Concurrant residents logged in.

Right, well, I think Sling is correct that they'll first blog the "official" landcutting policy, and then tee-up a draft of the traffic-gaming policy. We can be pretty sure that the initial traffic gaming policy will *not* include removal of Traffic from the Search metrics because it "will take some Development," but it seems as if that step is still on the table and maybe even tentatively planned; perhaps they'll reveal a bit more when they blog the anti-traffic-gaming thing. Obviously, patience is a virtue when anticipating LL action; they still insist that the Blog / Forums upgrade is in the works, now a full quarter late by my calendar.

Even the anti-traffic-gaming policy is one that we've heard many times would never happen because it will deflate LL's own concurrency metrics. It's at least encouraging that they're going to take this step.

And you know, if there are still bots around the grid being useful, that's cool with me. I think Elanthius's landbot flock performs a useful function, for example. Sure, he makes money from them, but that's only possible because the seller also benefits from the liquid, low-friction market for their unwanted land. And Group-invite bots are indispensable for many applications, notably large-scale land rentals. (It truly sucks, however, that these things are necessary. The viewer protocol is a sorry excuse for an API, especially from a systems-level view.)

(God help me: this is my 5000th post to these Forums. )

If that's what one of them said, then s/he is mistaken. They already did it once in an RC - replaced the traffic rankings with the All search, filtered by Places. It been in their roadmap for search for a long time. They said they changed it back because some people couldn't use the All search - it wouldn't scroll for them and stuff like that. From a post I read here recently, it appears that they've fixed that problem, so there should be nothing to prevent them from doing it again.

That's all they need to do to deal with the traffic bots and camping, *IF* they really want to do it, and it would be quick easy for them. It doesn't need any new policies of any kind. The value of traffic in the All search is so tiny that bots and camping wouldn't be worth bothering with.

The thing is, everybody knows that. It's been stated many many times and nobody objects to it (maybe the odd die-hard Places tab user). So why haven't they done it? For themselves, they don't want to do it, and I suspect that the blog announcement will be just a silly fudge, designed to maintain the numbers.

Sigh.

Average = Typical user.

I am not going to argue with you. As you are challenging me on this issue uselessly it seems.

*I* have asked people - non-forum people - and the response i get repeatedly is what i stated above already.

I do not need to be lectured with your view of "gaming traffic is misleading, etc..etc.."

What is misleading? What is that resident looking for when they do their search? How can you say it is misleading to all residents when there are residents who go to places with high traffic EVEN THOUGH they know people game traffic because they say it reflects the store cares to attract people therefore they go to see if anything is worth seeing?

Jesus, we can go around and around all day with this. I am not arguing that bots are wrong or bots are right. I am simply stating that from asking several residents from 2004 to 2008 rez dates what they think about places with high traffic numbers. And this has led me to believe overwhelmingly that the average or typical user - #1 does not care and #2 is unaware.

This entire thing you do over bots is really getting tiresome. What can you not recognize about LL's nonchalant attitude about bots and traffic that is telling you that THEY DO NOT CARE - and that the residents use of traffic and bots is FINE with them, it is called "acquiescence", and this is a case of it if there ever was on that existed.

I do not understand why someone feels the need to have a 100,000+ traffic rating in the furniture section. All you really need to do is generate enough traffic to appear on the first page of "Furniture" search returns and you will see a marked increase in your stores traffic and sales, we know this empirically.

Sling, why don't you turn your frustration to LL instead of the repeated appeals to the ethics of forum users who make up maybe 1% of the SL population. This is LL's fault. LL created the metric, LL has the power to kill it in a matter of minutes.

The month that we had "DWELL" created in 2004, i asked Catherine Omega if there was a way to reward people on my parcel only (Aqua Galleria) for being there so that i could increase my Dwell...That is just the way things work. If you want change - deal with the people who can make the change. Your fight in the forums is honestly commendable, but at the same time it is a public lesson in futility for the rest of us to bear witness to.

Sorry if i sounded harsh.

