1: Isn't this a crude invasion against my privacy?

If you feel that we're installing microphones to your living room, then yes. Are we?

• IRC is open by its nature, whoever can join the discussions and get kicked.
• IRC discussions travel around the globe, literally via interconnected IRC servers. The traffic is not crypted (unless everyone uses SSL) and it can be sniffed easily.
• Logging is trivial in any IRC client and widely used.

The point is that even without explicit logging efforts like Irchiver, it's possible and even probable that your sayings end up in someone's logs. We see that it's more fair to make it explicit than do it behind the scenes.

It's good to see that open efforts like Irchiver raise discussion about IRC privacy.

Note that currently we're developing theory and code for our search engine. All the logs will be used only for this purpose and they will be handled CONFIDENTIALLY. It won't be shown in public nor distributed.

2: Isn't this a bit big brotherie?

We let you be your own big brother. We believe that services like this operated by open communities can make their work even more efficient and fun. We are not interested in your business. We are interested in making science and code. Think Irchiver as one tool more for open communities to organize their work and help their users. We are not organizing a large scale spying machine. We need some data to test our ideas and then it's up to users how they will utilize the system.

3: Why should I trust you?

We don't expect you to trust us in the long run. We just ask for your understanding for our initial data gathering. Then if someone else decides to utilize the system, you have to consider your trust case by case.

4: Will you send me the raw data you collected from network XYZ?

No. You must contact XYZ administration for this.

5: Will all channels be logged?

Private channels will never be logged. Bot doesn't try to join repeatedly to channels where it has been banned. Eventually it depends on network admistration in which scale logging will happen.

6: opt-in?

We are afraid that we can't gain critical mass if you have to explicitly invite the bots. Remember that our statistical models need large amounts of data, so 10-20 channels won't suffice. We want to make this thing as easy and non-disturbing for the IRC users as possible. If people are positive enough towards our effort opt-in might work.

7: opt-out?

In our point of view, opt-out is analogous to Web crawlers. You may have robots.txt which bans data gathering. Correspondingly in IRC you can ban the bots from your channel. This should work well for public general purpose IRC networks. Users should be informed e.g. in MOTD about general logging policy.

8: Will you remove my comment X at channel Y said HH:MM?

Technically it's possible. We can do it, but still we would like to retain integrity of discussions.

9: Is this going to be public service, available on Web?

We hope that. [Clarification, Nov 21st.] Our main motivation is to study and develop theory and code for our search engine. We truly hope that we can come up with something useful for the community, so someone could utilize our code for providing a public service in the long run. Even then the service would probably contain channels only in some restricted domain.

Probably we'll provide a demonstration or a prototype system on Web eventually but it won't happen in the near future. Especially note that the data we're collecting now will never be shown in public nor distributed.

10: Can't you use just the raw data, no nicks etc. at all?

Yes. This is one possibility. Naturally it makes some interesting services infeasible, but still we could do something with such data.

11: Will you log / analyze / publish user IPs or hostnames?

No, never. We're interested in the contents of discussion.