Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
moozooh wrote:
First of all, SDA is inherently more competitive due to the nature of unassisted speedrunning (where TASes are better seen as collaborative work).
While it is collaborative work, and we acknowledge the works of the authors before us, sometimes an improvement or a run is due and no one is willing to step up. I would hope that bounties lead to runs on overdue or somewhat high demand content, that wouldn't get produced otherwise.
moozooh wrote:
In a TAS, basically everything is a question of time the player is willing to spend on a project. Depending on their knowledge, they may even have most of the job done for them using Lua scripts and somesuch.
Effort is definitely an issue, but I would say that anything that gets published is progress towards a great run. Even if a run doesn't meet my expectations fully, at least effort was given towards it that makes it easier for a future TASer to build upon.
moozooh wrote:
While a finished product does benefit community in a way, it also creates an unnecessary motivation anchor that may psychologically devalue regular contributions.
I don't see this becoming that radically popular that only bounty runs end up getting submitted. I see this as a way of motivating people to work on difficult runs that users want to see, but players may find extremely difficult to do well.
moozooh wrote:
I think, instead of pouring money into something that is already appreciated by hundreds and thousands of people, they're much better spent donated towards development of emulators and their functions, and to TASVideos itself.
While I completely agree with this, sometimes people aren't really motivated to put money forth without seeing something directly in return. If a player chooses to be modest and have the money donated towards the site, then it's win win. Some probably will, some won't. I doubt the amount of money going into bounties affects the amount getting donated to the site.
There are always requested runs, people don't take any merit from them.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
EDIT: Currently under discussion it seems. (Need more feedback from others.)
Want to make money TASing? Looking for a game to TAS? Check out the bounty list below.
Communication should be done between the user TASing and the user offering the bounty to make sure conditions are agreed upon. It is assumed that users will either use Paypal or negotiate some other form of payment. No one makes any guarantees to the bounty being paid out except the user offering it. We can't be responsible for people who don't hold to their word.
It is my hope that the requested videos are for the most part publishable material, as this only makes the site content better, but smaller demonstrations of unpublishable material are also acceptable. I would prefer that this list stick to games currently TASable, and somewhat substantial offers. ($1 is kind of stupid unless you're adding on to someone else's already placed offer.)
A case that I believe needs stated is that the bounty should be given to the first submitted case satisfying the requirements. It doesn't make sense for someone to make a great movie only to be obsoleted by ten frames a couple weeks later and have their bounty taken. I would not encourage 'fastest time by X date' for this reason.
Clarifications:
1: Game, System, Version?
2: Run type: 100%, any%, glitched, demonstration? (If a save glitch is found that skips most of the game, are you still planning on paying out?)
3: Does the run have to be accepted to qualify? (What conditions otherwise?)
4: Run time requirements?
5: Due date (Give a rough due date, you can always extend it.)
(Idea taken from the SDA thread. I will try to maintain this page fairly regularly.)
(Thread layout comments? I kind of just threw this together.)
Bounties:Example, not in effect yet.Secret of Mana
- $10 - Kirkq
- Must use 3 players if it provides the fastest completion, must be published.
- Due by January 1, 2012 (will probably extend)
Who will claim the unofficial title? (Courtesy of Kyman.)
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Kirkq wrote:
Mercury Lighthouse: a normal retreat glitch after going up the stairs in the room filled with water will take you to the aerie.
One possible strategy: recruit Mia, go down the stairs you came from, retreat glitch up the stairs, backtrack ~3 rooms, fight the Djinn, Retreat glitch out of the Djinn room. ((I think this takes you to the ply statue IIRC))
Sorry, the first strategy skips Mia, the second is if you want to grab that water djinn. You should retreat glitch straight into the aerie.
Doing both is probably possible (to grab the water djinn and not Mia) with a lot of backtracking/multiple retreat glitches.
I feel like you shouldn't be ground pounding the little boos.
0:12 - Long jump past, turn around, sideflip, dive through boos
0:20 - Jump kick
Maybe both of those don't actually work, but I feel that some other plan has to be faster. Fairly good execution, planning/strategy looks a bit mediocre.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Mitjitsu wrote:
Are you saying there could be some secret exit off one the maps that leads to the Venus lighthous? If so you'll still have to overcome the difficulty of beating Saturos+Menardi and Fusion Dragon.
Yes, this is a possibility, since I found a secret exit (in some cave listed earlier in the thread) to Vale. The game would then become djinn collection. My thoughts are that we may only really need 4 water and 4 fire djinn to beat the game. I want to find all feasible shortcuts before I start testing this.
