forums.marpirc.net

I just see the score non alpha made by Lordgaz is blocked by the one made by Diabolik on Alphamame.

http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0

But Lordgaz post his score before the Diabolik's one (11:23:17 for Lordgaz and 16:54:10 for Diabolik on the same day). So, why his score is blocked ? It's not logical, captain

Francois

first it's not "really" blocked yet since the score hasn't been changed. Though here is the question (perhaps ambiguously implied in the Blocking vote):

What to do with regular mame recordings (Lord Gaz made one) that have Already been made AFTER 3-3-3, and then someone comes along (diabolik) and makes a worse alphamame recording? This is the case for spaceatt.

Should the regular mame score be changed (blocked) because an alphamame recording now exists? Or should we go against the vote and NOT block regular mame recordings after 3-3-3 if they are submited before the first alphamame recording?

I am all for first way because it follows along with the blocking vote text "all recordings made after 3-3-3 will be blocked by alphamame recordings". But want to know what people think...

ok, I looked at that spaceatt score listings and don't see the result of this blocking. His score is still listed as 1st place...with 100 leaderboard pts awarded for it.

...so what was changed other than the description blocking note?

I figured it would still show the actual score of the game..but would be listed as 3rd with the 15% per place drop in the assigned leaderboard pts or something like that.\

so tar would still be in 1st with 100 pts.

diabolik in 2nd with percentage of score pts minus 15%....and then how would the pts for Lordgaz go? Would be just be 100-15%==85 -15%==whatever?

or would it actually start with that so called blocked "score" number?

I think whatever it's doing needs to be explained somewhere so we know what's going on there exactly.

Also, will this be specific for just the leaderboard pts? ie. if you look at the top 3 page will that still show the top 3 actual scores for the games? or will the blocking screw that up also? I would prefer the top 3 mean the top 3 actual scores for each game....versus some being real scores and others a "fake" score from the blocking.

Maybe that is what it's already doing or will do if the code is still being worked on...but can't tell.

First, it's not automatically implemented yet, this will take a while. The editors will manually be changing scores on a case by case basis if they should be blocked.

Here's what it'll do following the "block regular dos/win recordings made after 3-3-3 from beaing alphamame recordings.": what is proposed to zwaxy for site implimentation.

(1) When someone attempts to upload an blockableMame version (dos win regularmame), and a blockingMame (alphamame) recording exists with a score better or equal to than the about to be uploaded score, THEN don't allow the upload and give an error message explaining "you must record with a secure Alphamame Version to beat this secure game's score".

(2) When someone uploads an blockingMame(alphamame) recording, AND there are any blockableMame (dos/win) recordings that have a score with is greater than or equal to the current recording being uploaded AND which were uploaded after the 3-3-3 THEN change each such blockable recording's score to the new alphamame upload's score minus 1, and email the uploader of the unsecured score telling them that their score has been beaten.

I was considering zeroing the scores which would keep the placings a little more normalized, but decided to just change the score to the maximum amount allowable by the blocking vote (the lowest alpha mame recording's score minus 1) so they can still beat everyone else below the alphamame score.

***EDIT*** we are also considering doing the automated blocking ONLY when the alphamame recording in question has been CONFIRMED. since someone could upload an alphamame recording that really isn't one and change a lot of scores unnecesarily...

Chad, if I follow your first option it means no one can upload a regular mame inp for any game where someone post a score with alphamame.

I understood from vote that since 3-3-3 you can not upload a score (regular mame) greater than a score made with alphamame. If Lordgaz posted his score prior to Diabolik's one his score musn't be canceled/zeroed/whatever u want to do with his score.

Think about that : if only one score made with alphamame is enougth to cancel scores made with regular mame it means that I (or anybody else) just have to post a low score for each game to take the 1st place. It seems unfair !

That's an incorrect assumption. With the blocking rule, anyone can upload Tons of regular mame submissions as long as they are LOWER score than the alphamame recordings. (that's in the rule above)

And no you can't steal 1rst place away from any game just with an alphamame recording, because all of the recordings grandfathered before 3-3-3 can NOT be blocked.

It is true for new games and future recordings, since no recording with new games can be before 3-3-3, but alphamame is HERE and everyone should USE it for new games...

Chad wrote:That's an incorrect assumption. With the blocking rule, anyone can upload Tons of regular mame submissions as long as they are LOWER score than the alphamame recordings. (that's in the rule above)

That's what I understood, sorry if my words seems showing something else.

Chad wrote:And no you can't steal 1rst place away from any game just with an alphamame recording, because all of the recordings grandfathered before 3-3-3 can NOT be blocked.

Oki sounds fine

Chad wrote:It is true for future recordings, since no recording with new games can be before 3-3-3 with NEW games, but alphamame is HERE and everyone should USE it for new games...

I agree I use only alphamame and I'm pro. But until Barry release a 36 final based alphamame this could stop some players uploading cause of low performances since mame 60.

cool... I'm hoping there's no loop holes...

piot wrote:I agree I use only alphamame and I'm pro. But until Barry release a 36 final based alphamame this could stop some players uploading cause of low performances since mame 60.

right, but these players with obviously slow computers [if differeing mame versions make a difference between being regulation speed and not] don't have to submit to marp every game they play. They can practice and practice and when mame is fixed or barry releases a 36 version, then they can show off how much they have improved.

