-
Posts
10683 -
Joined
-
Quote:Technically speaking I think the best bet for future-proofing is going 1366. You get triple-channel memory and access to Core i9. All I've heard about the future roadmap with regard to sockets is that Gulftown will probably be a 1366 part. But since we're in-between processor generations (for Intel) I personally didn't think I was sacrificing much to get an LGA1156 system and at the moment you can get a better bang for buck deal on processor speed. I don't think you can go very wrong either way.I wanted Pyro's system to be as future proof as possible so I went with the new 1156 socket. Plus some of the tech sites
claim there is a problem with the 1366 socket that can potentially leave an air gap with some coolers... Intel won't
confirm (yeah big surprise right?) but their future roadmap looks good for the 1156, there is an i3 coming
soon but it will be budget option and should not affect the 1156's long term viability, so this
looks to be the best choice for performance computing at this time.
Quote:Sidenote: Long before I found the art forums and began posting anything I only came to the forums when I couldn't find
an answer to a question anywhere else... often times there were a lot of people speculating or outright
guessing/arguing with almost no hard numbers to support their claims.
There were however a few who really knew their stuff and took the time to do the work and research to actually KNOW
before they posted. After a time of wading through the tons of nonsense I began to just seek out a handful of
names and wait to see their take on a given subject... right at the top of that short list was Arcanaville
and since I never got a chance to say it before.
Thank you. -
Quote:You can add static routes using it as a gateway even for other local hosts, and it'll work (I'm assuming ICMP redirect is disabled on the appliance).Running it in a VM is certainly a novel idea, though the local network setup might be a little complicated to ensure you're routing through it.
I.e.:
C:\Windows\system32>route add 10.254.255.240 10.254.255.215
OK!
C:\Windows\system32>tracert -d 10.254.255.240
Tracing route to 10.254.255.240 over a maximum of 30 hops
1 <1 ms <1 ms <1 ms 10.254.255.215
2 <1 ms <1 ms <1 ms 10.254.255.240
Trace complete.
C:\Windows\system32>tracert -d 10.254.255.240
Tracing route to 10.254.255.240 over a maximum of 30 hops
1 <1 ms <1 ms <1 ms 10.254.255.215
2 <1 ms <1 ms <1 ms 10.254.255.240
Trace complete.
(I'm showing the trace twice for the benefit of the networking gurus out there that might be worried about getting an ICMP redirect which would override the static route with a collapsed route).
I can use the virtual appliance at 10.254.255.215 running in VMware as a router to get to a local system at 10.254.255.240. And actually, .240 is *another* virtual server running on my system; this should work regardless of whether the target is a virtual or real system connected to the local network, or more router hops away.
The one thing this won't work for directly is if you have the client and server literally loaded on the same workstation, and that computer has only one IP address. You might need to play some very weird networking games to make something like this work. It might make more sense to simply virtualize the server side under VMware to separate the two machines, unless debugging actually requires both client and server to be loaded locally under the same system context.
BaB: if you want to pursue this troubleshooting angle in more detail or your LAN techs have questions, you can just email me. -
Quote:WANem. You should be able to use this to simulate a wide area network's bandwidth and packet loss on your test system. It also comes as a virtual appliance so you do not need to install or configure it (you just need VMware workstation or the free VMware player to run the virtual machine).I remember at PAX a couple of years ago watching and playing our game on their tenuous connections and being dismayed at how frequently attack animations just completely failed to play. Even in situations where there wasn't any good reason why they shouldn't play. Ultimately, it's the server that tells the client what move it should be playing, and I kind of assume that something as simple as a lost packet at just the right moment, or missed tick, can cause the sequencers to reject playing a move. This type of stuff has proven to be nigh-impossible to replicate on a local client/mapserver environment...which is the only environment where I can actually monitor the state machine of both the mapserver and client and see what's happening.
