M, Pay Attention

I sent M Linden a letter -- I've only sent 2-3 of them so far, I think the others have been more open blogs simply addressed to him. One was to answer his initial blog. I think I've always pretty much said the same thing -- pay attention to inworld business, it's where your bread is buttered.

I use to write to Philip Linden all the time. I would get an answer perhaps one out of 20 times, sometimes very unexpectedly. I'll never forget that startling time I fired off an IM, when they messed with the sun prior to putting in that awful Windlight, and said in angry dismay,

"What the hell did you do to the sun? Put it back!"

And *instantly*, via Blackberry, got an answer, "Why, what's wrong with it?"

Scary, eh? Hehe.

It's like that town hall some years ago where I said I didn't believe Governor Linden existed, and suddenly, there she was in the chat, saying "Why do you doubt Me, Prokofy?" Wild!

No, those days are gone. Nowadays, M is like any other marketing puke. He will be very, very cautious to get involved in blogs, written statements, debates, controversies, etc. He is schooled to be always positive and always spinning. He knows just the right time to swoop down, deus ex machina, into Prad's blog, and say just the right thing, and which sims to grace.

So what he did with me -- he didn't answer. And he sent me a bear. ROFL. I asked for a bear just as a kind of check (so often you see Lindens seemingly logging in and getting your notecard inventory, but they never respond). And by doing so, of course, he's dissing me. He's saying two things:

o I'm afraid to answer you, as it is polemical and controversial
o I will not bother to answer you because that would validate you, and I'm definitely not going to make the mistake of my predeccesor and do *that* -- because people like you need to get gone from my service.

Philip used to practice the Linden Method (TM) on me, which was to never answer if he didn't like what I said, or if he felt it was "negative". Perhaps once, in three years, he made an outburst, and accused me of being "negative" (that's what criticism always means to such technolibertarian utopians) -- negative, because I hadn't heard from him in ages and wondered if it was still him really answering his email, as he had gotten so important (during the big SL boom of 2007).

Usually Philip only answered me if he was either really stricken by what I said, mystified or found it compelling as a math problem, or if I said enough positive that he felt flattered. It was a very odd sort of correspondence to be in, because you had to lob the ball in *just right* never knowing what would work.

Like, once, I idly wrote, "Oh! I get it now! On the Internet, nobody knows you're a dog. But on Second Life everybody knows you're a dog." This seemed so profound to me at the time.

And *instantly*, as if he was always hovering just above my avatar's right ear, waiting for me to "get it" about the great essences of SL, he wrote back, "Right. It's the Turing Test".

And then he fell silent...never answering me for about, oh, three months ROFL. That's how it is with these mysterious Lindens.

So here's my letter, FWIW. Others can take it up if they like.

Dear M,

Happy New Year! SL is as fresh and compelling to me today as it was four years ago when I first logged on and started my small business. Thanks for trying to make it better.

I don't get how Second Life works anymore, not that I ever had a really deep knowledge of it years ago, but I could understand how people like Philip Linden or Robin Linden or Daniel Linden went from A to B, even if I didn't agree with it.

I'm told the hippies have left the Lab, and now the marketing people are in charge. Good! I think...

Well, it's like Russia and communism. When they bury that...*thing* that's lying in state in Red Square, I'll believe communism is gone. And in Second Life, the day you allow companies to buy splash page ads or welcome area ads is the day I'll believe the technocommies are finally at bay. So is that 7 Days picture on the new web page an ad buy or a freebie? Is the aversion to commerce subsidizing?

But more importantly, I don't get the politics driving the code these days, since it's confusing and contradictory. So help me out here.

Explain this to me. Surely you don't read the JIRA...do you? Or do you?

How to explain this?
http://jira.secondlife.com/browse/VWR-5491

or this?
http://jira.secondlife.com/browse/WEB-382

(Two proposals vital to liberal democracy and free enterprise that were thrwarted for a long time, and now passed.)

I faced a pack of junkyard dogs for more than a year on these bug finds or feature proposals -- including tekkie Lindens who browbeat me and led the pack at times.

Now, all of a sudden WHOOPS it's CRITICAL and "WorkingonitLinden" and these issues are acknowledge as valid.

Did some big company or dev or university get this across to you? How did this work? Do you even know what I'm talking about?

I have only one suggestion for you, that will remove some of the agida that always occur when you do stuff like raise or lower prices etc.

Hire or assign a Linden to be SmallBusinessAdministration Linden.