LL is never going to do it.. You can count over 500 bots alone from the top 10 furniture listings (and 500 is a gross under-estimation, as i did a count several months back i think and it was pretty high i think). The bots increase concurrency which makes SL look legitimate and active. And LL does not want dropping numbers they want increase numbers.

But i would be happy as hell if 60% of the mainland derezzed and we dropped down to 20,000 REAL PEOPLE.

Well, to some extent, this is already the case. A business place in search is dependent upon, among other things, keywords that appear in your item names and descriptions. Example: our aviary wants to rank high for the keyword "bird" when searched, so we put the word "bird" in all of our item names and descriptions (twice in fact). Now let's take that one step further and assume that one can, hypothetically of course, rez a hundred trans prims, bury them underground, and name them each "bird" with "bird" in the description field, then click "show in search" for each. THAT would skew the keyword stats in ranking.

I'm just sayin....

Congratulations and best wishes for many more pleasant and productive posts.

Not at all.
You simply misunderstand.

I know that talking to Lindens doesn't get us anywhere the Lindens couldn't be arsed to be.
I have absolutely no illusion that posts here have the slightest effect on LL.

All that is happening is that I see some muppetry on the forums, and I call it.

If some people find it tiresome, then they might understand that I find the muppetry tiresome.

The problem with that approach is that it would be detrimental to the place's rankings, but I'm not saying why

Not tiresome enough, it seems, as you are always quick to jump in

OK, I have been fighting the urge to ask and am giving in....why?

You didn't put up much of a fight - your post came only 2 minutes after mine LOL

LL can't really believe that. They aren't stupid. If they're counting on residents being duped by traffic numbers they're setting themselves up for a fall. I think they're smarter than that.

Only experienced residents know that. People looking and joining, plus many newer users, don't know that at all.

You could limit everyone to 1 or 2 alts that are linked to your main alt... via email addy like it used to be... I am not saying to limit everyone to 1 account..

Google search generally identifies such practices as keyword spamming and has a means of lessening a parcels relevance for that term. Whether Linden Lab implement that is another matter, I'm not convinced they do.

It's not residents they're interested in, it's people who aren't here and high user figures are a good seller. WoW's figures are often questioned too, but the headlines are that WoW has however many million accounts these days.

Phil-- I am sorry if you were hurt by my generalization.. but focusing on my "Yahoo" comment, you are missing the overal point of my first post of the thread.. which is what are the real numbers... As far as real data.. if we HAD the real data from LL to this question that I have raised I am sure this thread would be bigger then that "Just let this one die". I am not claiming consipracy.. no tinfoil hats here.. it's a curiosity... we were told a year ago traffic would not be a factor yet... even by your own admission (which I am not attacking you personally) we still have 1000's of bots in operation....

I am not complaining that you have more traffic.. or that this person has more traffic or I dont have any traffic... This tread has nothing to do with traffic... It's about grid load.. and the decline in quality in the last year...

I am happy that you are a successful business person.. great

My Main point was lost along the way I am afraid.

The post of mine that you quoted was a reply to Arielyn

I wasn't hurt at all by your post. In your initial post, you called people like me "yahoos" and I responded to it. You also implied that traffic bots don't have any effect on sales, and I corrected you about that, as did 2 other people.

Your main point, about the amount of load on the grid that you imply, doesn't hold up. Every logged-in avatar causes some load, of course, but well-used traffic bots each cause a very tiny amount. We've been through it all before - over and over again - and nothing has changed. LL is about to post a blog about it all, so there seems little point in rehashing the same old arguments again. Suffice it to say that, when SL buckles under the strain, it's because of the number of active avatars that are logged in, and not because of the much smaller number of inactive, well out of the way, traffic bots that are logged in.

So I am to understand that a Bot is different from an active avitar? I mean a bot you can "pie chart", has a profile, you can boot them off your land.. and they do show up as a green dot... ok fine.. they do not buy anything, nor interact... I see your point.. however if we were talking about 1-100 bots on the grid.. sure it would not be an issue.. but from what I have seen the average bot user has approx no less then 5 bots running at once for traffic... Sure not everyone runs bots.. but I think there are more out there that do not admit to it.

