Has anyone else noticed how much better his English has gotten over the years? I just thought that deserved some positive recognition.
Excited to watch this run.
Meh vote. Would be yes for casting the best spell in the game on the best boss in the game, as well as alternate bat form... but then you had to go and diss the best sword in the game in the annotations. Shameful.
Seriously, though, fantastic commentary. This is the kind of thing I can have a friend watch and he'll enjoy even not understanding the game at all. And the run was entertaining. Yes, obviously. And I support the idea of obsoleting the 196% map run. "Best ending" is a bad goal choice for this game, for all the reasons we've heard argued before.
General question: do game shark cheats work generically that way? For somethin as in a color of an object and so on?
I realize this is off-topic a bit, but it's a good question and gives insight to how games work. Short answer: "generically" (you meant "generally", right?) speaking, yes for older games, probably not for newer ones. The coding for color is going to be tied to how the programmers implemented it, and in the case of the 3D Zelda games on the N64 they have tunic color set up to tie to a simple color code (probably somewhat similar to how HTML reads color codes for font, but I'm just guessing here). In newer games, however, color tends to be mapped in some way based on texture or other more complicated coding factors, and it's less common to have one color used for an entire model (this was standard back in the first generation of 3D games, before textures could be nice and defined and high-res).
Also, palettes were typically 4 colors in the NES era, 16 colors in SNES, etc. So changing colors was simpler and could be done in a few bytes back in those days.
That said it may still be that even the current gen Zelda games use simple color code. I don't keep up with this stuff any more, sorry.
The less obvious part is that, regardless of rules and methods involved, abusing a feature that provides the player character with unlimited or otherwise endgame-worthy resources at a subjectively early point where for us the viewers it would interfere with the progression of the game—the character's visible effort of dealing with its obstacles. Just like with real movies, we actually get a lot more entertainment out of seeing how the character is developing and being pushed to their limits than seeing how there is neither development nor actual limits. It's the limits that give shape to everything in the first place! They make us emotionally connect to the onscreen action, and where there is no limits, there can be no such connection.
Would you say maybe it'd be a good line in the sand to judge a cheat code based on how it's accessed? Button input on a title screen and the Golden Torizo debug code may both have similar results, but in this particular TAS the code is used in a manner that is effectively TAS-only due to the route. More importantly, even in a console playthrough, a serious speedrun that uses this cheat will still display adept execution in the setup process. I don't want to get stuck on that aspect of the code, but it's almost more like an earned bonus, given the difficulty of getting it in any competitively fast route. Maybe we could view it akin to, say, a run of Goldeneye that earned cheats along the way that required good play to unlock, and then activated them if they sped up the run. Obviously it would not obsolete a non-cheated run, but isn't there some real entertainment value there that should give us pause?
moozooh wrote:
Unlike a cheat code, stuff like ACE, memory corruption, and accessing hidden game-breaking debug features—i.e. completely ignoring in-game limits—at least get geek points for being ingenious and unexpected, but entertainment-wise they roughly amount to Luke Skywalker asking Obi-Wan to teach him how to use the Force, and then promptly killing Darth Vader with a lightsaber in the next ten minutes of the original Star Wars. Cheapening the experience is what it's called. I, for one, enjoy it a lot more when I see the awesomeness of an in-game character develop, not come "for free". This run's main (and only?) saving grace is that it ends at pretty much the same moment the "free" power comes, and before that comes some pretty hardcore "development".
I think the movie analogy is a little off, since games are repeated experiences and presumably viewers are not going to watch a TAS of a game they've never played expecting the storyline to be preserved. I get what you mean, of course. I think the argument should never consider throwing out the "movie" as it is. The appeal for these situations is more like an alternate ending, where the use of the code serves to cut out parts of the run that would be redundant otherwise compared to the non-cheated run. It's more like we have the original Star Wars, then we have two possible versions: one where Luke destroys Vader and explodes the Death Star with amazingly over-the-top superpowers within the hour, or one where he does the exact same thing but we still watch through the whole trilogy until the end. The rule itself should maybe be relegated to a guideline: something along the lines of "cheat codes should not be used except in specific situations, and TASers who want to use them should discuss their reasoning in the forums ahead of time."
Anyway, it's a moot point as the run was published. I think the rule is better as a guideline for the reasons I've mentioned, but in general I have faith in the publishers to know to judge such situations on their individual merits. Removing the rule does away with the "concreteness" of such an argument, and I'm just opposed to black-and-white reasoning in such grey areas as I think it creates noise in the discussion that doesn't contribute to the determination of the best resolution.
I absolutely agree with Warp that cheat-code runs have no business obsoleting non-cheated runs. Fortunately, I don't think anyone who has a say in the matter would seriously consider such a move.