I have my doubts that packet loss can be responsible for at least the issues I was seeing. Packet delay is theoretically possible, but seems unlikely to me. In any case, as long as your test system is fast enough, you should be able to use WANem as a virtual router between the client and server in your test environment with just some minor IP routing magic. -
I wonder how Piaget would have responded to modern experiments that suggest even infants less than six months old are capable of the mental process of deception, which requires at least a rudimentary notion of external mental states. There's almost no way deception could be learned by that time, which strongly suggests a genetic ability to construct at least rudimentary external mental state models and act in response to them.
-
Quote:Incidentally, I meant in the evolutionary sense. Evolution invented the brain mechanics to understand others before it then built upon that to create a sense of self. That doesn't mean one is actually dependent on the other in an actual functioning mind.See I don't completely agree with that sentiment and I tend to sight Aspergers syndrome as an example of why self-awareness precedes awareness of others.
Moreover, the inability to perform second-order reasoning about external minds in various autistic-like disorders including Aspergers seems to go at least to some degree hand in hand with impairment in self-awareness. The more severe the inability to comprehend other minds, the more likely it appears self-awareness is impaired as well. Highly functional Aspergers are not incapable of making mental models of other people, but the further down the scale you go towards mental deficits that impair second-order reasoning, the more likely you start to see self-awareness deficits such as (lack of) embarrassment. -
Quote:Actually, your starting and zoning times wouldn't be bound by disk anymore, but sorry to say no instant zoning with the SSD. One of the first things I did after getting the new system up and running was copy the entire game client to a RAM disk to see what would happen. Honestly, it was only marginally faster than running off my internal drive.1. Kick butt memory chip.
2. Solid State drive. If your game files were on this, your load times would be close to nil.
3. Cooling system.
I believe the bottleneck on zoning is due to the server having to send to the client an initial data bundle for the zone that is larger than the normal data stream it sends during gameplay. This is bandwidth-limited on the server-side (to prevent players with high bandwidth from starving out the other players) and places a limit on how fast you can zone no matter how fast your computer is. It seems to be about eight to ten seconds depending on zone.
On the other hand, *patching* the game client gets very zippy indeed.
PS: my guesses from the pictures were memory, SSD, and heat sink, but I didn't recognize the specific vendors. The memory was obvious, the SSD was obvious, but the heat sink was a lucky guess because I had recently done research on the heat sink changes between 1156 and 1366. -
Quote:A couple years ago, before the slow resistance was added, I did an analysis that attempted to normalize the costs intrinsic in those buffs relative to their cost within the power pools. I came to the conclusion that Quickness, even as a passive, was almost certainly significantly weaker than it ought to be, at a time when Super Reflexes itself was also weaker than it ought to be and therefore couldn't be applying downward balance pressure in Quickness.Even considering the run and fly speed buffs, slow resistance and the fact that it's a passive?
I'm pretty sure that even in SR's current better situation and the added slow resistance I could still prove Quickness has a much lower buff than it ought to have when factoring in the twin requirements that it should contribute to the set in a way that enhances set balance rather than hurts it (in either direction) and it doesn't provide enough buff relative to its cost (which, as a passive, is its power choice cost).
Its worth noting that Hasten, a power pool power with no prerequisites, starts off with no slotting offering +0.23 recharge, averaged over its uptime, compared to the +0.2 that Quickness offers. That's *without slotting Hasten*. Granted, Hasten has an endurance crash penalty, but cost averaged over its unslotted cycle time that cost is only 0.045 eps. That's less than Combat Jumping, whose endurance costs are considered to be practically zero.
That doesn't account for the other buffs in Quickness, but Quickness is already starting off essentially in the hole to a power pool power *before* I slot it. I'm pretty sure if I did the analysis again today, I'd come to the same conclusion. You can't make that sort of argument with analogous QoL powers like, say, Quick Recovery. -
Quote:Its technically part of early or late-stage field testing (in MMOs and especially this one, closed beta is more mid-term field testing and open beta is more late-stage field testing). Any time field testing is used as part of the implementation testing process, there is a way to provide feedback to quality control. Its not technically a component of the quality control process, but its an input to it.I guess it's a question of how you define quality control. I work as a software regression tester\deployment analyst for a financial services company. In my job quality control is something that can be defined and measured. Processes are set up to insure that every new software release is developed, tested and released in a predictable manner with a minimum of mistakes. Following these processes results in the release of a quality product.
