Posts for slamo


1 2
5 6 7
16 17
Post subject: Re: #6820: wesen's GBA Pink Panther: Pinkadelic Pursuit in 23:10.4
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
I think the difficulty is not grounds for rejection in this case because of the uncertainty of which one is faster. However, if a run on the hardest difficulty is only a few seconds slower I would be tempted to obsolete this anyway, since it's arguably a better representation of the game.
Level 5: Under Water Boss Thanks to the difficulty setting this fight is 2174 frames (about 36 seconds) faster than with the highest difficulty. However there is a WIP where the fight is done 24 frames faster. WIP: http://tasvideos.org/userfiles/info/64938183472332904
Any explanation for why this wasn't implemented?
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Apparently if you make a savestate before the power-off and load it, the game saves properly. I originally watched this run in Bizhawk with TAStudio open and watched it in multiple segments, saving the TASproj to bookmark it and then closing Bizhawk a couple times. This is probably why I didn't notice any sync issues. This doesn't seem to be any fault of the author, it's just some strange Bizhawk bug. I added 113 more frames before power-off, which was the minimum amount I could add to get it to save properly. It now plays back fully from the beginning for me. Here's the movie: http://tasvideos.org/userfiles/info/64480073615469364 If fsvgm777 can confirm sync on this, I will update the movie file.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
This is a good idea for a thread, I do like to cook/bake once in a while when I'm feeling up to it. I made some chocolate crinkle cookies for Father's Day, I'm pretty happy with how they turned out!
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Any progress on resyncing this? Do you have an idea of how much work it will take?
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Dimon12321 wrote:
EZGames69 wrote:
There’s two different firmware.bin files that I discovered that exist. One of them contains the DS splash screen and the other one doesn’t. It’s likely slamo used the one that doesn’t have the DS splash screen as it’s currently the only verified dump of the firmware in bizhawk.
Probably. I tried it without firmware.bin and the results are the same.
Try this one? I resynced it on a verified firmware file, I just had to adjust a couple loading parts and also redo 11-0 for some reason. http://tasvideos.org/userfiles/info/63170876617741887
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
I haven't set a firmware file at all. Bizhawk lets you run the DS core without one. Hopefully if one is required, it's just a matter of inserting frames at the beginning to resync it?
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
I was under the impression that nKit compression didn't have any adverse effects on the emulation, and I guess that's not the case. Chamale, can you please try resyncing this on a good dump? You may have to look into hex-editing the DTM file, if you're lucky then the loading times may not affect RNG. If it's not fixable then we may not be able to accept it.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
You did it! You beat Linux! Jokes aside, I'm actually very happy to see a demo of a PCem TAS. I know you have your problems with the input editor but it all technically worked. Voted yes by the way.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Bizhawk has a new DS core in the latest dev builds so I thought I'd test it with this game. The game is very sync-friendly so improving previous levels is easy. I'm not sure how this compares to the old WIPs in this thread since those files are long gone, but I'm pretty confident this episode is faster than those. http://tasvideos.org/userfiles/info/62279214089241435 Link to video
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
After some trials and tribulations, I've confirmed that this movie syncs. The link to the memory card in the submission text is dead, so I re-synced the verification movies and made a new memory card file. You can use this link to get it: https://files.tasvideos.org/common/SubmissionFiles/6501S/MemoryCardA_USA.7z Now I could use some opinions on something. This movie is not Vault-eligible because it's save-anchored. This basically means that it must be compared to other branches of this game, namely the Mercenaries: Castle run. Does this run offer new TAS material over the Castle run? If there's too much overlap, which run is a better representative? I'm leaning in one direction, but I'd appreciate some other opinions before I make a judgement.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Sorry this is taking a while to judge, I just have some concerns here. So as mohoc pointed out, there is a glitch that can be done to skip the Gravitron (saving and loading upon entering the Gravitron room) which is omitted from this run. While the speedrunners don't use save + quit in their any% branch, it seems kind of arbitrary to me to not use this in our branchless run. The Gravitron skip does not use the same glitch as the game end glitch branch. While the game end glitch requires you to save + quit, the actual glitch used to end the game involves text box storage and game state manipulation, while the Gravitron skip requires only a simple quicksave + load. I have personally tested it, and this Gravitron skip still gives you the smaller hitbox required to clip out of bounds and beat the game. Just so there is no confusion, this is not the same Gravitron skip that's mentioned in the submission text and done in Scepheo's run, that skips the entire Intermission 2 sequence. Was this skip omitted because it was not known, or was it intentionally left out? If it's the latter, why?
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
I never thought this WIP would see the light of day, but here it is. This is after I resynced the run on easy difficulty, I got to the flying section of level 5, realized I would have to wait a while and die intentionally, and then I lost interest. Link to video My WIP, up until the part where I stopped, is about 33 seconds faster than this submission (ignoring the 10 seconds lost during startup, which isn't gameplay so we don't care). It's not just the hit score counter that's causing time losses. There are some more efficient enemy jumping combos that can be done, and momentum is stopped in situations where this could be avoided. Also, there are some strange movement choices, i.e. why do you jump off to the right to start level 5?
