Posts for Chef_Stef

Posted: 6/21/2007 2:15 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

IronSlayer wrote:

You have GOT TO BE KIDDING ME....this is like someone saying that "Space Invaders", "Galaga", and "Pacman" are "really boring, really bad games".

I try very hard to qualify my posts, but I think this is one case where I'm being misunderstood. To me this game looks boring and/or bad. I tried to express that I've never played (or even heard of) the game before, so I'm just judging the game based on how this TAS completes it. It could also be that some games are good, but they're boring to watch as a TAS.

"Altered Beast" is one of the most influential, famous, revolutionary games ever made, and as well-considered by 20-somethings as "Super Mario Brothers".

Well, I certainly never heard of it before (I'm under 20, though, so maybe this point doesn't count). Again, a game can be really good, but unsuited as a TAS. Look at the entertainment rating of the published movie - it has a 4.8, hardly one of the most popular runs on the site.

Posted: 6/20/2007 4:33 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I certainly think that there are movies on this site that should be removed. The Altered Beast movie is probably one of them. However, that's irrelevant and should be discussed elsewhere. The issue here is whether this movie should be published; the poll question asks "Did you like watching this movie?", and my answer was no.

Posted: 6/20/2007 4:22 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I just finished watching this, and I cannot honestly say I was entertained by the movie. This seems like either a really boring or really bad game (the sound effects/music weren't very good), although maybe I would vote differently if I had played the game before. There wasn't very much action, and what little there was looked extremely simplistic. In my opinion, this is a bad game to TAS. For these reasons, I'll be voting no. Also, I found the "extra entertainment" extremely distasteful. In my opinion, we don't need that kind of humor to make movies entertaining to viewers. Publishing this movie will set a precedent that this sort of "entertainment" is acceptable, something that I strongly disagree with. (Note: The second paragraph is more of an aside, and was not the main reason for my "no" vote.)

Posted: 6/18/2007 1:15 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Well, the last part sounds interesting, but wouldn't 162 games be really boring? Isn't there a baseball game that has less games to play before the playoffs? Otherwise, I think the concept would look really good if done correctly, but manipulating several games without any controller input seems almost impossible... FractalFusion "only" manipulated eight rolls in his Monopoly movie, and he had a program to help find that manipulation. It might be possible, but the chances of any TASer finding the correct input before going crazy seem pretty slim.

Posted: 6/10/2007 5:45 PM
(Edited: 6/10/2007 7:03 PM)

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I wanted to mention that it seems incredibly difficult to glitch into a corner like that. I've only done it twice in all the time I've been SMW TASing (not counting SDW's enemy blocks). I couldn't imagive actually TRYING to get one of these to work, you have to get extremely lucky and have the hit detection fail at the perfect time to get into the floor like that. Edit:

Does this have any potential? I know the right-facing wall is solid (so it's probably not possible to get the key as small Mario), but the lower floor is not solid (when going through it from below). Would this save a significant amount of time as caped mario, by flying through as opposed to using Yoshi?

Posted: 6/10/2007 3:01 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I glitched into a wall once in Super Challenge World (a hack of SMW), here's the SMV: http://dehacked.2y.net/microstorage.php/info/183173573/SCW%20Wall-glitch.smv Is this the sort of thing you're referring to, Tompa?

Posted: 6/10/2007 4:20 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Fantastic run! I enjoyed the multitude of walljumps and shelljumps. This gets a yes vote for me, but I would recommend placing this as a concept demo because a) it certainly does demonstrate a concept, and b) it might be strange to have three SMW movies on the SNES page, but one with a semi-arbitrary restriction. I do have one other question - is it possible to use multiple walljumps to scale a single wall? I've done some testing, and it's possible to reach mid-speed oscillation back towards the wall after the first walljump, but for some reason, it doesn't seem possible to snag the wall a second time. Is there any conclusive information on this? If it's possible, both forest ghost house exits could be improved by scaling the large wall blocking off the doors.