In my mind the Training Room test server is not something that fits into that model. The game developers have no way of controlling the testing that goes on in the Training Room even during closed beta. They can put out release notes, specify areas of priority, ask for feedback and data mine but they can't make the players follow a testing regimen or report bugs. While I appreciate the fact that more testing is always better than less testing the Training Room server represents an uncontrolled environment that in some cases provides second hand, possibly incomplete data. As such the testing done on the Training Room server can not be considered part of the quality control process. -
The core parts of the game that subscribers pay for with monthly subscriptions go through more quality control than the optional stuff only some players will pay a small nominal fee for.
-
Quote:Its probably related to the bug I reported to the devs several times with regard to certain powers unrooting when they are executed under certain conditions. The common denominator was executing the attack while the player was moving in some unusual way, which caused the power to skip its animation. Because rooting is tied to the animation, if you skip the animation, you aren't rooted even if the power otherwise executes.I remember one member of the PvP ladder had gotten around rooting down to a science. He put up a video explaining how to do it (I believe using well timed jumps). I'll see if I can find it.
Edit: Here it is
hope that helps
It doesn't have to be superjump, I posted a video of shadow maul and sands of mu (both use the same animation) running completely unrooted without specifically superjumping, and without even needing to get off the ground at all in some cases. Unfortunately, the video upload converted badly, and the video is a bit unclear (although I still have the original and it was quite sharp: don't know what youtube did that time around).
I never found out what caused that bug, nor am I sure if it was ever even corrected. I have suspicions, but none I can prove (and its been a long time since I've looked at this problem specifically). I should try to re-upload the original and see if it comes out better this time. -
Quote:Those two statements are not mutually exclusive. Yes, Quickness + IOs can equal Hasten. In fact, Quickness is a little less than about half the strength of fully slotted Hasten's average buff alone (which is about 0.47). But you can get Hasten if you want even more recharge.Can "Quickness" + various I/Os = Hasten? Or should I take Hasten, as well, for added recharge?
I'd probably wait until much later in the build, though. At 16, a Claws/SR doesn't have very much to hasten in the first place.
As an aside, Quickness is too weak, provably so. I haven't said that in almost a year, so I thought I'd stick that in there. -
-
Estimated duration of mind control ray effects: 20 hours 18 minutes.
-
Actually, its far more common in comics than in the game.
Anything that hits anything else in comic books with what we'd describe as "smashing damage" due to impact (explosions, blunt force impact, etc - but not including non-directional crushing) is many times more likely to knock something back in the comic than if a similar situation occurs in the game, and conversely immunity to knockback is far rarer in comic books than in the game. -
Quote:Sam had it correct: bind KEY target_enemy_nearNot sure of the proper command, but you can change this under the Options Menu as well.
I used to do it all the time (before switching to keymap save files). I tend to map f to target_enemy_near so its easier to hit while my fingers are on the movement keys. Follow is usually exclusive to using movement keys so I remap it to g. (Sometimes r, but r is my standard key to toggle superleaping forward for characters with SJ). -
Quote:Yep, what you're missing is that the percentages are somewhat misleading. They presume there is a 100% that you can hit which means "all."I've noticed this every time I build a new toon and pick up a power like Combat Jumping to help with my defenses. When slotting for this power, I usually six slot it with maybe five Defenses Buffs and maybe one Jump. Sometimes I may max all six slots with the Defensive Buffs. When I then view my percentage of where I am at, it always shows it at around 66% or thereabouts. Unlike other power sets when you view them they will usually show what percent you are getting which is normally much higher in percentage. For my level 50 Blaster, I have her 6 slotted with the Red Fortune set and still can barely reach just under 70% with that. Is there something I am missing here about slotting defensive sets or is it impossible to get any defensive set to reach 99% or 100% capacity?