I mean help me out here.. If this is not looking at the big picture.. show me where I am wrong.. you may have rehashed this with someone else.. however I think there is more to this then what is acutally released to the public. I did not call for a blog post.. i just asked for data... If that is what it is going to take to shut us up... then by all means.. MR M. Linden.. show us the data.. Tell us "Anti-botters" that bots have nothing to do with the Grid slowing to a crawl impacting everything from the happiness of residents (yeah I know can't please us all) to the economy... and yes it DOES impact the economy when stuff will not rez or you get stale money tranfers

and sorry for quoting the wrong post.. my bad

I certainly can't speak for M, nor Phil, but it's true: trafficbots have very little performance impact on the grid as a whole or even the sim where they're located. That's because the bot builders program libsl/openmv to minimize interaction with the world. They do this in their own self-interest, to keep their network traffic to a minimum. So the bots send and receive the absolute minimum number of textures, animations, inventory requests--just updates of any kind. And they're usually parked well out of the way--not just to hide them, but to keep them from having to exchange update packets with other agents in their immediate area.

All this is in contrast to some other bots that constantly whack Search as often as they can without triggering the circuit-breaker timeout, or TP around the grid scanning for land, etc. Of course there are very many fewer of those than trafficbots.

The big caveat about all this is that we really don't know how much overhead is placed on some central services by a completely quiescent login session. I'm thinking Presence, mostly. It *shouldn't* be much--a trafficbot has no Friends list to update, and maybe at most one Group--but we really don't know all the innards of Presence. So we non-Lindens really can't know for sure if one average human-operated agent is equivalent to 10 bots or 100 or 1000. We do know it's far from 1:1, just by watching network packets and sim performance.

There are valid reasons to be anti-trafficbot, but those reasons are about Traffic-gaming and/or Traffic itself, not about grid performance.

What Qie said

Take my traffic bots as an example. There are 13 of them at over 4000m, and 3 more at 3000m. They are not in range of anything else in the sim. So, unless a person flies all the way up to have a look, nobody ever needs to download any data about them, and they don't need to download data about other people. In other words, once they are in, nothing happens to trouble the system's resources - they are not active. (I'll come back to that)

On the other hand, I have 6 dumb store models that are the load equivalent of afk campers. They do need to download everything within range, including the textures and movements of anyone who comes within range, and everyone who comes within range needs to download data about them - shape, clothes, etc. They don't move by themselves, so they don't need to continually download object textures as more come into range, like an active avatar does, but they do cause load, so they could be described as semi-active.

A real person avatar is usually fully active - on the move, continually downloading textures, movements, etc. as objects and avatars come into range.

So there is a *HUGE* difference in terms of load between an active avatar and non-active (well-used) bot.

I said I'd come back to something. Every logged-in avatar is bound to cause a tiny amount of work for the system to do and non-active bots are no different - but it's tiny. After that, the difference between an active avatar and non-active avatar is huge.

What happens when the system buckles? In the past, it's been the asset server that's buckled, and more recently, it's been the database. Unlike active avatars, non-active, well-used traffic bots don't trouble those systems unless someone decides to fly all the way up to where they are, which is very rare, and even rarer for it to coincide with a time when the system is close to buckling.

So you see that well-used traffic bots are not the cause of the system buckling. It's active, real people avatars that push it over the edge sometimes.

Not exactly true Philip.

If you go into an empty sim fly to 4000m and then run GLIntercept, you are still getting textures from 4000m away on the sim ground surface.

The simulator does a pretty poor job of filtering textures to that are within your draw distance or outside of it.

I am not saying they are doing a full time download like we do when we are around a bunch of things, but they ARE getting some texture data.

hah i called you Philip.

I wonder why any of your 13+3+6 = 22 bots have to download anything at all.
Surely the bot program could be written so as not to request any data other than that required to keep them in place?
The store bots are a load for other avatars, but they don't need to know anything about the world other than the object that they are sitting on.