Tateru Nino

There's not much point giving you a list of server names and addresses, because there's about 20,000 different server IP addresses and 20,000 different server names, on different networks across the USA. If you think that there's something like 4 addresses that you can plug into a firewall, you're wrong, sorry. Besides, the voice-chat doesn't even come from the same set of 20,000 server addresses that everything else does. That comes from a different company's servers (and I have no idea where and how many of them there are). Follow the instructions dealing with opening up and forwarding the appropriate port numbers, and you should have some success.

There's between 800,000 and 1.1 million Second Life users logged in per month - but there's very little data on how many per day. Based on concurrency data and a few not-too-unreasonable assumptions, I'd say that there's roughly 150,000 to 175,000 users on any given day. Only a small number of those, however, would be people who log in seven-days-per-week, though.

Ah, no - having payment information on file is not directly relevant to 17 U.S.C. § 512(i). I was talking more about the process for ©(1)©.

All card payments to Linden Lab are processed through a European payments processor called Dragonfish (whose main business is processing payments for online games and online gambling sites). Linden Lab may not accept prepaid cards because Dragonfish doesn't, or because they've asked Dragonfish not to (many businesses won't accept prepaid cards, and I'm not entirely sure why), or because something is broken in payment processing and isn't working the way it is supposed to (of which neither the Lab nor Dragonfish might be aware). Dragonfish's support of payment methods to the Lab has been a bit spotty in the past. My best suggestion would be to contact the Lab's billing support. They can tell you which of the three options it is, and if it isn't caused by a matter of policy, they may be able to get things fixed.

"are available to any person on reasonable and nondiscriminatory terms" That's referring to copyright rights holders, by the way. It means that you don't impose any undue barriers to rights holders like shutting them out of a service, or making them pay a lot of money or treat them otherwise differently to other users when it comes to their checking things out to see if their rights are being infringed.

It's actually pretty straightforward: The terms of service clearly aren't sufficient. Payment information on file means that there's a direct identity to file lawsuits against, if a copyright holder wants to. 70% or more of people have no idea what copyright is let alone what it covers or what they should or should not do (not kidding, it's really that high). A little education is not unwarranted. Going through the mesh-license process comes close to ensuring that you'll get stuck (if you infringe) with wilful infringement penalties (which are treble) in jurisdictions where those addtional penaltiesm are available, since it isn't just a shrink-wrap license. It provides an additional legal protection to Linden Lab. (ie: It can't be shut down for being negligent about this)

I know what causes it. What I don't know is why it fixed itself.

That's how most PR people sound. There's a particular tone that requires practice to achieve.

It is indeed an official account and it is most likely to be a marcom group/role account. Probably Pete is the most likely person to be using it, but it is possible that you'll find others posting with it as well.

It's generally bad news, though if the software is well-written it should neither crash nor fail if presented with a V6 address where a V4 is expected - that's just good programming, even before V6 was even concieved. Of course, over the last 15 years or so much software has been developed to handle either case. Most of the software you use today already handles it. I know a number of key pieces of SL's server-side software are IPv6 happy. I had a hand in the design of some of those, and in a couple cases with getting the IPv6 support in there. But that wasn't software produced by the Lab. What the Lab's level of software engineering looks like is unknown.

A lot of my blog traffic (which, obviously, is mostly from SL users and those interested in SL) is over IPv6 these days. When I switched on support for it, it was only 5% or so, but it has grown significantly through the year. IMO, the improvements alone are worth it. Also, we're not talking about cutting IPv4 support at all. Just providing the option for those using v6 to have access to the advantages.

