Welcome to the Second Life Forums Archive

These forums are CLOSED. Please visit the new forums HERE

Linden Labs doing everything they can?

Maklin Deckard
Disillusioned
Join date: 9 Apr 2005
Posts: 459
04-30-2006 18:15
From: Alex Fitzsimmons
Maklin strikes me as extremely selfish. I would like to see SL "fixed" (if that were possible, which I doubt given the open-ended nature of the game) for everyone BUT him. Since he doesn't care about anyone else's experience and would happily ruin theirs just to make his better, poetic justice would see only his experience ruined instead.

As for the poll, I voted "yes." Next time, try making a poll with real choices.


Not really, if what Sera_Cela Linden says is true, and it would break vendors, flight boosts, weapons that use physical combat, and cause ingame tooth decay</sarcasm>...it would hit me....ALL my vendors would break, my vehicles with physical combat systems would be broken (many of which I could not get updated versions of), most of the guns I collect would break, several of the items I sell on my vendors would have to be reworked due to code changes. Would I enjoy having to redo all the above or spend Lindens on replacements for those that were unfixable? Nope!

But I still support 'crippling' LSL. Sure, it would make the game better for me in the long run, but since the vast majority of SL players (by lindens own admission) are not content creators, but consumers, I dare say it would long term, benefit more than me (and irritate some content creators, which I honestly cannot say I oppose). :) What good is all this content if its down constantly and unavailable?


BTW, I really don't care all that much either way. Mostly only got involved in this thread due to one person being vacuum-attached to Phillip's posterior, having to jump to LL's defense after every post calling them to task. :) I understand taking up for them on occasion, but damn....someone's a better PR flack than their PR flack! :)
Sera Cela
A little bit of crazy
Join date: 15 Sep 2005
Posts: 197
04-30-2006 18:16
From: Maklin Deckard
And the lindens do exactly WHAT about it? Platitudes and 4-6 hr cleanups...but no preventions. No prosecutions (and they should modify TOS to allow revealing character names and what punishments were inflicted...they hide behind tos on downtime). Their lack of fixes and lack of serious enough deterrent punishments makes them as equally culpable as the nitwit that crashed the grid.

And no, I understand the issue. I understand when WoW finds something a player can do to cheat (dupe) or destabilize a server, they damn well fix the problem even if the players bitch and whine about features being removed or nerfed. And they have removed entire guilds of troublemakers....but LL tolerates the groups from a certain website that have (at least once in the past) had members causing problems. LL makes no effort to rectify the problem or compensate the premium players...WoW does.

LL is too busy genuflecting to the ingame capitalists and coders and won't do squat to fix the problem because it might make people mad if their flashy swords stopped being lagmakers.

First off, when wow adds any player generated content, or scripting ability let me know, otherwise stop trying to compare apples to oranges, bringing down servers to fix an exploit is not the same as a user bringing down the grid against the will of the company.

It's really easy to bitch and complain when you don't understand the facts. Sure i'm not happy the grid is down, however since I know how the system works and it's vulnerabilities I don't flip out and start calling the lindens incompetent. I get pissed at the moron who released the item, not start demanding LL cripples the game to prevent stuff like this from happening.
Introvert Petunia
over 2 billion posts
Join date: 11 Sep 2004
Posts: 2,065
04-30-2006 18:20
From: someone
...I agree, however your asserting that this is a "purely technical problem" which it simply is not, which makes your entire point invalid. Any changes to how the grid or lsl works impacts every single person on the grid. Usually in ways they don't even forsee. Remember the backlash when they stopped allowing objects to give other objects inventory? Now imagine if they put a limiter on rezzing objects from inventory, many weapons would simply be destroyed, some flight accelerators, "holo" vendors, all sorts of innocent bystanders would be gone. For example, I have this amazing sword, that will create 32 particle emitters that flow around me causing an amazing looking cherry blossom petal effect. Since there is no way for the asset server to track "generation" of items, things like this are simply undistinguishable from an object that will cause havoc accross the grid. And now your problary thinking "then why not have it start tracking generation" this is where the fact that I said it cannot be done without a significant change to the current system. When you rez an object, it's not tracked by the asset server anymore, it's tracked by the sim that it's on. Tracking generation is simply impossible without getting rid of the central asset server, which is not a minor change, and would open a whole new bag of worms, like seperate sim asset server crashes, which could be even more damaging to inventories.
I understand your point, but not every solution requires drastic changes as in the existing function breaking examples you cited.

