tea.mathoverflow.net - Discussion Feed (Roaming problems)

Scott Morrison comments on "Roaming problems" (5584)

2010-05-16T21:38:22-07:00

As Harald had guessed, I currently have no particular interest in who Regenbogen is. I found out what I did back when we were worried about multiple users. Once I heard from Anton that he'd been in touch with him I didn't follow the matter any further. Since anonymity had come up again, I thought the surprising easy with which I'd discovered what I did was germane.

Regenbogen comments on "Roaming problems" (5583)

2010-05-16T16:27:49-07:00

@Harald Hanche-Olsen.. For you, maybe it is so. But the mods did point out certain issues in private communications, chiefly things I did during my use of multiple ids. Anyway if I dwell on this matter too much I might say something about which I would be regretting later. So I am leaving this discussion here.

Harald Hanche-Olsen comments on "Roaming problems" (5582)

2010-05-16T15:42:48-07:00

@Regenbogen: I am not aware of any issues regarding your actions or behaviour on MO. You have no reason to be afraid of getting into trouble here, nor can I imagine why anybody would try to unmask you. Now relax and enjoy the mathematics.

Regenbogen comments on "Roaming problems" (5581)

2010-05-16T15:15:31-07:00

Ok, according to Scott the issue was having multiple accounts. For that it was not necessary to make inquiries at my institution.

If there are any issues about my actions or behavior, I would be perfectly willing to leave the site voluntarily. I do not want to get into too much trouble for anything. Without any hard feelings I will leave. But please do not do things like this.

Harald Hanche-Olsen comments on "Roaming problems" (5580)

2010-05-16T14:53:19-07:00

This is rapidly getting off topic, but: I seem to remember a story from way back where some would-be anonymous person on the net was successfully identified using textual analysis. Humans are creatures of habit, and we have our little idiosyncracies in writing which could make it quite easy for a computer program to identify different writing styles. Hence, if you write prolifically both under your own name and anonymously (or pseudonymously), a sufficiently motivated person will likely be able to identify your writings as coming from you.

@Regenbogen: As for why Scott would be interested in tracking you in real life, I see no indication that he is interested in doing so. His comment seems natural in the context of the present discussion of anonymity, however. There is very likely nothing sinister in it.

Regenbogen comments on "Roaming problems" (5578)

2010-05-16T12:41:50-07:00

@Scott Morrison.

I didn't make any attempts to hide my identity from the moderators. It was with the trust that you wouldn't divulge it to others. Such was Anton's assurance to me once and I took it at face value.

I do not quite comprehend why on earth you want to know who I am and track me in real life. You made such allusions about such matters earlier too. Ok, what if you know my institution? Who am I to you, more than just another grad student over there? There are hundreds of active users of this website and your moderator time is to be spent on each of these people. I am not any special. We never even know each others' existence before interactions at MO. If you have any problems with me using this website, you could have just told me and I would have been out of here by myself. This was something I had made clear earlier too. Without anonymity, I wouldn't be able to use this site.

Scott Morrison comments on "Roaming problems" (5577)

2010-05-16T12:23:15-07:00

@Regenbogen,

if you are actually that concerned about anonymity, I should remind you that anonymity is really hard. For example, I'm fairly sure that I know the institution you're at. This comes from various server logs. Further, knowing several people at that institution, and having had conversations with them about anonymity on mathoverflow and the potential problems of multiple accounts, I have a guess at least about who you are there. The mathematical community is rather small! Of course, I'm not going to use or reveal this information (with some exceptions, of course), because I don't particularly mind people being anonymous. I just wanted to let you know that it's very very difficult!

Regenbogen comments on "Roaming problems" (5573)

2010-05-15T10:11:42-07:00

Of course I do not want my identity to be discovered.

But, well, I had installed it for purposes outside MO.

Harry Gindi comments on "Roaming problems" (5572)

2010-05-15T10:10:08-07:00

I feel like you're overly paranoid about your identity being discovered....

