viewtopic.php?p=19067#19067
There are a slew of games (from the link above) that can't be playedback by about half of the people that try them -- even with the correct settings and mame version. Thanks to buttermaker and crew, we know this is a mame bug where certain mame games can be recorded on one machine and can be playedback on that machine but will never playback on another kinda of machine that might even be using the same OS. This may have something to do with the new mame_rand() or machines that store time in GMT vs. local time (that's something to look into BM) or something else but it's frustrating.
We know this bug is non existant in 62 and there is an alphamame62. Should we banish recordings with this feature that record with 63 or later and are in this game set? They can easily be determined by submiting a playback request and watching the "same here"'s fly.
I'm assuming this bug will be fixed someday (this usually means a year or so later with playback bugs :) so future recordings will be allowed with a future mame when the bug is fixed.
I wanted to get peoples thoughts before a poll i started.
Playbackabilty Bug Introduced mame63 : prePoll discussion
Moderator: BBH
yeah that's what i'm thinking, but buttermaker says the ini files are both nonexistant, which leads to something in the hardware. It would be SWEET if we could find the differnce then both sets of people can playback both sets of recordings but now Set A can only playback Set A's recordings and Set B etc...
-skito
can you rephrase the above to make it more clear?
Are you proposing MARP only allow pc versions of mame/alphamame of 0.62 or earlier now? ...that all versions since 0.63 have this flaw for many games? Are you proposing all scores submitted using 0.63 and later be removed from MARP?
I wonder if macmame has this same issue cuz many of the inp playback issues you guys seem to have with pc mames versus exact OS used or if mame32 vs mame vs dmame is used we never had for macmame.
I have not had 1 occasion yet where a macmame from anyone else hasn't played back correctly for me...although albeit almost all are for 0.60 and earlier...very few for 0.66...so perhaps I just haven't seen it yet.
Do you have a specific rom set etc. this is the case for for 0.63 and later? I want to try it in macmame 0.66 and my unofficial macmame 0.70 build to see if it's an issue with macmame also.
I can make a few sample inps and get other macmame players to try and play them back and see if they are all successful.
Are you proposing MARP only allow pc versions of mame/alphamame of 0.62 or earlier now? ...that all versions since 0.63 have this flaw for many games? Are you proposing all scores submitted using 0.63 and later be removed from MARP?
I wonder if macmame has this same issue cuz many of the inp playback issues you guys seem to have with pc mames versus exact OS used or if mame32 vs mame vs dmame is used we never had for macmame.
I have not had 1 occasion yet where a macmame from anyone else hasn't played back correctly for me...although albeit almost all are for 0.60 and earlier...very few for 0.66...so perhaps I just haven't seen it yet.
Do you have a specific rom set etc. this is the case for for 0.63 and later? I want to try it in macmame 0.66 and my unofficial macmame 0.70 build to see if it's an issue with macmame also.
I can make a few sample inps and get other macmame players to try and play them back and see if they are all successful.
I think we should remove inps for these games that have been recorded with 63 or later. I would dare say macmame63+ too since if it's happening in the pc world i'll bet it'll happen in the mac world, but isn't there only macmame60 out? AND the problem is NOT that the user can't playback their own recordings (this always works) ; the problem is that a group of other people with certain configs can't playback that persons recordings. I thought i said that in the original post though.
It affects a subset of games (NOT ALL GAMES) the current list can be viewed by the ABOVE link, here's a cut and paste of the current state of that list.
Mortal Komat and gang http://marp.retrogames.com/index.cgi?mo ... rt&tourn=0
metalclash http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
snakepit http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
cloak http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
raident http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
zwackery http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
paradise http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
and the nba gang http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
It affects a subset of games (NOT ALL GAMES) the current list can be viewed by the ABOVE link, here's a cut and paste of the current state of that list.
Mortal Komat and gang http://marp.retrogames.com/index.cgi?mo ... rt&tourn=0
metalclash http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
snakepit http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
cloak http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
raident http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
zwackery http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
paradise http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
and the nba gang http://marp.retrogames.com/index.cgi?mo ... 00&tourn=0
-skito
Ok...it seems an extreme action to negate all of the submissions using 0.63 or later just because 10 or so games have this issue.
if that's all out of the nearly 4000 rom sets...then you could add special rules for each of those particular games that 0.62 or earlier must be used.
I think given the alphamame blocking thing some games need those special rules added already to make it clear for players.