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
bunnywalk wrote:
I made this mainly just because it's my favorite childhood game, and I saw that there wasn't a TAS of it, so no, I don't plan on obseleting the WL2 100%... maybe. I'm waiting to re-do this to see if more people think this catagory is arbitrary when applied to TASing, I may just do a full 100% run once judging is finished
I was under the impression that you were working on an improvement to the movie? If that's the plan, you can just post the movie here and I can replace it. No point in publishing a 2 and a half hour movie if an improvement is imminent. If not, let me know and I can judge this. Personally as a judge, I don't think the "all chests" category is arbitrary at all. Chests are an easily definable and finite collectible that gives you a special ending if you collect them all. It also helps the category's acceptability that there are currently zero TASes of WL3 on the site right now, however, I do believe that a "100%" run would obsolete this run because there would be too much content overlap.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Oh boy, Skunny. I actually tried TASing this game for April Fool's at some point. I thought it would be somewhat entertaining to watch thanks to the ridiculous boosts you get from jumping on enemies. I tried running it on Hard difficulty and ran into a major problem on level 5. The only difference in difficulty that I noticed was that it starts you off with a lower timer; this becomes a massive issue in level 5 because even with TAS-perfect movement, the level is impossible to beat before the timer runs out. Just another example of the Skunny games being low-effort trash. One quick optimization issue I noticed is that you were running into a lot of the scenery in the Super Rosie sections. While this doesn't slow your character down, each hit adds to the "Hit Bonus" counter at the end of the level, causing you to spend more time watching the score count up. I should post my WIP soon to see how the rest of the optimization holds up.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
FYI, this synced fine for me on the Steam version, so I think you can use that if you don't have a DRM-free one.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Ok, I replaced the movie file (and fixed the rerecord count to match the original).
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
I updated the info for PCem. You no longer need to apply the patch to regular PCem, TASVideos now has its own fork. Keylie got it working with the new v15, so you can get it here: https://github.com/TASEmulators/pcem/releases
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
That's fine if it uses certain HTML features, it looks like features can be swapped around without reinstalling the whole thing. There isn't much documented yet on TASing with PCem, just the PCem fork and this issue for libTAS: https://github.com/clementgallet/libTAS/issues/208
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
I haven't heard of it until now. It looks like a really promising option, especially 98micro which gets rid of all the network/internet related stuff that we definitely will never need. My opinion is that as long as the game is emulated properly, it doesn't matter how we got there. And the smaller the hard disk image, the faster it is to load and save states. I'll try it out sometime soon. It's great to see some more people testing this setup. A new branch of PCem for TASVideos was set up recently, there are people currently trying to adapt the patch to the new v15. If you want to keep up with the progress or if you want to report any bugs that occur in only the patched version, you can go here: https://github.com/TASVideos/pcem
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
I appreciate the effort. I've updated the movie file.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Sorry that this is taking a little while. It's tricky to compare this to the Gens run. EZGames, is it easy to resync this if you edit some inputs in the middle of the run? There are still a couple (minor) improvements that the published run has over this run.
    * Stage 2 boss: The first hit on the boss can be done 10 frames sooner. * Stage 5, 4th area: One frame was lost, and I'm honestly not sure where. Don't sweat it too much if you can't figure this out. * Stage 6: The boss fights take the same amount of time but the published run gets in the hole 4 frames sooner.
Other than that it looks good. My best estimate is 116 gameplay frames saved so far.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Thanks to some hard work from keylie, PCem is now working properly in libTAS. I've been testing this with Windows 95/98 games with extremely promising results. I've updated the original post, check it for details on how to get it working.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Hey, have an encode anyway: Link to video
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Sorry about the last example, it turns out the fire triggers a warp point. Also, using an analog value of 127 instead of 89 seems to have no effect on the walking/rotation speed. My point still stands about the analog usage, though. Much of the movement is done imprecisely. Here's an example of what I mean: Instead of gradually moving the analog stick to the right, it should be done instantly. Doing this saves only a few frames in this case, but movement is done this way throughout the entire run, and it can really add up. You should also always travel in a straight line wherever possible, there's some strange direction readjustments being made throughout the run.
Experienced Forum User, Published Author, Reviewer, Expert player (2393)
Joined: 5/21/2013
Posts: 414
Some of the movement is definitely sloppy. Here's an example of a section that I very roughly improved: https://i.imgur.com/9ZsyD00.gif (big file, open externally) The pathing in the original run was very questionable in this example. (bad example, ignore) From looking at TAStudio, the analog stick movement was gradual and erratic. Also, the analog values top out at +-89 instead of the extremes of -128 and 127. My guess is that most sections were recorded unassisted with a controller.
1 2
5 6 7
16 17