Just to throw out a simple example of how this could be done without affecting the structure of the asset DB or gimping LSL. The "grey goo" attacks have a feature that could be exploited to detect and abate them - unlike the legitimate examples you cited above, the grid attacks lack locality of rezzing. One can imagine a meta-process that watches the asset rezzing and looks for patterns of rapid non-local rezzing.

This is similar to the heuristic used to stop TCP SYN flood DoS attacks that were threatening the entire internet by exploiting a flaw in the 1970s TCP protocol designed in an era relatively free of hostile crackers. If a network device sees a sharp spike in SYNs left dangling, they assume that a SYN flood is being mounted and start discarding SYNs from the originating addresses. Thus the response is targeted and kicks in only after detecting a known pattern of attack and affects only the originating machines. True, this also may cause a DoS of legitimate traffic from those addresses, but it is preferable to having all connections fail because some rogue machines are mounting an overload.

Have I proven here (or to myself) that this will fix it? Nope. Does it stand a decent chance of mitigating it? I think so.
Luciftias Neurocam
Ecosystem Design
Join date: 13 Oct 2005
Posts: 742
04-30-2006 18:23
From: Zonax Delorean
Well, I still bet if a 'real terrorist' was using that card, he/she'd be at Guantanamo already.

But in this case, too, I don't think it's too hard to trace the card to a person. It is, after all, the point of credit cards to have everyone and every purchase in a database, so everyone can be traced anytime.


A couple years ago I stumbled across a nice google hack that permitted anyone searching the right combination of numbers to find oodles and oodles of credit card numbers, names, and any other info you might want. Some people may remember this little bug. If I'd been a much worse person, I would have a big screen TV, a really nice car, and some gold capped teeth right now. And lots of fake SL accounts.


But I'm not that bad a person....most of the time.
Sera Cela
A little bit of crazy
Join date: 15 Sep 2005
Posts: 197
04-30-2006 18:25
From: Introvert Petunia
I understand your point, but not every solution requires drastic changes as in the existing function breaking examples you cited.

Just to throw out a simple example of how this could be done without affecting the structure of the asset DB or gimping LSL. The "grey goo" attacks have a feature that could be exploited to detect and abate them - unlike the legitimate examples you cited above, the grid attacks lack locality of rezzing. One can imagine a meta-process that watches the asset rezzing and looks for patterns of rapid non-local rezzing.

This is similar to the heuristic used to stop TCP SYN flood DoS attacks that were threatening the entire internet by exploiting a flaw in the 1970s TCP protocol designed in an era relatively free of hostile crackers. If a network device sees a sharp spike in SYNs left dangling, they assume that a SYN flood is being mounted and start discarding SYNs from the originating addresses. Thus the response is targeted and kicks in only after detecting a known pattern of attack and affects only the originating machines. True, this also may cause a DoS of legitimate traffic from those addresses, but it is preferable to having all connections fail because some rogue machines are mounting an overload.

Have I proven here (or to myself) that this will fix it? Nope. Does it stand a decent chance of mitigating it? I think so.

Your kinda agreeing with me in a wierd way. Like I said, it is 100% possible for them to do something about this. However they would need a drastic change to the system. People seem to forget that once an item is on a sim the sim handles it, not the asset server. Rezzed items are not tracked by the asset server. Also one sim doesn't keep track of what's on the nextdoor sims. It would be simple to create an object that flies around the grid dropping a single seed object in each sim, wait for a specific time and then start attacking every sim at once. All you would be doing is plugging up one way of doing it, and in return crippling LSL for many many many applications.
Sera Cela
A little bit of crazy
Join date: 15 Sep 2005
Posts: 197
04-30-2006 18:26
From: Luciftias Neurocam
A couple years ago I stumbled across a nice google hack that permitted anyone searching the right combination of numbers to find oodles and oodles of credit card numbers, names, and any other info you might want. Some people may remember this little bug. If I'd been a much worse person, I would have a big screen TV, a really nice car, and some gold capped teeth right now. And lots of fake SL accounts.


But I'm that bad a person....most of the time.