Posted: 6/4/2007 4:31 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I'm giving this topic a rather large bump because some might be interested in my progress on this game. First of all, http://tasvideos.org/1435S.html was the Mario vs. Donkey Kong 100% TAS I made (and later cancelled) awhile ago. The main problems were: -100% took, in this case, 1 hour and 50 minutes -I receieved many responses saying the movie was too boring, especially the later levels -The bonus levels took too much time and were also really boring. With that said, I did discover a lot of interesting tricks and techniques, but the general consensus was that an any% run would be a better choice for this game. I have been working on the any% version more recently, and I'm currently through 3-2. This is probably a good time to share the 3-2 WIP of my progress. I do have a few concerns, though... I'm worried that the any% run will still be too boring. I'm also debating whether going through the plus worlds is a good idea - completing the 6 normal worlds gives you a vastly inferior ending and credits (not to mention starting you on the level select screen again), but completing the plus worlds would likely double the length of the movie. Are there opinions about this?

Posted: 5/29/2007 3:01 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

My friend and I did some preliminary testing on this game awhile back ago... The main difficulty in planning a route is the large number of goals you have to complete inside each level. For example, in level 1, some of the more complex goals are to: -Get the slam move (to break open trash cans for tokens), requires 80 tokens to learn -Find the 4 screws and take them to the roller coaster -Get the Dune Dog car for faster movement and a few tickets (requires 50 tokens to use) This is a preliminary testrun up to getting 80 tokens (at that point it's best to save and quit so that we can get the slam move): http://dehacked.2y.net/microstorage.php/info/5779/RoutePlanning.m64 Here is a mostly optimized version of the first few parts: http://dehacked.2y.net/microstorage.php/info/5780/Rocket%20Testrun.m64 And, on a side note, I discovered that the race with the corn car can be done without the dune dog by riding the corn car (it's very difficult, though). I have a movie, but I can't host it at microstorage because it starts on a savestate. Edit: A few other notes about optimization: -In the first level, there are a number of extra texts. Most of these take a long time to move through, so we tried to plan out where we could cancel out the texts. The "first ticket" text at the very start is skipped by landing on the ticket switch, the "first token" text is skipped by entering the midway, and Tinker's "sensor array is cut" speech (activated by getting a ticket in the first level) is canceled by landing in the exit of the midway. -We discovered that the ticket sitting on the platforms in the water can be collected by driving the dune dog towards it, hopping off the edge of the water, and jumping out of the car onto the platform. -There is another break later on, in level 2 (paint misbehaving): we found that after activating ticket D (where you paint the giant objects; collecting it activates an event where a building emerges from the ground), you can drive the paint car from the slope above, into the ticket, and far enough forward that the rising building will carry you upwards, starting you on the roof. This allows you to grab the ticket on the roof, enter the building from above, and complete the whole room backwards! By the way, I can't promise any progress on this (it was just a side idea for my friend and I), but I just wanted to share this information.

Posted: 5/22/2007 12:16 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Wow! I never thought this would get obsoleted... The play here was amazing, and I was continually surprised by the fluidness of dispatching roomfulls of enemies at a time. I'll add my Yes vote to the heap of other Yes votes ;)

Posted: 5/20/2007 3:07 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

FODA wrote:

Guys, this is mario 64 we're talking about. It can generate more interest than 40% of the games on the site, which are quite obscure. Do you really think this will get less views than dunno, GBC Deja Vu 1&2 or NES "Sypha path" Castlevania 3 (JPN) (examples picked for no particular reason)??? I think this game deserves more than 2 runs...

I really don't want to revive this old debate... but I have to disagree here. Are you saying that more runs should be accepted for Super Mario 64 just because it's more popular? You're probably right that this run would get more views than many others on this site. But does it really matter in the long run? Most of the tricks in this TAS can be seen in the 120 stars TAS. Don't get me wrong - I do think that this TAS is of high quality, but it seems a little unnecessary to publish this when we have other runs demonstrating the same tricks and routes.

Posted: 5/19/2007 11:11 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Highlights for me: -First Koopa the Quick race -Mario spinning on top of the pyramid -Chuck-ya glitch (it was awesome seeing Mario swim with "nothing" in his hands, then suddenly BLAM he's holding a block or some other item) -Bouncing around in TTC to avoid the red coins -BLJ in Rainbow Ride to get that far-off star I think the best parts of this TAS are the stars with strategies different than in the 120 star run. The places where you wasted a little time for entertainment were good, also. Unfortunately, most of the stars seem to be collected with about the same strategies as in the 120 star run. Perhaps if taken alone this TAS is very interesting, but I found myself fast-forwarding through a lot of stars simply because I had seen them before. I can't say that I was entertained all that much by this run because most of it is already done in the 120 star TAS. I'm giving this a "meh" vote. Edit: Maybe this could be published as a concept demo?

