Re^6: Dealing with An(?:no|on)ytroll

[tirwhan]

First you say that posting with content like...are clearly trolls

No I didn't, reread my post. I said that post (emphasis added) 1 is clearly a troll but I couldn't tell with certainty whether 2 or 3 are without further information. I was talking about three different posts (and said so).

Really? If I'm a company, I know who each AM is?

If you're a company with even a hint of a clue about network security you can easily find out whether a given AM is an employee of yours because you control your network gateway and can read the posts (if you're so inclined). Anyone who believes that their employer cannot find out a given AM post originates with them either trusts their employer not to care/check (in which case, why should they be worried about an IP address display) or is a fool.

Look, we're going round in circles here and I'm sorry but I have a feeling you're either trying hard to misunderstand me or our perceptions of reality do not overlap sufficiently to make understanding possible. Please give me a valid case scenario where both of the following are true:

1. A user does not want either their real-world or virtual identity attached to a post they send to PM
2. The user cannot create a user account and log in with that.

If there is such a scenario I'll agree that you're right and retract my proposal. If you believe the requirement of logging in to achieve anonymity is too harsh and 2. would pose too much of an inconvenience to people, I would like to agree to disagree at this point.

But your proposal is to list every ones IP address.

No, I'm only proposing to list AM IP addresses. Posts from logged-in users already have a unique identifier, their username, a second one is not necessary for any purpose I can see. This is where the car analogy breaks down, because cars only have one way of uniquely identifying them.

[Perl Mouse]

If you're a company with even a hint of a clue about network security you can easily find out whether a given AM is an employee of yours because you control your network gateway and can read the posts (if you're so inclined).

Most certainly. However, you can't go back in the past and find out who wrote some stuff you don't like. Which is a gigantic difference.

Please give me a valid case scenario where both of the following are true

That's not the point I'm making. Your arguments can as well be used to get rid of anonymous monks all together.

If there is such a scenario I'll agree that you're right and retract my proposal. If you believe the requirement of logging in to achieve anonymity is too harsh and 2. would pose too much of an inconvenience to people, I would like to agree to disagree at this point.

That's not my point at all. If convenience is an issue, here's an extremely convenient one: don't reply to trolls. Doesn't require any work on your side. Doesn't require giving up anonymity of anon monks. Doesn't require code changes.

But your proposal is to list every ones IP address.

No, I'm only proposing to list AM IP addresses.

Now I'm confused. I make an anology to single out all red cars (and not blue cars, yellow cars, black cars, white cars) and change them because a single red car is a nuisance, and you claim that changing all cars is a better anology.

So, what is it? All cars (that is, IP addresses listed for everyone), or just red cars (that is, IP addresses for AMs)?

Posts from logged-in users already have a unique identifier, their username, a second one is not necessary for any purpose I can see.

Well, posts have a unique identifier in the sense that each post only has one user name. The user name may be reused in a different post. However, there's nothing enforcing a person to use a single unique user name. Users can have as many user names as they want. And believe me, there are users with different user names.

Look, we're going round in circles here and I'm sorry but I have a feeling you're either trying hard to misunderstand me or our perceptions of reality do not overlap sufficiently to make understanding possible.

Indeed. You seem to think your solution is actually going to solve the problem. I claim that the price of that solution is very high (giving up anonymity - like giving out phone numbers) and that it's very doubtful that it will actually solve the problem.

[tirwhan]

you can't go back in the past and find out who wrote some stuff you don't like

Yes you can, if you collect netflows, use an http proxy or have only slightly paranoid firewall logging enabled. Just compare the timestamp of the node with your logs.

All your other points I've already replied to in other posts, please reread those carefully if you're really expecting an answer.