That was the number range scan they implemented. That's been fixed (mostly) now.
Maklin Deckard
Disillusioned
Join date: 9 Apr 2005
Posts: 459
04-30-2006 18:26
From: Sera Cela
It's really easy to bitch and complain when you don't understand the facts. Sure i'm not happy the grid is down, however since I know how the system works and it's vulnerabilities I don't flip out and start calling the lindens incompetent. I get pissed at the moron who released the item, not start demanding LL cripples the game to prevent stuff like this from happening.


You're EXACTLY the kind of CONSUMER businesses love. Always willing to make excuses for them, to be unhappy but not at them. I still remember the time when folks DEMANDED service, not rationalized and defended poor performance.

Down is down, whether they pull the plug or a player causes the crash. If they pull the plug for an exploit, the exploit is their responsibility due to poor programming and QA. If the game goes down because of a player, its their responsibility for not coding to prevent such occurances. Yes, in both cases, the exploiter and the global attacker share responsibility (they found the bug / hole)...but the companies BOTH are to blame for providing the mechanism. Its how they respond to it that counts, and LL just apologizes and goes on like nothing happened...that is why I blame LL.
Sera Cela
A little bit of crazy
Join date: 15 Sep 2005
Posts: 197
04-30-2006 18:39
From: Maklin Deckard
You're EXACTLY the kind of CONSUMER businesses love. Always willing to make excuses for them, to be unhappy but not at them. I still remember the time when folks DEMANDED service, not rationalized and defended poor performance.

Down is down, whether they pull the plug or a player causes the crash. If they pull the plug for an exploit, the exploit is their responsibility due to poor programming and QA. If the game goes down because of a player, its their responsibility for not coding to prevent such occurances. Yes, in both cases, the exploiter and the global attacker share responsibility (they found the bug / hole)...but the companies BOTH are to blame for providing the mechanism. Its how they respond to it that counts, and LL just apologizes and goes on like nothing happened...that is why I blame LL.

Funny... looks like I'm saying they need to keep the downtime to about 10 minutes. But of course, you don't care about the facts. So Rage on!
Zonax Delorean
Registered User
Join date: 5 Jun 2004
Posts: 767
05-01-2006 02:41
From: Luciftias Neurocam
A couple years ago I stumbled across a nice google hack that permitted anyone searching the right combination of numbers to find oodles and oodles of credit card numbers, names, and any other info you might want. Some people may remember this little bug. If I'd been a much worse person, I would have a big screen TV, a really nice car, and some gold capped teeth right now. And lots of fake SL accounts.


That's true, however, if you were such a big scale fraudster, the FBI would mobilize much more powers to find you -- and the IP address you logged in to SL would be a good start.

I don't think a purely 'law enforcement' solution would help, though, but even that has to be tried.

There could be different 'trust levels' for users: basic users could do many things, for example, but maybe not create self replicating object (only in sandboxes). Or maybe they would have stricter limits on self replicating capabilities. Premium users who are 1 month old could have looser limits. If that's not enough, maybe you could 'sign' a special contract or authenticate yourself more with Linden Labs to get 'more trusted' and more rights. Also, scripts made by a non-trusted user would never get trusted rights, even if given to a trusted user.

But this is just one proposal. There are possibly many others that could be implemented, and would help.
Zonax Delorean
Registered User
Join date: 5 Jun 2004
Posts: 767
05-01-2006 02:44
From: Introvert Petunia
This is similar to the heuristic used to stop TCP SYN flood DoS attacks that were threatening the entire internet by exploiting a flaw in the 1970s TCP protocol designed in an era relatively free of hostile crackers. If a network device sees a sharp spike in SYNs left dangling, they assume that a SYN flood is being mounted and start discarding SYNs from the originating addresses.


I think this could make a good solution, too.
Selador Cellardoor
Registered User
Join date: 16 Nov 2003
Posts: 3,082
05-01-2006 03:00
What a strange thread! When it got to the Arab terrorists inworld it became almost surreal.

I think there must be a solution somewhere, but it doesn't lie in the direction of crippling the world. Zonax had a good idea, except that I believe the last idiot, or perhaps the one before last, had been a citizen for several months.

My own feeling is that the Lindens need a reliable system of getting the rl address of new residents. Then they will be in a position to prosecute and if necessary take legal action to recover damages.
_____________________
1 2 3