September 2018 at 16:48:35 UTC, SashaGreat wrote:
>> PS: By the way the CAPTCHA is awful, look what they throw to us:
>
> If you have a better idea of a CAPTCHA that would be easy for D programmers but hard for spammers, please submit a pull request:
>
> https://github.com/CyberShadow/dcaptcha

First I didn't want to sound harsh, and by the way I sent the message without complete it.

I did by head. But how a newbie would suppose to do that?

You may say to open the compiler and try it or go with the online version, but it isn't too much?

And by the way, after you do once why need to do every time?

S.G.

On Saturday, 22 September 2018 at 17:19:41 UTC, SashaGreat wrote:
> I did by head. But how a newbie would suppose to do that?

For that challenge, you only non-obvious thing need to know is the syntax for the modulus and ternary operators, which are present in many programming languages. You can do tthe calculation with a regular desktop calculator.

If that is too much, you can run the code on run.dlang.io. In this case, you only need to know how tocall a function and print its result.

If that is also too much, you can ask for help in the #d IRC channel.

The software suggests the above two options. In my opinion, it is a reasonable compromise, but I'm open to suggestions. (Note that at least once, a spammer managed to get through the CAPTCHA precisely by simply asking on #d, with a good samaritan providing the answer without inquiring further.)

> And by the way, after you do once why need to do every time?

It is needed to prevent flooding. However, successfully solving the CAPTCHA a number of times across a period of time while logged in will whitelist your account.

On Saturday, 22 September 2018 at 18:56:28 UTC, Vladimir Panteleev wrote:
> On Saturday, 22 September 2018 at 17:19:41 UTC, SashaGreat wrote:
>> I did by head. But how a newbie would suppose to do that?
>
> For that challenge, you only non-obvious thing need to know is the syntax for the modulus and ternary operators, which are present in many programming languages. You can do tthe calculation with a regular desktop calculator.
>
> If that is too much, you can run the code on run.dlang.io. In this case, you only need to know how tocall a function and print its result.
>
> If that is also too much, you can ask for help in the #d IRC channel.
>
> The software suggests the above two options. In my opinion, it is a reasonable compromise, but I'm open to suggestions. (Note that at least once, a spammer managed to get through the CAPTCHA precisely by simply asking on #d, with a good samaritan providing the answer without inquiring further.)
>
>> And by the way, after you do once why need to do every time?
>
> It is needed to prevent flooding. However, successfully solving the CAPTCHA a number of times across a period of time while logged in will whitelist your account.

But what is there to stop a spammer from doing the same?

I mean it's fairly easy to grab the captcha code and run it through a D compiler and then post the result automatically.

It's no rocket science, so it really doesn't do much in preventing I think.

Really it can be automated like:

1. Copy the code
2. Go to run.dlang.io
3. Paste the code
4. Compile it
5. Wait for the output
6. Copy the output
7. Paste the output into the input field
8. Submit

It would take anyone familiar with basic http macros less than 10 minutes to automate that process, even less using a programming language if it's mass automation.

The forums for D might just not be popular enough for any "bots" to bother I guess?

I would suggest some real captcha software that are used by the majority of sites.

And on top of that maybe a flag system.

People being able to flag posts and if a specific post is flagged by enough people then it'll be "hidden" until moderation takes action by either making it "visible" again due to invalid flagging or deleting it because it was a valid flag.

This can help not only against spammers, bots etc. but also when there are trolls making troll posts etc.

On Saturday, 22 September 2018 at 19:09:24 UTC, bauss wrote:
> But what is there to stop a spammer from doing the same?

Spammers are not going to exert that much effort in order to be able to spam 1 website, so that the moderators then change their algorithm and block them again.

This is the key. Spammers win by targeting classes of websites running the same engine, or using generic bots that detect arbitrary forms, or employing humans to do it for them. To defeat them, you must pose a challenge specific to your website, so that people who are not interested in your website's topic will have difficulty solving, but not the other way around.

> I would suggest some real captcha software that are used by the majority of sites.

No, those work MUCH worse than the above.

On Saturday, 22 September 2018 at 19:09:24 UTC, bauss wrote:
> And on top of that maybe a flag system.

This exists, but is only visible to certain users.