Mitjitsu wrote:
Would it be practical to skip Mia? After all she plays the role of the sterotypical healer during the boss fights.
It is not very practical if there are no glitches to be gained. She comes at level 10 with a water djinn, which makes the Saturos fight much easier. (Collecting another water djinn in Mercury Lighthouse seems optimal.)
Mitjitsu wrote:
EDIT: Out of interest for GS2, does anyone know the password to transfer max stats over to the characters from GS1?
I'm sure this exists on GameFAQs. (EDIT: I sort of misread this, I thought he meant maximum completion/djinn)
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
t3h Icy wrote:
1) Since there are 28 Djinn and 21 possible available slots, what happens? Do you cap out on the Djinns you can pick up, or do they get put on the right hand side where Mia would be, or what?
No clue, worth tinkering with. I haven't really looked into the ramifications of skipping Mia yet. I wonder if it skips dialogs, etc.
t3h Icy wrote:
2) Are all sections of the game, including dialogue and the like possible without glitching out the game?
Don't really know specifically what you mean by this. It's probably something I haven't looked into
t3h Icy wrote:
3) How do you actually skip getting Mia? >_>
In the room where Mia joins you, use retreat on a shoulder button without enough PP and go back out the door you came in. (I think there is a statue you need to move for Mia to progress or something.)
t3h Icy wrote:
Because with what was mentioned about the game competition for the Lure Cap, it sounds like it may have been intentional that you could skip Mia.
Since this seems to be the only way to skip Mia, I doubt it was intentional. The thing is that the game assumes you have Mia at Colloso, and since all characters are taken out of and put back in your party at that point, another glitch occurs.
Definitely let me know what you test so that if I or anyone else gets around to running this, we don't have to redo things you already tried.
Definitely check out retreat glitching and save glitching (both detailed in this thread). Also my lua script removes random encounters and sets Isaac's PP to zero.
EDIT: Also, I've tried skipping areas up until Kolima Bridge. I've only had luck skipping through dungeons instead of bypassing them entirely, but it's extremely possible that it is entirely possible to bypass later areas. (It may be very possible to glitch past Babi Lighthouse into Venus Lighthouse for example.)
EDIT2: I never actually went off map in Mercury Lighthouse. As stated previously, off mapping in one of the early caves had a secret door to a cave requiring whirlwind in Vale. It is very possible that hidden warps to other locations exist. How useful? Who knows.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Randil wrote:
How about a run that starts from a savestate right before round 0, shows off this kill screen (i.e. not kill the birds as fast as possible but rather as late as possible), then resets the game, and does what this movie does?
It isn't even really a normal TAS if you start from a savestate at the end of the game, and would probably just provide confusion because the circumstances were forced by a savestate.
It seems to me that this is creating somewhat arbitrary guidelines in order to make the game publishable. It would essentially be a quick, novel demonstration that may be better suited for a youtube video.
Props for the demo, but I don't think this game is good for publishing under conditions of the potential based on current knowledge.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Oh awesome.
I really hope my negligence to finish my run in a timely manner didn't detract from any technical tricks in the later levels. I'll be sure to check it out when/assuming it gets posted.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
A few notes as I go on. I'm trying to do a couple levels a week.
Hopper Land: It is possible, but somewhat difficult to do an angled jump on the cliff a ways left from the exit. This requires only one wall jump to beat the level.
Junfalls: It is unnecessary to bomb either of the platforms. Jump at the higher one and drop a bomb and you will grab it. From there you can jump up the slope of the waterfall.
Midterm Tuesday, hoping to start Cool Cave Wednesday.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
(Submission Publication) wrote:
We realise that as of this time (June 2008), the target tool-assisted completion time for this game has been lowered greatly. For an example, see submission #2012. However, we don't want to replace this movie with one that has a lower quality of play, since that would go against the site's standards.
May I note, "June 2008"
I don't know why anyone feels the need to come up with a formal site mechanism that says this statement for 1-2 movies. This published grue thing is not necessary by any means.
Retroactive obsoletion is another completely fine issue, for that one movie where the completion conditions of a published movie were similar, (albeit not exactly the same).
The debate on unpublication is an entirely separate issue, which I vote yes to for the 2 movies that don't really satisfy the site criteria. If that other game has no potential to make a good run and was poorly played it should be unpublished too. If it could make a mediocre run and was poorly played it may as well stay.
-
With regard to poor runs, I feel like the worst published runs should be somewhat representative of the minimal threshold at which to accept a movie. I myself obsoleted Dragon's Lair. It's a really hard game. Also, there was an AVGN of it. The movie does suck, but it is worth watching to a spectrum of individuals. It has some minimal value to contribute to the site. Even most of the oldest runs are still watchable by some value. On top of that, if you want to get rid of that run, I don't really care. I wasted a bit of time obsoleting it to scratch a 2004 movie off the list. I'll live.