A Linden to care about INWORLD business. Not OUTWORLD business. But INWORLD. Those makers of prim hair or shoes that you smushed into one paragraph at the end of your holiday picks. Those rental agents like me. Those provider of all kinds of little services and entertainment and smaller time consulting, i.e. not the Electric Sheep but the electric hedgehogs who take real estate pictures or shops for girlfriends or whatever. Inworld business. You know, those Positive Monthly Linden Flow People whose numbers are tanking lately?

Glenn Linden does devs, Zero Linden does AWG, Jack Linden does land, Zee Linden does sales and LindEx, etc. -- but you need a Linden who holds office hours, writes blogs, and meets with SMALL business owners INWORLD and feels their pain.

If he will feel their pain early on, you will not feel their pain later in the form of bad press, angry blogs, stampedes to OpenSim, etc. etc.

*Every single resident protest you experienced in the last 2 years was from the SMALL BUSINESS PEOPLE of SL. Not the devs. Not the big corporations. Not the educators. But SMALL BUSINESS. GOM. VAT. Casinos. Island price hikes. Island price crashes. OS price hikes. SMALL BUSINESS INWORLD pays the tier; it is the flywheel of the world. Pay attention.

Sincerely,

Prokofy Neva

P.S. please send your bear.

Comments

[ina]

prok, you always make me laugh. (in the good way that i really need) I LOVE YOU!!! ***HUGS***

[Ann Otoole]

Not one Linden has ever sent me a bear. Not even when I asked for one. So I decided I am not a collector of Linden Bears. I might make a new outfit that has teddy bears impaled on spikes though.

Yes a business & economy liaison would be a good idea. I know of an excellent Linden to have this role. Workingonit Linden. The perfect choice. Workingonit Linen could also serve as the voice of the blog. This is because we all know what it means when Workingonit Linden is assigned to a task.

[Prad Prathivi]

You got a bear?!

I feel cheated, now.

[Prokofy Neva]

Yes, I nominate WorkingonitLinden as the inworld business Linden. He has a proven track record of accomplishments, and is a can-do kind of guy.

[Ciaran Laval]

You've raised this idea before, so have I, I've emailed them about it too. Such a role needs to be filled.

[Prokofy Neva]

I realize they won't do it, because inworld business makes them uncomfortable. Inworld business means sex, and what's worse than sex for the Lindens -- land arbitrage.

I get all that.

Still, it doesn't hurt to put out these requests so they do a little thinking about damage control a little earlier and try to be somewhat more accommodating through the existing networks they do have.

Example: if they are contemplating removing traffic so that they can remove the howling about bots, I'd like them to have a Linden who meets not with club owners, who screech and howl against bots because they are in a bots arms race, but with the other kinds of businesses that do not use bots, but who rely on merited traffic to sort results in search/places and make sales.

I'd like them to hear out the concerns about that sector, without being in a group in which every five minutes, either a tekkie idiot like that awful Hermit Barber, or one of the anti-traffic screechers from the clubs forced to use them to compete with other clubs, holler and drown everyone else out. Those people don't make up the bulk of sales in SL, so they shouldn't make an economic determination like this.

I'd like the Lindens to lay out really cogent reasons why they *cannot* keep traffic as a metric, especially as they now have added other metrics, so that Search/all does not order by traffic. There's really no *ideological* reason why they can't leave search/places ordered buy traffic as one tab, and have search/all be ordered with their love tags or whatever the hell they use to order stuff.

Suppressing traffic information is about suppressing freedom of the media, and that is vital to a free economy. It is a free economy's real indicator, once you get over your obsession with the few parcels that are botted up.

Then I'd like to hear why they don't solve the bot issue with a simple CHARGING of bot accounts as premiums given their resource use. Once a club owner has to pay $9.95 per month per bot, we'll see the bot fleets decimated.

Another way they could deal with this is by accepting ARs on bot fleets as abusers of resources on shared mainland sims.

Then that leaves how you address fleets of them on private islands. But there, the libertarian ethic will likely prevail. If someone wants to put 50 bots in a box on their land, who are the Lindens to stop them?

They could selectively deal with this and create a concept called "Traffic Fraud". And they could agree that for those parcels that are abuse-reported as using bots to gain their traffic, they will remove the ability for that parcel to show up in search with its traffic.

Say, now there's an idea! Ima put that on the JIRA. I wonder if that is technically possible. A week of that, and the bot problem might be gone forever! lol

[Darien Caldwell]

But wouldn't that just lead to another form of AR abuse, with unscrupulous business people ARing their competitors to get them delisted from search? I don't know that ARs have ever been a good solution to any problem in SL.

