Posts for zidanax


1 2
8 9 10
14 15
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Once a lair's been sealed, it doesn't really matter whether or not we bother to kill any more monsters because we'll have enough souls to last until we reach the next lair.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Sounds good with Fillmore. Please put up your progress when you're done with Fillmore so I can see what you came up with. Now that I'm done with initial run-throughs of all the action sequences, I'd like to go ahead and do another sim test to see how your change in strategy in Fillmore and my change in Kasandora strategy affects everything else. Hopefully it won't throw off anything (most importantly, the number of turns after you leave Aitos before its population grows again). Otherwise we may want to come up with something else for Fillmore.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
If it is, I'm not sure how I would activate it. In Aitos, the platform moving down past the hero causes the hero to go into the wall, which of course isn't supposed to happen. But with Marahna, the platform won't even start moving until the hero reaches the middle of the platform, so when you try to go to either wall, you just stop at the wall as expected. Also: Northwall Act 2 WIP 1 Saved 241 frames up to the boss fight. As usual, I'm not bothering to use stardust since the patterns may change next time. Dunnius, any progress yet? Now that I've done first run-throughs of all the acts (I already did tests in Death Heim earlier), I think I may do a first run-through of the Northwall sim section, continuing from the sim testrun that I posted, and see how I can do compared to Ogreslayer's run-through. EDIT: initial run-through of Northwall SIM,continuing from my second sim test: Northwall SIM test I also updated my growth spreadsheet on google docs I think my current sim test clocks in at 28 less construction phases than Ogreslayer's. I wonder if the final run might even be under an hour.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Marahna Act 2 WIP The level branches off into two paths at one point. I chose a different path from Ogreslayer because I think the one I chose is faster, but I'll test the other one later to make sure. I saved 207 frames up to the boss. I didn't bother using the scrolls I had against the boss, because I don't want to bother banging my head against the keyboard trying to figure out the best time and place to use stardust when I know I'll just have to do it again later. As a result the boss fight is 435 frames slower than Ogreslayer's. Northwall Act 1 WIP Saved 168 frames up to the boss. Again, I didn't bother to use stardust, so the boss fight is 221 frames slower than Ogreslayer's.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Finished Marahna Act 1. I managed to save 177 frames. A relatively easy level to TAS, I'll move on to Marahna Act 2 and maybe even Northwall tomorrow. Marahna Act 1 WIP 1
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
The question is... will I continue to be that lucky with stardust patterns? I wasn't so lucky when I went through Kasandora Act 2, even after waiting for a bit outside of the boss's room, so the fight was slower than Ogreslayer's. We can hope that things turn out well when the run is put together,during which I'll play through the levels one more time to see if I can squeeze any more frames out of them (I find it works well to compete against a previous run I've made). Interestingly enough, the Aitos boss fight took almost exactly the same amount of time from entering the boss's room to the boss's death as Ogreslayer's fight, except I only cast stardust once. Just turned out that way.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Finished Aitos Act 2. Saved 154 frames. I may change my strategy when I come back to this stage, since you have to decide where you want to get hit to bypass stuff, and where you want to bypass monsters through other methods. You don't have enough life to get hit as many times as you'd like. By the way Ogreslayer: I'm impressed that you managed to come up with that method for bypassing the last (5th) red blob in the big corridor without getting hit. Especially since you didn't have frame advance. It was a pain for me to get it to work, so I ended up following your smv very closely in that little segment. It seems that the timing and position of the landing on the platform to the left of the blob has to be absolutely dead-on or else you get hit when you try to attack the red blob. Aitos Act 2 WIP
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Nice find with the modified construction pointer movement to save time. I have an idea for something to experiment with in the sims.We know that where people build depends on where we move the construction pointer, as shown by having to move the construction pointer in a particular way in Fillmore to avoid a fire. Does the length of the "Town Under Construction" part vary depending on where people actually build? If it does, then I wonder if it would be possible to save time by the leaving the construction pointer in particular places while waiting for the people to build. Naturally, the "Town Under Construction" part would have to be shorter enough to make up for the extra time spent moving the pointer. Also keep in mind that the 32 frame estimate I threw out is just that, an estimate. For all I know, optimizing both your strategy and mine would show the difference to be bigger. Even so, I doubt the difference would be more than 600 frames, which is the time gained in Kasandora using my modified Kasandora strategy, which requires leaving Fillmore earlier (e.g., with your strategy). EDIT: Here's Aitos Act 1. I managed to save 514 frames. I actually managed to save frames on the part where birds carry you over. Probably lag reduction because I spent more time on the ground. I also managed a very nice time-saver elsewhere. Perhaps one of you guys have already discovered it, but I was rather pleased, because I'd been wondering if it was possible, and I barely managed it. You'll know what I'm talking about when you see it. :) Aitos Act 1 WIP 1 Also, it turns out that it's possible to zip in ActRaiser: Zipping example. Unfortunately, I'm not sure where I'd be able to use it. Unless you guys can think of another way to jam into walls where it would actually be useful instead, of, um killing the character.