Actually, when you slot to 66% what you are really slotting to is "1.66x". In other words, you're slotting to 1.66 times higher. There's nothing magical about slotting to 100%. After that comes 101%, then 102%, then 103% - you are basically increasing the dial from 1.66, to 2.00, to 2.01, to 2.02, and so on.
The reason why you can't get much higher than about 60% for some things, and 95% for other things, is because enhancement strength has diminishing return limits.
Just to be sure you understand, defensive powers are quoted in percentage strength: Focused Fighting offers 13.875% defense unslotted. That "percentage" doesn't mean you dodge 13.875% of all attacks: there's a formula which basically says critters have a 50% chance to hit you by default, and 13.875% defense means you *subtract* 13.875 percentage points from the attacker chance to hit: 50% - 13.875% = 36.125%. That's all that "percentage" means. Obviously, the goal is not to hit one hundred percent, but fifty percent, because 50% - 50% = 0. Except the game enforces a floor of 5%, which means defense of 45% is the limit.
Enhancements have a percentage, but those are just multipliers like I mentioned above. A power with 10% defense that has 66% enhancement is not 76% (10% + 66%), its 10 percentage points of defense boosted to 1.66 times the original strength, or 10 * 1.66 = 16.6. That power would then offer 16.6 percentage points of defense.
Its important to try not to get the "%"s mixed up. Think of defense powers as offering points of defense, and enhancements as offering multipliers to those powers.
A good source for more information on these areas is the paragonwiki page on Attack Mechanics. That article covers the topic more generally, accounting for more effects. -
Quote:I believe it can, but that's not relevant to whether computers can be intelligent, because intelligence is not ordinarily considered a binary property, but a range.Whether you believe computer intelligence (which is reached through programing) can ever reach human intelligence; we can argue back and forth.
Quote:My definition of intelligence is sentients (Self awareness). It's possible we have a disagreement on what intelligence is.
My own theory of consciousness is that it is the conclusion to an evolutionary arms race to provide our ancestors with the best possible tools to survive in complex societies. Without getting into too much detail, evidence strongly suggests that awareness of others comes before self-awareness. We first evolve to understand how others will react in order to understand how to interact with them in societies. We build models - simulations if you prefer - of the other people around us. Eventually, that collides with the ability to plan ahead, and instead of just thinking about how others will react to us, we start thinking about how we will react to how they will react to how we will react, by making a model of ourselves and adding it to the process. Eventually, that mental model of ourselves becomes consciousness.
Its our ability to apply intelligence to simulations of our own mental state that I think form the basis of what some would call "general purpose human intelligence." We're not genetically programmed to be general purpose thinkers, but we are genetically programmed to be thinkers of thinkers and that's what allows us to apply our thought processes to any possible problem that our mental model of ourselves discovers. We can think about how we want a mental model of ourselves to be, and then attempt to behave as the model suggests we should, in a feedback loop. In a sense, humans have virtualized minds.
There's something ironically appealing to me about the notion that the solution to the Chinese Room is not that somehow the simulation becomes as good as reality, but rather that actual awareness *is* a simulation also, just a biological one, and there is no "real" awareness in the sense the Chinese Room suggests. -
Quote:Define "malicious." I was serious.It doesn't befit you to resort to malicious arguments, Arcana, especially since you've shown you can do better.
Quote:A person can very much work beyond his or her "original programming,"
A computer does not have intelligence as it can not operate outside the scope of its program.
If you believe humans aren't even programmed in any sense of the word, then that statement is irrelevant. If you believe humans are programmed in some sense of the word, then the statement implies that the property of intelligence is the ability to transcend the laws of physics. In any case, the question generated a useful response: Tokyo claims not to believe human minds are programmed. In that case, the question becomes why do computers have to show an ability to transcend programming to be considered intelligent, when that requirement doesn't apply to humans.