The idea of charging premiums for bots sounds the most appealing to me. But then the question becomes, how does LL verify what's a bot and what isn't? I'm sure they can make pretty good guesses, but knowing for certain is another matter.

[Prokofy Neva]

You can't stop progress because of the possibilities of such abuse. The Linden examining the case can tell if there are 50 bots in a box or not.

We're forced into using this model, as it was the model for how they made progress on ad extortion. First, they made it possible to AR ad extortion in a soft policy indicated in a blog. Then, eventually, after cleaning it up, but backsliding, they finally hunkered down and put it into the viewer as a category you could pull down and AR formally. It's still not clear to me how that wires up to the TOS, and I complained about their vague crimes that aren't in the law but are in ARs, in a specific JIRA, which they acknowledged to fix but then...did nothing with (there are worse problems in it like vague categories like "fraud" or "libel" without sufficient definition.)

In any event, the idea of charging premiums isn't one that seems to grab the Lindens.

The way you verify a bot is by examining it. You are intelligent. It isn't. Here's my Prokofy test, which I find more effective than the Turing test. I ask them if they want to earn $50. If they are not bots, they answer quickly.

Of course Ll can tell bots logging on. This urban myth that they can't, spread by literalist tekkies, is silly. Bots have patterns of activity. They never talk, they never move, they override the normal AFK log out, etc. etc.

Why is that so hard to understand? And I dare say there are coded commands coming in with those log-ons that leave a footprint too. I simply refuse to believe you can't tell bots logging in, as automated programs have to be used to log them in, a, and b, they behave like automatons once logged in.

A week of being delisted from search for traffic infusion will end this problem promptly. I think it's brilliant. It requires political will, and it's the kind of political will we can summon from our laissez-faire libertarian technocommie Lindens, because it will not require them to punish bots, whom they love, and it will not require them to remove a metric and free media communication, which should go against their libertarian conscience.

Instead, they only have to chalk it up to "resource abuse".

[Dale Innis]

It's reasonably easy to automatically detect a bot that hasn't been written to evade automatic detection. Once you put in some degree of automatic detection, it become reasonably easy to write a bot that evades it (it moves around a bit, runs the occasional emote or gesture, says random things). Once someone's written such a bot, it's again relatively each to automatically detect it (although you get a few people kicked off automatically because they got distracted by their dog just after logging in or whatever). And then it's an arms race.

Automatic bot detection would definitely take care of most of today's bots, and might well lower the number of undetected bots in the world. It would also miss some bots, and start to false-positive on a certain number of non-bot users, and would require constant updating if it was to remain viable. LL needs to consider the tradeoffs involved and decide just how much automatic detection makes sense; I don't think the right answer is obvious.

(I speak from some experience here, having been involved with automatic heuristic virus detection Back In The Day; it's not entirely dissimilar.)

[Yumi Murakami]

I think there's still a fairly relible way to detect bots: force them to rebake the avatar then check that they did it right.

[Dale Innis]

Requiring a rebake would detect a fair number of the current bots, but again as soon as the botmakers realize that that's what you're doing, they can just have one or more pre-made bakings lying around, and send one of them back to satisfy the request. And if you start doing really elaborate things like giving them a random piece of clothing and requiring them to put it on and rebake, the real-live residents that you subject to the test are going to get pretty annoyed. :) It may be that there's a really clever way to solve this problem and avoid an arms race, but my strong suspicion is that there isn't. (The ol' Halting Problem is hard to get around.)

[Prokofy Neva]

Dale is a tool -- and a douchebag.

His claims here about bot detection are absolutely like the problem of "if you can see it you can copy it" and he's giving the stock tekkie wiki answer, which is "We will throw up our hands and do nothing, and pretend it is impossible because we don't want it to be detected, and we want to be able to copy."

that's all that's about.

People who need to detect bots get them detected. They don't fuss about arms races and people making up routines to stay ahead of them; it's like copying, and obfuscation. They don't say "Just because we can't do this 100 percent, let's give up and commit suicide". Everybody knows what a bot is in SL. It's not rocket science. What they do. What they don't say. How they are hunched over, etc.

If there is an AR on a bunch of 60 unmoving avatars in hunched mode, or showing as Caspers, or standing straight and starting, or not logging off for hours -- a Linden responding to the AR, seeing a huge box up in the air with 60 of those things hunched over, staring, Caspering, will not have to work too hard to avoid mistakenly banning the real person. They can accept that it's a problem of resource abuse and traffic fraud and take action.