For example..hang-on...use version 0.36-final as a special rule for now. This prevents someone from playing it in alphamame despite the cycle drawing issue...and block down the others.
...so why not just break that down where for those games you list only 0.62 or earlier can be used. and if there is an alphamame 0.62...that alphamame blocking still applies if that version is ok for the game.
I'll have to try a test with cloak and dagger...that's on your above list.
The thing is in the past I have seen numerous reports of certain inps only playing back on a pc running win98...or XP etc. that are OS specific. This has never happened with macmame.
if the 0.63 and later issue for pc mames is similar to what caused the issue of which OS was used for older versions...then it's likely this isn't an issue for macmame...cuz we never had the first issue.
However, if the code base has changed in 0.66 or the unofficial 0.70 so inps are same format as PC mames...then perhaps it has been introduced now.
Like you said...you aren't sure exactly what is causing the problems. It can easily be something that's pc specific.
Anyway...that's how I feel about it...a case-by-case basis for those games...special rules to only allow 0.62 or older for those particular games...instead of applying that to all nearly 4000 rom sets.
if that's all out of the nearly 4000 rom sets...then you could add special rules for each of those particular games that 0.62 or earlier must be used.
I think given the alphamame blocking thing some games need those special rules added already to make it clear for players.
For example..hang-on...use version 0.36-final as a special rule for now. This prevents someone from playing it in alphamame despite the cycle drawing issue...and block down the others.
...so why not just break that down where for those games you list only 0.62 or earlier can be used. and if there is an alphamame 0.62...that alphamame blocking still applies if that version is ok for the game.
I'll have to try a test with cloak and dagger...that's on your above list.
The thing is in the past I have seen numerous reports of certain inps only playing back on a pc running win98...or XP etc. that are OS specific. This has never happened with macmame.
if the 0.63 and later issue for pc mames is similar to what caused the issue of which OS was used for older versions...then it's likely this isn't an issue for macmame...cuz we never had the first issue.
However, if the code base has changed in 0.66 or the unofficial 0.70 so inps are same format as PC mames...then perhaps it has been introduced now.
Like you said...you aren't sure exactly what is causing the problems. It can easily be something that's pc specific.
Anyway...that's how I feel about it...a case-by-case basis for those games...special rules to only allow 0.62 or older for those particular games...instead of applying that to all nearly 4000 rom sets.
-
Buttermaker
- MARP Seer
- Posts: 788
- Joined: Sun May 19, 2002 9:06 am
Buttermaker...that's exactly why I asked for clarification. If Chad is only referring to those specific games...fine..whatever...then it's just making a special rule for those particular games.
His first post didn't state it that clearly.
..so I wasn't sure what to even discuss. I covered both bases...
His first post didn't state it that clearly.
..so I wasn't sure what to even discuss. I covered both bases...
Re: Playbackabilty Bug Introduced mame63 : prePoll discussio
Does this mean half of the playbackers succeed?Chad wrote:viewtopic.php?p=19067#19067
There are a slew of games (from the link above) that can't be playedback by about half of the people that try them -- even with the correct settings and mame version.
If so, I think it's enough. There has allways been inps that other can playback and other cannot, reason unknown.
If someone else, other than recorder, can playback(confirmed)...Imo recording is valid then. If only recorder can playback, not valid.
May have something to do with different type of processor(AMD, Intel) etc...OS not allways the cause, for sure!
How you know this is a bug at those Mame vers?
exactly TJT...I always felt the MARP policy for PC mame inps that if 3 people can't get it to play back correctly...then it's zeroed out is a tad flawed....cuz you seem to do that even if someone else is able to get it to correctly playback.
You should only do that when at least 3 can't get it to playback and no one else has been able to get it to playback. if someone got it to playback successfully, then it's confirmed...even if many others can't play it back.
You should only do that when at least 3 can't get it to playback and no one else has been able to get it to playback. if someone got it to playback successfully, then it's confirmed...even if many others can't play it back.
Remove those recordings? Are you kidding me? Just because some people can't get an inp to work, that is no reason to remove it. I can't get tons of confirmed inps to playback at all. Someone at some point got them to work, otherwise they would never have been confirmed in the first place. That is fine with me. I am in windows XP and shouldn't have such lofty expectations that every single inp at marp playback properly for me.
I know several people who can playback my cloak inp perfectly. Just because it won't playback on some systems is no reason to throw it out.
By your logic since the PC users can't plyaback MAC inps and vice versa, ALL marp inps should be removed. The important thing here is that some people can get the inps to playback properly.
