bumped as the original post still applies
should be in rs2 section
also what in the hell is dynamic memory?
Info that is... random?
[quote=“super_, post:21, topic:549”]bumped as the original post still applies
should be in rs2 section
also what in the hell is dynamic memory?[/quote]http://en.wikipedia.org/wiki/Dynamic_memory_allocation
that doesn’t make sense given context
Probably analyzing the client in memory while it executes. Haha this post is shit, but it was written years ago so you’ve gotta go easy on it.
the post isn’t that shit it just shows pplsuqbawlz talking a little out of his ass e.g. “quantem mathematics”
dynamic live analysis isn’t faster than pattern matching bytecode or sourcecode however so that rules out your idea
the interesting part about this thread however is that it offers statistical classification and the KDD cycle as solutions to the updater problem
The interesting thing about this thread is that, if nothing else, it’s painstakingly clear that psb is a narcissist, or egomaniac at the very least. 
Let me reiterate… this is a shit post.
The red flag should go off well before it talks about the speed of an updater instead of reliability. What he calls “assumptions” has its very own field in the world of computer science - machine learning. I can assure you that, in general, a machine learning approach to updating is not “the fastest” type of updater you can write as classifying large datasets (in this case using training data from past client revisions) is an expensive task… which isn’t to say it’s not a reliable approach.
…and by “machine learning” I mean to make a distinction between supervised and unsupervised learning (artificial intelligence)
pplsuqbawlz once sent me the code for “assumptions”… I can assure you it was not based on machine learning. Instead it used static decision trees for classification 
lol!
You’ve earned yourself a spot in my sig along with Mopman. Don’t take it personal.
EDIT: Max sig length is 5 lines… oh well, I liked Mopman’s better, anyway.
It was machine learning in the sense that the static trees were used to “learn” which classes corresponded with code (classification). The trees were not themselves updated with incoming data using a decision tree algorithm, and therefore in that sense it was not really “machine learning”. I called them /static/ for a reason.
P.S. SmartFoxServer is written in Java. 
http://docs2x.smartfoxserver.com/api-docs/javadoc/server/
You’ll notice that distributed are /clients/ in AS3, C#, Objective-C, and Java as well; the server, however, is in Java.
Well, I don’t have Bishop with me to cite anything definitively, but I fail to see how the way you update input for your decision tree influences whether or not you are using decision tree learning. And decision tree learning is a subset of machine learning.
P.S. SmartFoxServer was originally developed for Actionscript.
EDIT: ffs, that’s SmartFoxServer2
There’s a difference between updating input and using input to update and refine your decision tree. In pplsuqbawlz’s case, the decision trees were not automatically “tuned” by machine learning algorithms actively.
On the subject of SmartFoxServer, the original client was developed for ActionScript originally, I think. The server is Java though…
» Server: The server minimum requirements are a 500Mhz CPU with 64Mb of RAM and an operating system capable of running the Java runtime 1.4 such as Windows, Linux, MacOS X (10.4 or higher) Check our product comparison table for a list of features supported .
» Server: The server minimum requirements are a 500Mhz CPU with 64Mb of RAM and an operating system capable of running the Java Runtime 1.5 such as Windows, Linux, MacOS X (10.4 or higher) Check our product comparison table for a list of features supported .
I think I see what you’re saying. Instead of training a decision tree, he just threw together his own decision tree, manually, without using an algorithm like ID3, on attributes (client patterns), to generate an optimal decision tree.
So basically, I have underestimated how shitty this post is.
This is the server-side doc for the original Actionscript version that I supposedly never used in a proxy bot because it doesn’t exist: http://www.smartfoxserver.com/docs/docPages/serverSideApi/
Here’s the system requirements for that version, stating its dependency on the JVM because it was written in Java and just happened to have an ActionScript interface. You chose to use it apparently, despite the fact that “Java is a bad choice”. That’s all my comment meant.
Oh.
In that case, let it be known that my sig’s meaning does not commemorate the day Mopman made an idiotic comment.
Instead, let it commemorate the day that Mopman was a tricky bastard.
And we all lived happily ever after.
lets hug
Does this mean you could visit any website and it would go through their server and not be linked to you? Or how do they have it sent? Doesn’t look like an iframe, does it?
[code]