Posted: 5/19/2007 12:40 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

A few more screenshot ideas:

Frame 135456

Frame 144264

Frame 185966

Frame 190784

Frame 191664

Frame 199624 Any opinions?

Posted: 5/18/2007 4:20 AM

Post subject: Programming a brute-force bot (that's useful for a TAS)

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I am somewhat interested in how BisqBot and other brute-forcing bots work. I understand that the bot has to check memory addresses and restart if certain conditions are met, but I don't quite understand how a few other things are done. How does the bot actually "play" the game, i.e. how does the bot advance the emulator/rom to be able to compare memory addresses? Related to the previous question, how does the bot restore a previous position? Through a savestate? I might want to create my own TAS-bot sometime, so I'd greatly appreciate any responses informing me of the basics for programming one.

Posted: 5/10/2007 3:05 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

It's a good thing I didn't complete my version of the run... you have an incredible eye for entertainment (especially misty star road!), and the play is very precise, too. Unfortunately, I wasn't as surprised about the new glitches (as I already saw them when you showed me your progress awhile ago), but I still greatly enjoyed this run. Yes vote from me.

Posted: 5/6/2007 3:10 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I was surprised at how much time was saved over FODA's run... I thought this run was amazing, and I was never bored when watching it. I personally wasn't bothered by the camera switching or the unnecessary damage, but there were one or two things from FODA's run that I missed (such as the zoomed out camera during the penguin race). My favorite levels were Tick Tock Clock and Rainbow Ride, the former because of all the fast-paced wallkicking through the level, and the latter because of all the crazy leaps Mario takes, skipping big portions of the level. "Yahoo!" is probably the best way to describe this TAS. I thoroughly enjoyed watching this, so it gets a big yes vote from me.

Posted: 5/3/2007 3:56 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I loved the last two bowser levels, it was amazing to see Mario move that fast and bounce around like he'd had to much coffee :) The graphical glitches in the last bowser level were funny, too. I was a little disappointed that the new route prevents all the glitches in Bob-omb Battlefield, but the Hazy Maze Cave glitch (almost) made up for it. Yes vote from me. I wonder if we'll ever get sub 1:15...

Posted: 5/1/2007 3:08 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Themattman wrote:

Just curious, how much time do you think that you could really shave off now since this looks like a "perfect" run?

Let's see... -I waste ~30 frames to do entertaining tricks -It's probably possible to collect the 20 rupees in better places than I did (maybe 1-2 seconds) -If ANY new trick is discovered, that's another second (or more, depending on the trick). Example: In an earlier version, I picked up the bombs from the clockbombs in the 3rd dungeon. Skipping them saved about 2/3 of a second. If a major sequence break is discovered, such as skipping the 3rd song or skipping the walrus outside the desert, whole minutes could be taken of the run's time.

Tompa wrote:

Could that be a 100% run possible? =P

If I ever get the motivation to work on another Link's Awakening run, I would have to know a lot about the planned route, whether to use the black and white version for screen warps, etc. But I might consider this in the future.

Posted: 4/30/2007 6:02 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Takarifreak wrote:

I'm used to playing the original, where I know it IS possible to enter the Egg before the owl begins his speech (in real time), since I've been sucessful a few times doing that (only maybe 5% of the time though). Is this simply not possible in the DX version? I can see you're taking advantage of moving the few pixels forward when he lands and before he talks; is that as far as you can go?

I extensively tested this, and I concluded that it's impossible in the DX version. Link can only move 3-4 pixels forward before the owl speech starts, but you have to go forward another 3-4 to enter the egg.

Takarifreak wrote:

Also, is it possible to beat the level 5 boss (what the hell is his name anyway?) in one cycle at all? Not that I'm experienced with TASing or all the funky fighting techniques used, but it sort of looks like to me it MIGHT be possible with perfect timing. But hey, I'm not the expert here.

Well, I still have to hit him 3 times in his 2nd cycle, and even that was only because I discovered the spinattack cancel trick. Part of the problem is that the boss moves back inwards really fast, so it becomes very difficult to hit him with regular slashes (since Link stops during a slash). The spinattack is the most efficient way to hit him at that point. Thanks for showing interest in the run, by the way!