If you want to go to a universal everyone had better be able to play it or it gets thrown out' standard, I will quote from BBH's favorite game of all time:
Good Luck! You'll need it!
I know several people who can playback my cloak inp perfectly. Just because it won't playback on some systems is no reason to throw it out.
By your logic since the PC users can't plyaback MAC inps and vice versa, ALL marp inps should be removed. The important thing here is that some people can get the inps to playback properly.
If you want to go to a universal everyone had better be able to play it or it gets thrown out' standard, I will quote from BBH's favorite game of all time:
Good Luck! You'll need it!
Last edited by zlk on Sat Jun 28, 2003 2:02 am, edited 1 time in total.
That's not the whole rule, if three people (including the author) can get it to playback it gets unzeroed. Which is why i didn't start a poll right away. Most of these recordings made by the buggy mame are clear from the "flawed" playbackabilty rule, because half of the people CAN get it to playabck. But i just thought it out of consideration that people who know they are recording in a mame that has playback bugs might care, if all they had to do is recording in a mame a few versions earlier their recording would be viewable by all.LN2 wrote:exactly TJT...I always felt the MARP policy for PC mame inps that if 3 people can't get it to play back correctly...then it's zeroed out is a tad flawed....cuz you seem to do that even if someone else is able to get it to correctly playback.
But it's clear points and rank are all people care about, i think i'm tired of trying to maintain a usable site, when all people care about their points wether they playback or not to anyone else... forget it.
-skito
Chad, personally I don't care about the points. It is just the fact that I spent a lot of time making that cloak inp. Many people have gotten it to work. It is even confirmed. I feel bad that not everyone can get it to playback. If you want to have a vote, that is fine by me. If you want to zero it, I will be annoyed but it is not the end of the world. I just hope you can understand where I am coming from.
hehe....oh, I agree inps should playback...cuz that is the first objective of MARP..a library of submitted inps for others to download and view and enjoy and perhaps improve their own play from watching others.
It just seems to me if you know you have an inp for a game that is finicky about playing back correctly...and perhaps even OS-dependent...that someone else saying they can't play it back correctly that doesn't match what the player was using shouldn't count toward the "3" that can't play it back correctly IMHO.
For example I have seen some post in the playability forum they couldn't get this or that to playback correctly..and it count toward the 3...when it's likely an issue of audio settings or something similar they overlooked.
The fact you have once 3 correctly view it it gets unzeroed is great...but truthfully, once a score is zeroed out are others still attempting to play it back to get i reinstated? I doubt it...so you never even have a chance of getting 3 to play it back successfully.
if you have inps that only playback with those using Win98 though perhaps that's still sort of a good thing cuz you likely have some others still using Win98 or similar that would be able to view that inp correctly...but all using Win2k or XP can't.
Luckily I haven't seen 1 case of anything like that in macmame. All inps successfully play back using different mac CPUs and OS X or 9 or 8, etc.
Maybe in some cases it's dependent on the Direct3D stuff in pc mames?!?!? I saw that one thread where use of -d3d allowed you to playback an inp you couldn't otherwise...or not using it when recording allowed inp that others could playback fine when they couldn't when you used d3d. None of that stuff applies to macmame obviously.
It just seems to me if you know you have an inp for a game that is finicky about playing back correctly...and perhaps even OS-dependent...that someone else saying they can't play it back correctly that doesn't match what the player was using shouldn't count toward the "3" that can't play it back correctly IMHO.
For example I have seen some post in the playability forum they couldn't get this or that to playback correctly..and it count toward the 3...when it's likely an issue of audio settings or something similar they overlooked.
The fact you have once 3 correctly view it it gets unzeroed is great...but truthfully, once a score is zeroed out are others still attempting to play it back to get i reinstated? I doubt it...so you never even have a chance of getting 3 to play it back successfully.
if you have inps that only playback with those using Win98 though perhaps that's still sort of a good thing cuz you likely have some others still using Win98 or similar that would be able to view that inp correctly...but all using Win2k or XP can't.
Luckily I haven't seen 1 case of anything like that in macmame. All inps successfully play back using different mac CPUs and OS X or 9 or 8, etc.
Maybe in some cases it's dependent on the Direct3D stuff in pc mames?!?!? I saw that one thread where use of -d3d allowed you to playback an inp you couldn't otherwise...or not using it when recording allowed inp that others could playback fine when they couldn't when you used d3d. None of that stuff applies to macmame obviously.