You're assuming that completed signups turn into actual logins, which really isn't the case in 99+% of cases. For example, during the 'boom times', 30% of new signups never logged in at all. Another 30% didn't log in for the first time within six months of signing up. For the remainder, the median time before new users stopped logging in was 48 hours.* * These are very approximate figures. Now, I don't see any reason that new-user retention should be significantly improved from that (and actually, I can see quite a few reasons why it should be much worse) - and your own take on economic performance (it seems to me) reinforces the point that retention has slid significantly. I cannot pin down exactly what user retention (out to seven days), actually is, but I'd be surprised to find it significantly over 0.1%. The sudden surge in completed (a very important word, that) signups began when the registration process was refit. That is, there doesn't seem to be any noticeable increase in the number of people commencing the registration process, only in the number of people getting to the end of it. Or, in other words, picking a name, putting in an email address and a password (and confirming it) and making a few other additional choices was simply too difficult or complicated for approximately 40-50% of people who attempted to sign up. Streamlining that process may have gotten them through the signup phase better, but frankly, how well do you think people who found the previous registration too difficult would be coping with their first login session in-world? I'm thinking: Not well. Attrition of the user-base is inevitable - even if the platform is flawless and Linden Lab does nothing wrong. There are circumstances which may accellerate that attrition, and you can probably think of more than a few. Assuming those losses aren't replaced with fresh-blood, you'll see... well, you've seen it already, right? That's why we're having this conversation Where are all those completed signups? They're not using Second Life.

There is a distinct, statistically significant increase in completed new user registrations, yes. It started at a specific time and has been ongoing since then, and it is demonstrably not the product of the registration of alt accounts. I am not saying that alt-registrations don't happen, but it is clear from the available information that this increase in registrations is not due to that.

Not all of those are still live anyway. The last-60 figures, for example (and a couple of others) were frozen quite some time ago.

and no one bothered with the official feeds? The official data feeds have always been somewhat erratic, sometimes failing to update, or coughing up old data. It takes quite a bit of data-quality processing to sort that out. Not for the faint of heart so much.

Depending on which country you live in, it may be unlawful for you to send identifying information to a foreign party - for example, your drivers' license number, passport number or national ID number/SSN. These numbers are, in every case that I'm aware of, not a part of publicly-accessible records, and can't be verified in any case. (Exception, SSNs appear next to other personally identifying data in USA voter lists that are publicly accessible records).

If you want an answer from a Linden Lab employee, you'll need to contact the Lab's support people, or its PR department.

Actually, Linden Lab developed the first hypergrid prototype in conjunction with opensim, back when. Linden Lab, however, abandoned the idea some time after.

Tried the tool out. Discarding outated results the number seems to be ~8 regular sims per server (I only checked sims I know that weren't lightweight or special-class sims), however just determining the number of effective cores on a system isn't as clear-cut as it once was. Physical cores no longer have a 1:1 relationship with effective cores, and that can be someone dependent on system configuration.

Last I heard, the number was variable. Homesteads and Openspaces for example, used to run at least 4 sims to the core (hence they used to be called 'quad' sims). Nowadays, I hear talk that it might be 8/core or even 10/core. Regular sims used to be 1/core, but might now be 2/core. How many cores per server isn't known to the best of my knowledge.

The number is in the thousands - there are presently 31,250 simulators and several sims run per server, plus there are a number of inventory servers, asset servers and utility servers. Last I was aware, the servers were located in colocation facilities in San Francisco, Dallas, and the District of Columbia. Your connection jumps from server to server as you move between sims. No user-accessible servers have been deployed outside of the USA to-date. There have, at various times, been plans to deploy some simulators on servers in the UK, Australia, China, Japan and Korea, but none of these have apparently ever reached the stage where hardware was deployed.

Probably 'Rhiadra' which was a first name that was available in the latter part of 2008. Nyll is right, however - nobody who can check the status of your account or make any fixes to it is likely to see your message here. You should file a support ticket, or call customer-service.

If they're flagged as scripted agents, they don't generate any parcel traffic. That's actually what the flag is for - exempting an account from the traffic system. Some people set it even though they're not bots.

I did some math back during Second Life's peak, when the number of ARs moving through the system was easier to track. At peak times, each on-duty member of the abuse-team had to deal with a new AR every 20 seconds. So, that was 20 seconds to decide whether to take immediate action, ignore it, or defer the report for later investigation to try to determine if it was correct. It's difficult to judge what the load on the abuse-team might be right now, but I don't imagine that they have more than a minute or so, most of the time to make a decision on each report. I imagine that leads to bad decisions, or things getting forgotten or ignored. I suppose the Lab could find a way to put more people on that duty, if we were willing to pay more for them - or willing to have fewer resources allocated to other areas.

"never to take at face value a graph that doesn't have zero as the origin of the Y axis" I have charts that show that as well, and which also show a larger range.