I realize it's a can of worms, but this is a good time to really revisit the "cheat code rule" issue itself.
That rule is antiquated. It serves an important purpose, but it could not have possibly been made with an awareness for this sort of situation. Moreover, what it means for a run to get published has changed, with the addition of the tiered system. Finally, "executes arbitrary code" was a nonexistent category when the rule was developed.
To those who are arguing that the Earthbound TAS is an exception that is permissible where this run is not: it is not intended for players to use the debug "cheats" in either run. In EB, it is accessed through errant functionality in the game's "Talk To"/"Examine" code. In Super Metroid, it is accessed through erroneous code related to health, pausing, and the "green gate" glitch mentioned by Garrison. In either case, you have a run that could not be recreated in the manner seen here by console play.
To me, this is the key point. Is the cheat code accessed or used in a manner that emulates real play? Because if the goal of a TAS is to show off amazing technical feats, that is an important question. And the answer to that question for either TAS is "no." You will never see a real time run of this route, and viewers cannot possibly see this sort of run and feel "cheated" by what happens (pardon the pun), at least not in the way that they would be disappointed by a game that gives your character permanent invincibility or one-hit kill abilities based on well-known controller input sequences on the title screen of a game.
I think the spirit of this rule was made to avoid people creating runs where technical precision in managing elements like health, enemy damage, and other mechanical factors could be nullified with cheat codes that would remove a large factor of what makes TASing unique. I don't think the rule was made to preclude a route that makes the current real-time run faster than the TAS run would be without it.
Correct me if I'm wrong, but as this glitch requires plasma beam there is probably no way to compete with the real time record (somewhere around 18 minutes 20 seconds in real time from Ceres to finish, I believe) if the debug code is not used, right? Rejecting this on the grounds that a cheat code is used may run the very real risk of having a TAS never capable of being faster than real time strats. I can't think of an example of where that has legitimately happened in the broader speedrunning/TAS world yet. I think that alone is reason to publish this run.
Anyway, long-winded post on my part. The TL:DR is that the rule was made in a different time, and it may now be detrimental to TASing because the TAS scene itself is quite different than it was back then. For those who simply feel cheated by any sort of codes or "total control" elements: your "no vote" for entertainment is certainly justified from that perspective, but realistically I think this is exactly why we have tiers now, because actually not publishing this run at all would be passing up on the absolute fastest completion of the game, and that was the debate that led to tier creation in the first place.
Well, I voted yes. I'm not up to date on the current state of TASing this game, but I remember optimization was very difficult, and a lot of people seem to have put it on the back burner.
I'm sure there's more optimal ways to do this game, but that's sort of the exact problem: games like these have so many possibilities that need testing that it's almost impossible to conquer, and they end up never getting a run because everyone who attempts them soon gives up. Short of someone breaking down the exact code that handles population growth, this game will never have an optimal run. Taking into account pollution, traffic, crime, land value, power, commercial/industrial/residential spread, RNG... yeah.
I understand objections to its slow pace, but that's the purpose of tiered publications, right? I hope we have a place for this. To me, this is one of those "childhood games" I look back at nostalgically; I play it on emulator every few years, and it still has appeal. I thought this was a pretty good TAS, and it's long overdue for some love here.
As a side note, I thought the PlayStation would have better 2D graphics than this. Most SNES and Genesis games have much better graphics than this.
Keep in mind: this was a PC game, running on DOS, ported over to console in a day where console ports of PC games were not exactly commonplace. Blizzard was also not nearly as large a company as it is now, and I think the entire development cycle of the game was less than a year. I can't even remember what the resolution options for these games would have been. 600x480?
I'm sure it would have looked a lot prettier if it were developed for console.
Well, as far as publication goes, what's the rule here? If a trick is ridiculously difficult to optimize, would it be better to not use the trick and have optimal play without it, or better to use the trick, doing it without perfect optimization, and just accept that there's probably time/frames to be saved?
Hourglass doesn't fully support the mouse (from what I'm still seeing)
Thanks. I know very little about it. I imagine whenever it gets there, it might be better for entertainment purposes (camera control) and easier for manipulating RNG maybe... however, that just makes this run all the more impressive, being limited to controller scrolling.
This was a really fantastic watch. Well done. Definite yes vote. Abusing the AI was quite impressive.
May I inquire as to why you chose to run the PSX version instead of one of the PC versions + Hourglass?
This was really fantastic. I don't suppose anyone has a total count of dead lion occurrences? Because I laughed harder each time I saw one.
And this run definitely colors a dinosaur. It also contains a cat being eaten by a zombie which should really be a category in and of itself. Yes vote. Hope you guys can do the newest Scribblenauts game in time.
Encode looks to be working as of now. Watching, will comment after.