I don't really know how bad the couple runs we are talking about are, but if they provide no minimal value to the site, and don't look to improve in value with an improvement in quality, they may as well be removed by an overwhelming majority vote.
Again, we are talking about like two movies in each case. This doesn't need to be any sort of sweeping generalization.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Just got done. =p
http://www.youtube.com/watch?v=XNeyTnt3s6U
Input file with same progress: http://dehacked.2y.net/microstorage.php/info/708806835/Begin%20Hopper%20Land.m64
Progress will probably continue to be uploaded one world at a time. This run is in the home stretch now. The run is currently at 32 minutes and 61k rerecords. I've been doing this for too long. xD
Regarding World 4: I get to showcase my favorite find of the run in Storm Valley. There's a few fairly linear levels that should let me progress quickly. The boss in Heaven Sky is going to be a bit annoying, as I'm not exactly sure what criteria kills him the quickest, and I want to make sure it looks clean. I have a cool, optimal strategy planned for Natia now too. World 3 clocked in at an unfortunate 12 minutes, World 4 should be around 9-10 I'm guessing. Hopefully World 5 and credits fit in one video. I digress. Enjoy!
Will I meet my self-imposed April 1 deadline? I doubt it. I may come close though.
EDIT: Bolban smokes too much weed. LOLOLOL
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
RadioKJ wrote:
For some reason I can only encode the first 18 minutes before Mupen crashes with an error suggesting I'm out of memory.
This happens to me almost religiously on my Bomberman Hero video playback. After I run through what is probably exactly 18 minutes, an error message continuously pops up.
I'm recalling the rest of this from memory, so it may not be entirely accurate. This also happens after 18 minutes of non-video recorded gameplay, and I'm pretty sure it happens after 18 minutes on other games as well.
My guess is it is some sort of memory fill within the emulator or perhaps the graphics plugin, but this problem is very familiar to me. The only solution seems to be to reboot the emulator to get it to run flawlessly again for 18 minutes of gameplay.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Pretty cool. It's quite interesting for the first few levels and then the novelty wears off a bit, but it is definitely cool to see it TASed.
If it was optimized well enough (by your judgment), I think it would have a pretty decent shot at publication. It's short and interesting.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Captain Forehead wrote:
I assume that "him" is the planet?
Him being the creature on the planet which will also have to be revisited.
neskamikaze wrote:
they're worth 100 and they'll set you under a 5, with the exception of marine/jet/slider levels.
Remember enemies that explode/kill themselves are sometimes worth extra points. (Daggers etc.)
Move Stone is finished, 81 seconds of pushing blocks. <.< I'll make it my goal to kill Bolban tomorrow and post the video....and do some homework. xD
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
He already skipped 50 points in Heavy Room if I recall correctly, but it's easily unnoticeable.
The game has no recognition system that you ever got 50 points over the 'maximum criteria', so I am quite vehement that it should not be in the criteria. Also, by your system of 'maximum score per level', we would have to kill numerous temporary enemies in every level, and spend more time grabbing things in water levels. On top of that, maximizing the amount of points in water/sky levels would probably force us to backtrack or slow down considerably. If you are saying maximum score on 'non-bomber device' levels, you are getting pretty arbitrary.
One step more extreme, in order to get maximum score on Planet 1 Endol, we need (I think) 6 health. So, we would have to revisit him after gathering 400 crystals.
The game has a fairly set definition of 100%, it unlocks a secret world, then it unlocks Millian's treasure hunt (which may be worth a debate of inclusion.)
EDIT: While I'm at it, I might as well keep you all updated. I'm currently on the second to last level of World 3. I may upload the other half of World 3 within the week. =) Unfortunately World 3 stays split in 2 parts because it runs about 12 minutes.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
I hate to nag, but 4050/4000 is a bit too obviously suboptimal. I would recommend restarting from before the skewers and skipping that blue crystal right before the heart. It will look much better, and the platform at the top of the tower will probably be in a better position.
Other than that, your work is much higher quality than previously. Make sure to pay attention to points you could be scoring while waiting. (Secret room was a bit obvious that you should be killing things on the bottom of the level while riding around the top. You did better in Sky Room.)
Good quality overall, just pay attention to level routing a bit more.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
I'm going to try to give some general constructive feedback for TASing. It takes a long time to be able to effectively and efficiently make a polished TAS for submission.