Quote:I'm not of the opinion that any system which can fake intelligence can be considered intelligent. For one thing, humans are idiots and some can believe anything is intelligent. After all, how many people are completely convinced their toaster is talking to them? For another thing, humans are idiots, and even ANOTHER PERSON can't always manage to come across as intelligent, much less a machine. What "seems" intelligent is an inherently flawed test, because it's subjective and because we're predisposed to see faces, patterns, reason and intelligence in places where it doesn't exist. Gods weren't invented for fun, they were invented because people were SURE there was some kind of intelligence behind the sun and sky and the health of their crops.
You can't simultaneously claim to be an expert in cognition, and claim not being capable of judging a Turing Test even in theory. -
Quote:Not really: that's an exaggeration. That can only happen in situations where the spawn density is so high this becomes a problem regardless of initial spawn distribution and regardless of knockback.Which would then have spawns overlapping each other on some maps, leaving them clumped together anyway in large teams. So maybe I don't get the whole 16, but I get 8 of this spawn and 3-4 of the other.
Quote:That's assuming anyone still wanted to play the game if they mobs were running all over the place. As I stated in another KB post, scatter is annoying. KB is annoying. I don't care if it's good mitigation (it's not), it's a PITA. -
Quote:Not necessarily, no. What I'm trying to say is that when the sets were first created, they were created using a formula to set the damage per attack. And then, secondary effects were sprinkled around based on conceptual desires. These were only moderated, not balanced. There was (and is) no rule that even says a set *must* have a foe secondary effect. Believe me: if there was I would have used it long ago to appeal the fact that MA lacks it. They tended not to want to make such sets, but Martial Arts came very close at the start. Alternatively, Energy Blast could have had a hodge-podge of effects like Assault Rifle. And the devs continued that practice after launch: witness Archery, which has at best a +10% accuracy bonus above the standard weapon bonus (which itself is an unbalanced "give it out just because" bonus) and very little foe secondary effects in its attacks.Again we are saying the same thing. It's a side line to the larger point, but the difference between saying "Fire has no secondary effect, so it does more damage" and "Fire's secondary effect is more damage" is a matter of syntax. The larger point stands--if Energy Blast did not have Knockback as a secondary effect, it would have a different secondary effect or do more damage.
There was *no rule* that said that secondary effects must be balanced numerically, there wasn't even a rule that said all attack sets must have foe secondary effects, and in fact to the best of my knowledge from launch to today there has never been an attempt to balance secondary effects across all attack sets. Basically, if the original designers didn't think Knockback was the cool effect energy blast should have, today we'd very likely be Archery with energy damage and no accuracy bonus. -
Quote:I've been having debates surrounding the Chinese Room for over twenty years. It has almost *always* been used, at least outside hypothetical discussions of cognition, to act as a foil to the Turing Test conjecture. The Turing Test conjecture states (essentially) that anything that can pretend to be intelligent to a human being is intelligent. The Chinese Room attempts to argue that a person that doesn't understand Chinese can nevertheless theoretically pass the Turing Test in Chinese without any comprehension of Chinese. Since the person obviously doesn't have any understanding of what he's doing, it suggests that there's no reason to conclude that a computer that passes the Turing Test necessarily does either.No. The concept has nothing to do with the room being intelligent or not. The person is the thing being considered here. The person does not understand chinese and will not understand it, no matter how simple or complex the rules are.
It does not prove the person is unintelligent, just that they do not understand Chinese.
The argument is designed to illustrate the difference between simple mechanical interaction and the human concept of understanding as a form of intelligence. I would certainly not think the Chinese Room argument is a strawman.
The question of 'what is the bridge between neurons and understanding?' is a vital question to our existence, and one we haven't figured out yet.