Edit: Yes vote. Nothing I can say that wasn't already said. Route's great, interesting throughout.
Looks good (and yes vote) but I don't get it. Why do you evert to 6-7 at the end? What happens in level 7 where you're at 7-7 throughout and suddenly you end up in 7-8 even though there's no eversion point nearby? And what's with all the going back and forth in level 8? Please enlighten me!
It's explained in the comments: a glitch related to dying in certain places allows the player to toggle eversion when he shouldn't be able to. Technically you have to collect all gems to use the eversion point in level 7, but using this glitch lets him evert without doing so. As long as you complete the level in an everted state, you go to the extra "castle" level.
I assume he uses the glitch more here because it's faster in some cases to die and trigger it than it is to get to some of the eversion points (the areas loop until you evert, and the eversion points are usually pretty far in and take some time to reach).
Sorry. Feel free to delete the previous posts if you so please. Back to regular Zelda discussion (in other words, mostly silence until someone finds something new, I suppose).
julianface: not to beat a dead horse, but go look at his post history either here or on SDA (where he was also banned). I've actually become convinced he may have some form of autism based on his behavior and attitude. Sad, but it doesn't mean we should tolerate it here, as his presence generally caused negativity and conflict due to his constant and unhelpful/condescending attitude. "Guys, you need to test this. no, I can't test it, I'm no TASer. but test my ideas! why haven't you tested them yet? because they're bad? I still want you to test them. I'm smart, I came up with things, test them!"
Let's let the dead horse rest. Any way you spin it, you don't get to demand the testing of your ideas if you aren't willing to do some work yourself. That's just selfish, and it's an insult to the intelligence of the people here who actually TAS the game or at least take the time to plot out and think out routes/tricks/etc. by comparing times from previous runs (he didn't even do that!).
I find it very hard to believe you couldn't beat Ingo on a brown horse in a TAS, but I'm sure you guys know better than me. I mean, you just need to horse superslide, right? What, we don't have a way to do that yet? Lame...
Earthbound's any% had a similar shift in glitching (from completing most major plot points to skipping to the end a certain way into the game) and I believe that obsoleted the published run there. Not sure if it's a reliable precedent, but it's the closest example I can think of for the given genre of game and style of glitch.
Unoptimized, as in there is room for improvement. An optimal TAS is done to the best technical ability known at the time. This ensures a high level of quality and makes for a faster/more entertaining run.
Vykan is saying that your low rerecord count suggests that the run is not optimized and could be improved easily. As a rule, if you want a TAS of a game published here you will want to maintain a high level of optimization, which generally results in a higher rerecord count than what you currently have.
That looks amazing.
In the forest of silence I noticed some lag when a lot of enemies were on-screen. I'm guessing you already tried killing off some as you ran before the lag started?
When he says it might be possible in TAS via overflow, he means that if you could find a way to specifically manipulate write functions in this game, you might be able to abuse this. Think how the Super Mario World run does some silliness with Yoshi and enemies to skip straight to the credits, or wrong warp in Ocarina of Time... finding a way to write values in ways they aren't supposed to be (abusing programming errors, basically) could be used here.
Unfortunately I am not aware of any means we have to mess with write values in SM64, though as I only follow this game casually there may be ways to do this. I'm sure those who have TASed this game have a better idea. For the immediate future I don't think we'll see anything come of this.
Well, for what it's worth, I've been following the thread and am interested... I just don't have the ability to view your m64s where I am. I hope you pick up the project again in the future, though I understand where you're coming from.
Is there a practical use of the hookshot glitch in OoT for TASing? It seems unfortunate that such a cool glitch has so few implementations. Obviously it won't be used in any% with WW, so I guess that leaves MST w/no RBA... I've been puzzling over this one but I can't think of any other uses for it. A 100% run could probably make use of it but it seems like we won't see that TAS until the emulators are more reliable, if I understand correctly.
I fear this game is reaching a point where it's so broken that even the cool new tricks are hard to find uses for. Another example: CosmoWright at SRL managed to glitch the Trials by overloading actors until things were despawning, and it looked pretty cool, but again there's no way practically to use this in a TAS, is there? Unfortunate, I suppose...
Well, looking forward to whatever progress is made on this front. The guys here continue to impress, so I can't complain.
Pheenoh wrote:
I won't bother continuing this; I'll just join everyone else in ignoring you.
^
Once I stumbled upon the discussion leading up to his ban at SDA I realized the merits of this approach.
Knuckles is a good choice for such runs (maybe Tails in S3/S&K) seeing as he has some tools to use to avoid rings that make it more interesting. I wonder if Sonic could be done as impressively... it seems like you'd either be forced to use zipping glitches or have levels that are simply unbeatable.
At any rate, that is pretty cool use of the game physics and momentum. I stand corrected.