Chad wrote:They can practice and practice and when mame is fixed or barry releases a 36 version, then they can show off how much they have improved.

You're right !
I play most of the games in contests (ie with a deadline) that with I'm talking about no submition for slow computer owners. But of course for record submition you have all your life and practice for a while could be an advantage for those guys.

Well It seems we are OK now you have explained your thoughs.

btw, back to the lordgaz score, i've notified him by email instead of changing the scores right away [which is really what i could do] but since this is sort of a grace period the [vote is new] i'll give people ample time to resubmit with alphamame. I'll ask barry to put a news item about the passing of the vote and a louder pointer to use of the alphamame to preserve leaderboard points for new recordings.

Chad wrote:I was considering zeroing the scores which would keep the placings a little more normalized, but decided to just change the score to the maximum amount allowable by the blocking vote (the lowest alpha mame recording's score minus 1) so they can still beat everyone else below the alphamame score.

There is a flaw in this approach.

Any player could decide to essentially screw other players. If I saw new scores for some game using a regular mame. Someone else could just submit an alphamame inp totally sucking at the game intentionally getting a really really low score. This would reduce that regular mame score to their score minus 1 point according to your above...removing all of the leaderboard pts they were awarded for that.

Otherwise you end up where you might score 25 million with a regular mame...so then someone submits a score where they kill 1 enemy and then kill off their men just to essentially "zero" out your score...even if a totally bogus play of the game etc it would accomplish that...even if 10th place etc. Maybe instead of zeroing it out you make it the alphamame score minus 1 if within 25% or 50% of the regular mame score...if not within that percentage of the score then no effect. This would prevent others from just uploading totally nonskilled bogus inp of gameplay just to zero out regular mame scores...At least then if you had scored 25 million an alphamame player would have to get 50% or 75% of your score to overtake it and knock yours 1 point below them....and also perhaps if just for 1st place..or the top 3 places..or something like that.

I think that would be a strong enough effect and deterant still for those using Windows to use alphamame.

Just brain storming a little bit...you don't want this alphamame thing to have any methods of being abused....which the above would be IMHO.

As far as your decrypting tool thoughts...perhaps..but it seems those same hackers can already hack up alphamame the same way and extract that routine...just would take them a little longer to separate and figure out exactly what part is that.

Maybe that could be a function within alphamame even? Then it's all mixed in with the code is it currently is...no difference as far as hacking..but would allow extraction of the original inp if asked for by others or wanted by others.

yeah, maybe if the difference is great enough and the alphamame score is lame enough we might just ask the alpha mame person to at least make an effort twoards blocking.

But really we're trying to get the great scorers to use alphamame, and lowering their regular mame scores is a convincing way to do it, btw we tried asking nicely doesn't always seem to work :p.

I'd rather see someone put up their best effort even though it's a lame score in alphamame than someone get away with an insane score while cheating undetected.

oh, I totally agree...and not really many cases you could even do that currently...3-6 months down the road of many are still using non-alphamames for Windows then I think it's a possibility a player could decide to upload an inp just to zero out a nonalphamame score.

As long as players are warned of that possibly happening when they submit with regular mame then they get what's coming to them I guess...but a warning to that effect might be quite "effective". hehe

No one would want a great score to be zeroed out by anyone just uploading a minimal score with alphamame not even trying to play the game for score.

oh..while I'm posting(hehe)....I am guessing you guys have some PC utils to work with inp files for stuff like autofire detection etc. from what I have previously read at the forums. Is any of that open sourced?

It would be cool to port something like that to the mac so macmame inps can be tested for autofire etc.

I can easily see some macmame players not knowing autofire or run like hell cheat in track & field etc. aren't allowed...so a util to easily check for that or how to check for that would be nice.

I watched and confirmed a shooter with macmame a couple nights ago. It was obvious autofire wasn't used but it made me wonder if I had one that I suspected what could I do to confirm that is the case?

the program you speak of is called analinp and the source contains the decryption of m35tg3 (and easily deducable encryption) which since it's almost trivial to figure out anyway it's not that special, it also contains a ref to the alphamame decryption call linked in at compile time.

It's been pretty much accepted that this program be kept underwraps because firstly it might make it easier to elude the autofire detection if autofire desiring users understand how it works and secondly, it's a pain in the ass to support, i.e. it's sometimes difficult to interpret the numbers for slowdown detection, even though it does have an YES/NO printout that is a pretty good indicator if an autofire mechanism is used.

Once we have a nice support document for it we might consider letting it out but the consensus is for the time being that we should keep it private. It might be able to analyze mac inps since it can do any type of regular mame version (including old mame version with out headers), but my guess is that it'll need some work to get all forms of macmame inps to work if they have freaky frame sizes. if you have a macmame inp in question just send it over and i'll give you the results of a run with it.