But that's a strawman for the reason I mention above. Put simply, it focuses on the *person* but the person can't pass the Chinese Turing Test. The room does.
In effect, we considered the computer (the human) without its programming (the book) and said that because a component was provably unintelligent (within the context of the test) the entire system was. That's no different - no different - than saying that since individual neurons aren't intelligent, a bag of them isn't either.
Why can't a human plus a book be intelligent - within the context of the hypothetical - even if neither in isolation is? Its human bias that makes that seem obvious, because it plays on our belief that if a human isn't considered intelligent within the context of the test, no system with the human in it is either.
Searle (the original author of the Chinese Room) himself had this objection to the contention that the room + the human + the book as a system was a mind. He suggests that if the human memorized the book and left the room, the system would be reduced to just a person, and since we already established that the person doesn't understand Chinese, the fact that the "system" can is meaningless, since the system *is* just the person.
Wrong. Just because the person memorized the book, doesn't mean anything. What mattered wasn't the physicality of the book, but its instructional content. Memorizing those instructions didn't destroy them, any more than converting the instructions on a paper card into electrons destroys them. The system is still Human + "Instructions that used to be in the book" which the person is processing. Its really no different than the human tatooing the instructions all over his body, Memento-style. Tatooing on the skin, or memorizing in your brain, has no distinction within the context of the Chinese Room.
One of the problems Searle has is he fails to delve into the nature of simulation verses reality. He dismisses simulations as "not being real" which is technically true, but fails to examine *why* simulations aren't real, which is a question of relativity. But that's a very, very long discussion. -
Fire doesn't do more damage because it has no secondary effect. It actually does exactly the same amount of *base* damage based on recharge as every other set. Its just that its secondary effect *is* damage - specifically DoT. And in an exception-that-proves-the-rule situation, Fire is not required to pay for any of its DoT damage in endurance or recharge costs - because its considered a secondary effect.
-
Quote:The irony is that it proves a person following instructions has no intelligence.Aside from Arcana's postulation, computers do not have intelligence. This is illustrated easily by the Chinese Room thought experiment.
Put a person in a room. The room contains a book filled with instructions detailing inputs and outputs in Chinese symbols. The person in the room receives an input of chinese symbols and then uses the instructions to construct an output of chinese symbols and sends it back out.
The person in the room does not understand Chinese at all, and can only follow instructions.
Computers don't understand, they just function, and hence cannot be truly intelligent.
(currently speaking)
The strawman aspect of that argument has always been that the "lack of understanding" of Chinese is predicated on the belief that the instructions the person is executing are trivially simple. It masks the actual metaphysical problem with discussions of intelligence (and consciousness, which is not the same thing). And that is at what point does mechanics generate intelligence. After all, none of your brain's neurons are individually intelligent. They each function in a predictable way that a machine could emulate. Intelligence is an emergent property of the collection of those neurons. Looking at the Chinese room from a different point of view, if the person in the room was given so complex a set of instructions that the Chinese room could pass the Turing test, who's to say that while the *person* wasn't acting with intelligence or understanding, the room was.
In other words, it relies on bias to make the listener leap to the conclusion "obviously rooms can't be intelligent: either the person in the room is, or nothing is." But that's no different from saying "either a neuron is intelligent, or its not: certainly a white rock with eyes that holds a bunch of them can't be intelligent." -
Quote:Not really. Energy Blast's attacks have the damage that their recharge dictates. No more, no less. Its secondary effect is not in any way considered in that computation.Energy Blast does less damage than it otherwise would, because it has a secondary effect.
Ironically, the only way taking the knockback away would somehow increase the damage of the set is if the devs datamined that all energy blast characters were collectively averaging lower performance - in other words if knockback itself was materially contributing to better levelling performance: in other words, if it was so valuable its value was measurable across the playerbase. In that what-if case, the devs might have considered tweaking the set in ways that ultimately (but not necessarily directly) increased Energy Blast's damage. -