hummm,i believe the minigames will go for real time,so you won't be able to pause trick much,right?
TAS i'm interested:
Megaman series, specially the RPGs! Where is the mmbn1 all chips TAS we deserve? Where is the Command Mission TAS?
i'm slowly moving away from TASing fighting games for speed, maybe it's time to start finding some entertainment value in TASing.
hummm,i believe the minigames will go for real time,so you won't be able to pause trick much,right?
Someone on Youtube made a similar comment. But I don't get it. Do you really think story mode TAS will have minigames with 3 minute pausing? If you haven't realized it by now, story mode TAS will not have any minigames at all (if all goes well anyway), except for Frightmare. So all is fine. :)
Piece Out minigame seems pretty tough to manipulate. Not sure I will do it.
Later Skater probably requires me to look at speed addresses, and I have no clue how to set up cheat engine and lua.
i'm not really sure what is possible or not,happy to hea... read it.
TAS i'm interested:
Megaman series, specially the RPGs! Where is the mmbn1 all chips TAS we deserve? Where is the Command Mission TAS?
i'm slowly moving away from TASing fighting games for speed, maybe it's time to start finding some entertainment value in TASing.
Luckily I have not gotten burned from this game yet, so here is a small update.
- The run has been restarted on Japanese to save time by faster text. Saves 8.2 seconds from power-on to fading into the first map. Will save up to 1 minute or so in the whole run.
- People here, on my Youtube and on my Twitter all want Hard difficulty to be used. So the run will be Hard, period.
- Unlike the English WIP, I have now researched the 6 maps and come to the conclusion:
Toy Sea Future Rainbow Pirate can be cleared the fastest, but Rainbow is risky to manipulate (might take too long)
Pirate can work as an alternative for Rainbow Pirate is slightly faster than Rainbow
You don't want to miss out on Future Toy Sea because they are good. Cake is terrible and should be avoided.
I have found a system start time that yields a convenient RNG seed. Took me like 2 days of testing... :
Mon, 01 Jan 2001 00:01:27 GMT (d7 c8 4f 3a at 0x81 in the DTM) and confirming A press at Input: 8068 yields Rainbow Sea Future Toy.
I have yet to find a system start time that yields RNG seed for Sea Future Toy Pirate without delay frames.
- As for why Rainbow is risky to manipulate: DK has to be in the exact right space for Rainbow which I can achieve only by adding delay frames until it works. There are 18 possible spaces, so I hope it can be manipulated quick enough. Then, by waiting on Minibowser dialogue, I have to manipulate Piranha plant on the exact right red space (the bottom one) and a Hammerbro on nearby space after that. This might take too long but I'll test it later.
I started a new run and got until the 2nd dream. But now I have been struggling for a whole week to get the 1st triple dice roll. It just doesn't want to happen regardless what I try. So maybe the run is dead and I have to start over..
Edit: Ok got a 964 dice roll within 1.5 sec of gameplay finally... only took 8 days of trying, jesus...
Edit: Ok got a 964 dice roll within 1.5 sec of gameplay finally... only took 8 days of trying, jesus...
Actually, it takes about 2.5 seconds which may be considered bad. I have checked about 10,000 possibilities by trial and error and only 3 are the triple dice roll I want. Then I found out that I made a small mistake - not making sure that a certain space has no field effect so I can't travel on it. This wastes about 1.3 seconds additionally.
Basicly, all triple dice rolls I ever manipulated were really easy but this particular one turns out to be a pain in the ass. I'm going to have to check thousands more possibilities so I can save those 1.3 seconds back.
When I'm done with the 2nd dream, I'll post an unlisted WIP video.
Edit: I finished more testing. I don't think I will be able to get the best case scenario. It really sucks... It's not very motivating to take an unavoidable 2 second time loss. Not sure if to restart the run, but that would mean all my effort so far was for nothing.
I have not been feeling motivated to continue on my WIP which was cursed with horrible RNG. I lose about 3 seconds, most of which is because I can't land on the space right after the red field. Instead, I have to go through an intersection. It looks bad and I can't bring myself to accept it to be in the run.
Dead WIP: http://tasvideos.org/userfiles/info/31881149986154958
(J version, Dolphin 4.0-9215)
The next course of action would be to either just accept the bad RNG anyway (which I can't see myself doing) or restart and possibly try to manipulate Future Dream to be first (because it requires a lot of RNG to go in my favor) so we can have it out of the way asap.
Welp @ 50+ hours of work lost.
Joined: 1/16/2008
Posts: 358
Location: The Netherlands
Oh that's really unfortunate :(
I think this is another run that could/would greatly benefit (and in this case be saved) from lua scripting :X
Anyway thanks for sharing the WIP... and I hope you find motivation again anyway :)
Yes but the luascript would have to be made. A script that tells what future RNG does what isn't easy to make. I have been doing pure trial and error so far and it did work until that one case where it didn't.
Joined: 12/1/2012
Posts: 103
Location: New Donk City
Been looking into RNG in Mario Party 5; found what appears to be a classic RNG function (x * 0x41c64ed6 + 0x3039), but it's only called in one place and the result appears to be immediately trashed. Really bizarre; I'll continue looking into it, though.
EDIT: Tried a number of techniques, couldn't find anything else; my guess is the random function actually used is too 'naive' for any of the usual routes to find it, and without breakpoints on memory write, I doubt I'll be able to tie it to an observable use of randomness easily.
EDIT: Tried a number of techniques, couldn't find anything else; my guess is the random function actually used is too 'naive' for any of the usual routes to find it, and without breakpoints on memory write, I doubt I'll be able to tie it to an observable use of randomness easily.
Could this be useful to track down the RNG mechanics? Unless you had already been using it...
Exhaustive tutorial + explanation on the Dolphin's debugger
I'm not as sad about the timeloss anymore, so maybe I will continue the WIP sometime. Just focusing on Mario & Luigi at the moment.
Joined: 12/1/2012
Posts: 103
Location: New Donk City
I had been using CE and the Dolphin debugger in tandem, yes (and I tried again after break-on-write was fixed). No luck in finding anything manipulable, though.
Dark Deal on nico did Triple Jump in 63.32 m (=207.74 ft, according to Google).
This beats the current record of 188.37 ft.
I'm also updating the records page with some more records by Dark Deal.
I checked Dark Deal's videos again and I saw he improved these scores:
Chomp Romp13:20* --> 13:11**
Piece Out (1 min)85,750 pts --> 85,899 pts***
* at 9:48 in the video
** at 1:55 in the video
*** at 13:35 in the video
I'm also updating the