forums.marpirc.net

http://marp.retrogames.com/inp/rcl_atet ... n35tg3.zip

Well... people were complaining about Negative 1's breakout recordings.. then it went to Roncli and Ms Pacman scores..... and how there were requests for them to do something with TGMame or Alpha Mame....

I watched this whole thing... no frameskipping... not only does it playback 100% [and well over round 135, that's for damn sure]... but I couldn't detect any flux of the game speed or FPS...

Perhaps this is the beginning of some real proof to humble those who naysay?

Kelly
-Who's floored.... if you know the patterns and can adjust to the speed... it's damn well near easy to get high scores!

Uhhh... since when have you been able to read the recorded speed in a TGMAME inp?

Kelly might be talking about what speed the recording "playsback", yeah i get 100% on it too when playing it back, that says nothing about what the true recorded speed is. Incidentally the recorded speed reported in the inp averages 100% with no pauses detected and you need a program to see that. The reported speed distribution looks like he used autoframeskip on a computer that isn't faster than 500mhz (a guess) . With alpha mame you can tell the speed while playing it back. m35tg3 is of course a bout a few hundred thousand times less secure than alpha mame.

Okay.. so it was a bit misinterpreted.. but I am curious Skito... in what regards is it less secure... [just merely curious as to this so as I can learn and better understand]

well as far as encryption goes you can be quite secure with 2048 bit encryption where the private (inp) bits are completley massaged into the encoded data making it quite impossible for one to hack the encoded message or you can be trivially secure like hexencoding a binary file where someone can easily deduce the encoding (get the binary file from the hex file and vice versa just by looking at the encoded file and not knowing the method used to encode it). TgMame is way to the less secure side of that scale.

Thanks for the information... but it makes me wonder why the TG Mame wouldn't be created with a bit more security... perhaps Barry would be willing to make the changes needed to have his fine AlphaMame become the TG used Mame?

What do you think Barry? [Note: I've only had a chance to look around it a bit since my time has been quite little]

The main problem with strong encryption is that isn't compressable. A perfect Pac-man recording wouldn't be much fun if it were a 50MB download.

mahlemiut wrote:The main problem with strong encryption is that isn't compressable. A perfect Pac-man recording wouldn't be much fun if it were a 50MB download.

50 meg? that is not much! Also think about it.. it is a perfect pacman recording! I bet both you, BBH and Benjos would be glad to download it over a 56k modem and spend a few hour watching it without f10! *joking*

Ok seriously... 50 meg!!! you really mean that they will be THAT big because the heavy encryption? Why?

QRS

qrs, think about it some... inp files are 90% zeros... all those consecutive zeros in a row all over the place get compressed big time.

When that file gets encrypted, you will have almost random looking bytes in the data so instead of 50 zeros in a row it might be 30 different characters in those 50 bytes....not able to compress nearly as well.

Apply that to the entire inp and you likely do end up with a file that only compresses like 10% at the most unless the encryption scheme has it where there are still lots of the same character repeated...but likely isn't like that.

I have posted what i think would be another approach and would allow the compression of the inp etc. as well as free sharing of the inp as we have it now at

viewtopic.php?p=15459#15459

I guess for these kinds of steps to be taken with alphamame etc. many must have hacked the inp files in the past to submit faked games? That's sad.

Ahh ok. Now I understand why. Thanks LN2.

Add to all this that the more complex an encyption algorithm gets (complex does not necessarily mean good), the more tendancy it has to munch on your CPU time for playing the game. Which in turn would impact on the program's usefulness in competitions featuring lots of slower machines.

Agreed Kelly

There are lots of guides that show you how to utilize patterns in the Pacman games that practically let you play forever or at least till the game freezes. If you have the time and the iron man will, you can get huge scores on these games with relative ease.

Dax, you can get in the ballpark...however, in pacman since you use that as an example, there is a huge difference between a 3.3+ million score and a score of only like 3.1 or 3.2 million. Only a handful of players have reached 3.3+ million. I submitted a score of this a couple weeks ago. I have submitted a couple others that would have been more than 3.3 million if I had reached the split-screen. The difference there is points from eating the monsters on the early boards where they turn blue.

A great player will only miss a few monsters, if any at all...and a much more advanced understanding of the game and how the monsters move around versus just blindly running a 9th key pattern.

Yes, anyone can find or design their own 9th key pattern without too much work. However, I found even after making a decent 9th key pattern I have only made it to the split-screen once..and then was on my last man. It's a lot harder than it initially seems running a 9th key pattern...for 235 boards and about 4 hours worth. At least it's hard in mame using a keyboard. You mess up here and there and end up having to wing-it to get out of that board. Often patterns are where if you mess up in certain spots you are almost guaranteed to lose a life.

I actually found teaching myself how to group and eat all the monsters over about a 3-4 week period easier than running 9th key patterns to reach the split-screen. I have had more perfect monster-eat games(eating all possible monsters) than I have games where I have reached the split-screen.