Technical solution to eliminate desync in single-player sessions

"

syrioforel wrote:

"

ScrotieMcB wrote:

You're not really serious, are you? I can barely code and this is a joke. This doesn't even need a hack. It literally hacks itself.

This kind of attitude pisses me off so fucking much. If a piece of code or mathematics or whatever is clearly wrong, you should be able to show it's wrong with a counterexample. Just do it, and the conversation ends, and you've clearly proven yourself correct.

If instead the provided argument does not have enough rigor for a counterexample to be constructed (maybe it's too handwavy or isn't specific enough), then state that this is the case and why.

If it's written like shit and unreadable, say why it's written like shit and unreadable. Make some intelligent point to further the discussion, or at least have the good sense to step out of the way.

qwave clearly is, at least the very least, putting thought into his posts. Of course he's being serious. He's spending time trying to develop and communicate an idea, and you're trivializing his effort. Regardless of whether or not his method is sound, you're being rude and unduly disrespectftul here.

As someone who has written several long and well-developed posts on this forum, who has clearly made an attempt to try to improve the game and community, I'd have hoped you would know that responding in such a dismissive manner accomplishes nothing.

I appreciate some good trolling when the poster clearly deserves it, but this shit has to stop when the intent of the post is clearly a constructive one.

Yes, you've made other intelligent counterargument in this thread, and I am taking your post out of context. However, it does nothing but add stupidity to what at least started off as a constructive intelligent thread; one that yielded interesting discussion and responses from GGG staff.

</rant>

EDIT: this was written before your most recent post.

EDIT2: as well as your edit of the quoted post.

"

qwave wrote:

Therefore, the timestamp and delta have to be consistent, otherwise the server will identify a hack. Signing the timestamp means that the client cannot 'buffer' packets. Because the actions must coincide with the signed timestamp.

"

And where does latency play into this? I mean, if someone has a lag spike, their cheaters?

"

qwave wrote:

"

And where does latency play into this? I mean, if someone has a lag spike, their cheaters?

Latency doesn't effect it. Why would it? The server only cares that the delta is consistent with the timestamp, signature, and seed.


ScrotieMcB: It has all gone so far over your head. Please understand what the timestamp is used for: delta

The server can do anything it wants to validate the timestamp, but the point is, the server can validate the actions that the client performs from the base timestamp against the delta.

"

http://pastebin.com/raw.php?i=t2Z9K6hh

"

syrioforel wrote:

This kind of attitude pisses me off so fucking much. If a piece of code or mathematics or whatever is clearly wrong, you should be able to show it's wrong with a counterexample. Just do it, and the conversation ends, and you've clearly proven yourself correct.

"

It was the word buffer that got me. So if the client was sending packets, and I 'held them' for a few seconds (after they are created of course) and then sent them on their way (without tampering with them), the server would accept them?

"

Everything you would need to hack it would be sitting right there in the client for you to copy into your hack program.

"

qwave wrote:

"

It was the word buffer that got me. So if the client was sending packets, and I 'held them' for a few seconds (after they are created of course) and then sent them on their way (without tampering with them), the server would accept them?

No, you have to sign the packets with a timestamp. The timestamp has to match the time that you performed the actions. A delta is used to make each timestamp 'relative' to the previous one so that your CPU clock doesn't impact it.

"

qwave wrote:

"

It was the word buffer that got me. So if the client was sending packets, and I 'held them' for a few seconds (after they are created of course) and then sent them on their way (without tampering with them), the server would accept them?

No, you have to sign the packets with a timestamp. The timestamp has to match the time that you performed the actions. A delta is used to make each timestamp 'relative' to the previous one so that your CPU clock doesn't impact it.