Posted: 4/30/2007 1:07 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Tompa and I tried for a long time to pick up items from a distance, and we came to the conclusion that there are two types of items. Type 1 is items like hearts and rupees (the sword on the beach is somehow in this category), while type 2 is items that bring up a text box after you pick them up, like the golden leaves, the slime key, etc. For some reason, it seems that only type 1 items can be picked up from a distance. Unfortunately, there is almost no use for picking up type 1 items... one of the biggest timesavers in this run would be to find a way to pick up these type 2 items from a distance.

Posted: 4/30/2007 4:10 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

I'm really sorry that I haven't worked on this for so long, but I do have a new WIP for everybody. I hope everybody enjoys the latest progress!

Posted: 4/20/2007 10:32 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Wow, I was so sure that the previous version was already perfect... That's a really nice trick with the doors! I was entertained enough by all the wobbling (the run was short, so it didn't get boring), but the random glitches near doors was even more interesting. I'm voting yes on this run.

Posted: 3/7/2007 10:53 PM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Here's the run through dungeon 5: http://dehacked.2y.net/microstorage.php/info/4296/Testrun4.vbm There were a bunch of changes, mostly minor, that I won't bother to explain here. I was mostly more creative in finding ways to minimize the time (e.g. trying to figure out the best time to kill a group of enemies rather than killing them as soon as possible). I also discovered another trick, useful when the hookshot and sword are both equipped: if you charge a spin-attack, then press the hookshot button as soon as you release the spin-attack, the spin-attack will be cancelled. I use this trick on the boss, and I'm able to get an extra hit in as a result. Also, I've redone from the beginning of dungeon 4 (again) because of another new technique Tompa told me about - jumping while using glitched pegasus boots dashing will send you zooming in odd directions. It's usually faster to do this "super-jumping" when I have to cover the whole screen (otherwise I'll just bump into the wall on the other size), and when the passageway is more than one block wide (if I only have one block of room to spare, I'll just hit the wall instead of jumping through). Total time saved: 9315 frames, which is 2 mins and 35 seconds.

Posted: 3/1/2007 8:28 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

Done up to the Mambo song: http://dehacked.2y.net/microstorage.php/info/4168/Testrun4.vbm It's technically 10446 frames faster, but Tompa found another timesaver that I hex-edited into the movie - since I'm now using the warp hole to get to dungeon 5, it's quicker to get the Ballad of the Wind Fish on the way, rather than immediately after getting the Angler Key. It saves the time of a S&Q, or about 150 frames. So, the "real" time saved so far is ~6740 frames (the extra ~3700 frames was from skipping the song for now). Improvements include more precise movement in the dungeon (as usual), and manipulating enemies into better positions. You might see a few delays in movement; it's luck-manipulating the enemies. Edit: I'm now up to the beginning of the 5th dungeon. Instead of using the Mambo right after learning it, I instead swim to the nearby warp hole. Also, I've saved another item switch by keeping the feather/boots equipped until the beginning of dungeon 5, where I'll switch to the sword and skip the text at the beginning. Total time saved at the start of the dungeon: 7763 frames, or 2 minutes and 9 seconds. It's crazy how much time I'm saving off my old run... By the way, I'm not going to bother posting a WIP because there's really nothing interesting that I've done (since the last one I posted above).

Posted: 2/28/2007 7:48 AM

Quote

Chef_Stef

Editor, Experienced Forum User, Published Author, Active player (467)

Joined: 5/23/2006
Posts: 361
Location: Washington, United States

The issue that always comes up, then, is: what counts as a glitch (versus a trick or an exploit)? In Link's Awakening DX, the game I'm currently TASing, there are a multitude of tricks/glitches, some being much more significant than others. For example, it's easily possible to clip the edges of walls and walk a few extra pixels into them, allowing for small movement optimizations. Is this truly a "glitch"? If not, then what about a trick that allows me to clip the wall and exit a room before some event occurs (e.g. an owl flying down and giving you advice)? One may argue that only glitches that skip parts of the game should be banned. Well, in Link's Awakening, it's also possible to skip the text telling you what's in a chest (by using ABStartSelect). Boss text can sometimes be skipped in the same way. Another example: It's possible to destroy the level 6 boss in Link's Awakening before he even starts talking (or attacking you). Does that count as a part of the game that is skipped? What if it could be achieved by glitching, and it were more entertaining to do so? It seems to me that the only way to judge such a collection of "glitches" is to review them individually, and have some list that says whether each one counts as a glitch or not. This, in my opinion, is too subjective.