On this line of thought, I note that your previous submission had 465 rerecords for 525 seconds of gameplay. This is less than one rerecord per second of gameplay.
If we consider information here: http://tasvideos.org/MovieStatistics.html
Average movie length: 24:08 (1448 seconds)
Average rerecord count: 28818
This suggests that an average published movie on this site has roughly 20 rerecords per second. This would imply that you should try roughly 20 times per second of gameplay to save frames.
Now obviously if you are just walking to the right at a fixed speed, there isn't too much to do, but this is where you should try to take a fastest route that is creative. When you get to a point where you accelerate, decelerate, or have to solve a problem for the fastest solution, this is where your rerecords should come from. You should be satisfied that you solve the problem to the best of your ability before you move on. You should know that it isn't within your capabilities to save any more frames. You should be using frame advance almost all the time when trying to implement the solution to these difficult problems. This is where we look for optimization.
Now obviously this takes a lot of patience and perseverance to hold yourself to a high quality standard, but it is the only way you'll get better at TASing.
For now I would say roughly 10 rerecords per second is a good goal for a short game like this for a new TASer. A movie file of this length should have roughly 5000 rerecords, give or take plenty. It's just a rough gauge. Most experienced TASers can use roughly 800 rerecords per hour (I meant this in terms of real life time, not game time.), so you can quickly see how minimal 465 is. Obviously if you become more experienced you will be able to get to that level, but we all had to start somewhere.
It's not actually about the rerecord number, it's obviously about the level of quality. So the rerecord count is more of a persoal gauge for you to see if you're spending enough time on relevant segments. It's obviously easy to waste reloads or even hack the rerecord count high (so we never really consider it fully), but it is good for both you and us to understand statistically how much effort was put into the TAS.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
Some leisure reading: http://tasvideos.org/forum/viewtopic.php?t=7502
Summary: Bisqwit relinquished the server. The server is currently being hosted through a company through a pay per year type thing. Adelikat and Nach are running the server. Bisqwit wanted to disconnect the site from his life or something to that effect, so he was away for a while. He is now only around periodically, but we're always happy to hear from him. =)
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
True wrote:
I purposely made parts look sloppy if it lost no time.
This philosophy is ****ing terrible.
True wrote:
No, if it doesn't get published simply because someone doesn't like MY aesthetic choices, which I find more entertaining than a bunch of follow-the-song "dancing" or other crap (...)
Which do you find more entertaining, your aesthetic choices, or arguing about this? I'm thinking the latter is the only reason you actually chose to create the run that way.
True wrote:
it took just as much effort to make it look aesthetically sloppy without technically being so as it would to make it look somewhat polished
No it didn't. You were being both lazy and defiant.
True wrote:
This is a minor protest and guess what?
So wait, are you protesting, or is this actually more entertaining to you. I'm confused, Saturn.
True wrote:
But at least I've got the balls to do it instead of just getting in line and taking it.
Right now it seems to me as if you're fighting against the viewer response more than the judges decision. I am strongly of the opinion that if a majority of the viewers find something more entertaining, then the site should feature the fastest, most generally entertaining movie. By 'generally entertaining', I mean the author should get his head out of his *** and do his best to feature entertainment during relevant portions.
Ironically you compare it to the Megaman 3 run which is a completely different issue. That run at least did its best at the time. If you have an issue with judging that run, it's still sitting in the workbench. Go voice your opinion, though I doubt anyone cares to listen to any sort of extremist rant. Perhaps you could try constructive comments?
I did watch about half the run, it isn't too bad. The run is a bit long for my tastes, the game engine is fairly bad, but exploited fairly well. Normally I would probably place it in 'weak yes vote' territory. I suppose if anything, since this run takes an intentional poor approach to being entertaining, that it should be rejected so as not to start a precedent that the speed of runs is not the only thing that counts. This is my personal opinion having nothing to do with site rules. I don't want to see you go beat the Super Mario Bros 3 run and stand still for 5 minutes of autoscroller. This is the precedent you are trying to set.
Normally I'm not one for destructive posts, but I really couldn't pass up this opportunity.
Experienced Forum User, Published Author, Player
(208)
Joined: 7/7/2006
Posts: 798
Location: US
I don't really know the technicalities, but the emulator settings should 'best represent' how the hardware would process it.
finalfighter wrote:
We ask other TAS players.
1:Which PPU near a real machine?
2:Should we not use the glitch with BombManStage?
Use the PPU that is closest to a real machine, and exploit the game as much as possible based on this 'best representation' we have of the hardware.
I have no knowledge of PPU accuracy though.
Bablo wrote:
Who will claim the unofficial title? (Courtesy of Kyman.)