Regenbogen comments on "Roaming problems" (5571)

2010-05-15T09:53:16-07:00

Well, what I do was to connect my mobile via USB to the laptop to use it as a modem and then use from the laptop. It was running linux and I had recently installed the privoxy proxy server. It turns out that privoxy somehow mangles things up if I restart it. I take down privoxy, everything works well.

Sorry for troubling you guys!

Scott Morrison comments on "Roaming problems" (5569)

2010-05-15T09:10:16-07:00

@Regenbogen,

I've never had a problem, either. As far as I can remember, I've logged in once per device I own.

Harald Hanche-Olsen comments on "Roaming problems" (5567)

2010-05-15T06:19:34-07:00

@Regenbogen: It might be helpful if you told us what kind of mobile you are using, and what web browser you are running on it. Maybe others have the same, and can confirm the troubles. I just ran an experiment myself, with my iphone (and safari): I logged in, using the wireless network at home. Then I turned off wireless on my phone, so my IP address would be totally differnet and followed a link to a question. I was still logged in.

One problem with debugging such things on mobiles is that the impoverished user interface may not let you examine individual cookies. But if yours does, look for cookes for mathoverflow.net and compare with the cookies you get on a “real” computer.

Harry Gindi comments on "Roaming problems" (5565)

2010-05-14T15:46:22-07:00

This is going off-track.

Awful train pun. Your question has been answered, so I see no reason to get this thread "back on the rails", as it were.

Regenbogen comments on "Roaming problems" (5564)

2010-05-14T15:44:49-07:00

This is going off-track.

Harry Gindi comments on "Roaming problems" (5563)

2010-05-14T15:43:14-07:00

Then what do trains have to do with it? Anyway, it seems likely that your "mobile" is "broken" or "not saving cookies". Maybe you should update the browser or change the settings.

Regenbogen comments on "Roaming problems" (5562)

2010-05-14T15:31:12-07:00

Well, you cannot use the mobile most of the time inside the train. And never when it goes underground.

Harry Gindi comments on "Roaming problems" (5561)

2010-05-14T15:28:45-07:00

Then wouldn't you be "on the rails" rather than "on the road"?

Regenbogen comments on "Roaming problems" (5560)

2010-05-14T15:20:09-07:00

I didn't say that I drive. There are places on this earth where trains are an integral part of life, you know.

Harald Hanche-Olsen comments on "Roaming problems" (5559)

2010-05-14T15:15:38-07:00

Like Harry, I never see this problem. I can't even recall the last time I logged in to MO. I am sure it was months ago. And since then, I have accessed MO from home, from the department (wired and wireless, which gives me different IP blocks), via three different VPNs, and from three other locations widely scattered geographically. FWIW, I use openid delegation via http://www.math.ntnu.no/~hanche/hanche/ (there are weird, pseudo-technical reason why I use this URL) and Firefox 3.6.3 on Mac OS X 10.5.8. And I don't block cookies. I count no less than 11 cookies from mathoverflow.net.

Harry Gindi comments on "Roaming problems" (5557)

2010-05-14T14:41:26-07:00

I don't get logged out when my IP changes (I just checked this when I brought my laptop to a friend's house earlier today). Your "mobile" must be broken (or it's not saving cookies). Meanwhile, why are you accessing MO while driving "on the road"? Isn't that a little dangerous?

Also, the title of this thread makes me picture a majestic herd of roaming problems.

Regenbogen comments on "Roaming problems" (5556)

2010-05-14T14:36:55-07:00

Consider a situation in which you access MO from various locations. For instance, you use MO on your mobile and you are on the road.

Then when you move to a different location, your IP changes. And then all goes havoc with the MO page. You get logged out, you see the annoying "first time here? check out the FAQ" banner page(annoying because I am not a first timer) and all that. And logging in using the mobile is harder than with a PC keyboard.

The issue is with MO. Google, yahoo, myopenid, all that has no problem with roaming. Even if the IP changes, the cookies track you carefully.

The same happens if you access MO through a proxy. Or if you have a dialup connection and your IP changes the second time. Of if your internet connection is mobile, and you are given a dynamic IP address. You are forced to login again and the unnecessary banner message appears. It should be enough to have the cookies and your session should be preserved even if you change your IP.

Is it possible to fix this problem?