Dale's insolent fisking here is merely once again about trying to privilege the artifacts of tekkies' will as if they should never know any restraints. Of course they should. And that starts with putting restraints on Dale, first, otherwise he goes hog while privileging technology over human beings.

We saw how historically, that got IBM specifically and mankind more generally in trouble.

[Ann Otoole]

If Linden Lab had any intention of dealing with the traffic fraud they would have already done so.

Traffic fraud makes those residents on line numbers go up.

Sadly it has become a general consensus that bots are now required to survive in business and more and more people are setting them up resulting in more and more competitors increasing their bot counts. It is a bot arms race. If 20,000 store owners add 2 model bots each (2 model bots makes no serious difference in traffic so it is generally acceptable) that is 40,000 additional connections to the communications, groups, presence, and asset systems. But next week we will see people adding a couple more bots. And a few more the next week. And in a short time the grid is shut off to logins for longer periods multiple times a day.

I don't really see any way around Linden Lab creating policy to deal with resource abuse related to the bots in a box thing. Bots can't be outlawed because not all bots are bad and we need advanced NPC units developed. So the only possible solution is enforced policy.

But it is Linden Lab's grid/system. They get to do, or not do, as they please with it.

Programming anti bot systems is a waste of time. Whatever you can code there is some bot maker that can out code you. The L$ farmers running the thousands of camping bots have nothing but financial incentive to defeat any attempt at bot detection. An enforced policy on the number of campers or bots allowed per parcel will take care of the issue. We just have to make sure Linden Lab doesn't try to make the policy say one avatar can only stay so long on a parcel. The L$ farmers would simply program rotations in.

[Maggie Darwin]

I say we build a bear for Workingonit Linden.

Or maybe a set of three...modeled along the lines of sanbiki no saru

http://en.wikipedia.org/wiki/Three_wise_monkeys

[Dale Innis]

I didn't say (or secretly mean) that we shouldn't try to detect bots just because it can't be done 100% correctly. I just pointed out that although (some of) today's bots may be easily detectable by some particular method, as soon as that method is in wide use, people will develop bots that it doesn't detect.

That doesn't mean we shouldn't try to detect them at all; it just means we should realize that we will do an imperfect job, and take that fact into account in our plans.

[FWord Utorid]

if someone buys a sim for $1200 usd and $300 usd monthly, and they want 100 clouds of smoke with nametags in it, they can have it.

What sort of hubris encourages you to believe you can tell them what to do on basically what is their property?

You are afraid of technology because you do not understand it and are too lazy to learn, and thus, you do not have the advantage.

In a world where technology is the foundation, it is foolish / stupid / handicapped / retarded / immensely dumb / beyond belief / there are not enough words to proclaim the level of ignorance you support when you try to put down technologically proficient people. It is, in essence, calling the Emperor naked, and you should have no surprise when your cries of foul play fall on deaf ears when reviewed by the technically proficient people who run the show.

The irony is this: In order to facilitate the anti-technology solution you are proposing, you are going to have to employ those "tekkie wiki" types you have purported to despise so readily. In essence, you've shot yourself in the foot. These people will never in a thousand years assist you in your drive to extinguish their advantage. So even if there was a brilliant mechanism for detecting bots, by alienating those who could provide you with it, you lost before the war began.
On the totem pole, your priorities will rank dead last because you disrespected the community you needed most to get what you wanted.

Game Over

[Prokofy Neva]

Because I'm not a rapacious anarcho-capitalist and techno-libertarian, FWord. I'm not a Randian or a Friedbergite or anything of the sort. I don't think property rights so trump every other right that somebody can get to do what the fuck they want on their property endlessly and impact the freedom of others and their property value -- and that doesn't matter.

The notion that 80 cloudy bots on a sim doesn't harm anybody is hedonistic licentious libertarian hogwash.

Oh, I understand technology *very* well. I live inside it in SL. My criticism of it hacks at your abusive control of others, and your interests, and you don't like it -- understood.

Technologicall proficient people do not get to harm others with their mechanical "intelligence" that bots around and hogs resources, inflates traffic, prevents log-ons, and harvests information against people's will -- to name just some of the damage that unrestricted botting causes.

And it will get worse.

It's not about technology. It is about something very old and very organic and very human: the lust for power and the greed for gain.

The Lindens will do fine detecting bot fraud, like they did with ad farms, that everyone said "couldn't be done" lol.