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Yep, I know about that shortcut. In fact, I think you may have told me earlier. I finished Kasandora Act 2 and saved 376 frames. It would've been more, but I don't want to spend tons of time manipulating stardust patterns for the boss since the exact patterns may change when I go through this level again. At the moment, the boss fight is about 130 frames slower than Ogreslayer's boss fight. The waiting before entering the fight was an attempt to manipulate the stardust patterns since waiting before entering does change the patterns, but it only helped some. Kasandora Act 2 WIP 1 EDIT: I looked at Ogreslayer's run and remembered that you can build a road over the square in Kasandora that has a pyramid... even while there's still a pyramid there. So I thought "Why not use a route that doesn't cast rain on the squares beneath the pyramid and northeast lair?" I made an SMV of that strategy, and unless I'm messing something up, it seems to save about 600 frames. The catch is that I can't get a 5th house in the beginning phase of Kasandora, and the rate of population increase is slightly different, so I can't leave Kasandora one phase early. But with Dunnius' Fillmore strategy, we'd have a few extra people, so we'd still be able to leave Kasandora one phase early. Even if Dunnius' strategy is slower playing through Bloodpool/Fillmore, it doubt it's 600 frames slower. Any loss in time in Bloodpool/Fillmore from Dunnius' strategy will probably be more than made up by it letting us use my faster strategy in Kasandora. There are just a few things to check: Can Bloodpool support the extra population? How soon after you leave Kasandora will Kasandora experience another construction phase? What about Aitos? It's vital that Aitos still grow on Marahna's first construction phase. If those things turn out OK, then we will almost certainly use Dunnius' Fillmore strategy. Here's my Kasandora strategy.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I doubt that waiting in Fillmore and picking up the scroll would be a good idea. The time to enter Fillmore, plus the time to exit it, plus the time to go through the messages they give you (One of which I manage to skip before leaving for the first time, but wouldn't be able to skip if I came back), plus the time to pick up the scroll would nullify time saved from picking up that scroll. The time for those events is about 700 frames. I have not been able to save that many frames with any single boss, and I will be picking up some other scrolls that will be used against non death heim bosses. When scrolls are effectively used , one scroll will generally provide the majority of possible time savings against any particular boss, with any second scroll generally providing less than 100 frames in time savings, if any at all. I would like to stress how little time most scrolls save in comparison to how long it takes to pick them up, once you can slash at the speed of TAS, so to speak. EDIT: Currently working on Kasandora Act 2, and I just wanted to share a funky bug I found. I doubt it'll save any time, but it looks kind of funny. You might want to play the smv in slow motion/frame advance. Kasandora Act 2 bug
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I did a quick check of how the two sim strategies for Fillmore compare when Bloodpool is taken into account. I tried to not build a road to the left of the bottom lair, using your strategy. When I did that, and tried to measure how long it would take to wait another phase in Bloodpool, as would happen with your strategy, I got that your strategy would take about 32 frames longer when time spent in Bloodpool is taken into account. But that's only a rough check. The best way to compare would be to try to optimize both strategies all the way to the end of Bloodpool and compare time spent in those two sims. Whichever way it goes, the time difference between the two strategies will probably be very small. EDIT: In the meantime, I've been TASing some of the action sequences by themselves. I've done Bloodpool Act 2 and Kasandora Act 1. Bloodpool Act 2 I actually lost 4 frames, but I didn't have any sources of magic. I would've actually saved time if I could've gotten the boss to appear off of the floor twice in a row right at the beginning of the fight. Kasandora Act 1 I saved 236 frames here. I talked earlier about making some basic lua scripts to mess around with casting stardust. Even though the one for waiting a variable period of time was only semi-useful when I tested it against the Bloodpool Act 2 boss, it actually worked rather well here as long I moved the character around to different positions to try the script. I actually saved 28 frames in the boss fight in spite of only having one source of magic instead of two. Having done WIPs like this will make the work go a lot faster once we actually put everything together since I'll have a good idea of what to do in each stage. I might even be able to hex some of these in, though I wonder about the boss fights, since some of them use stardust. But even just hexing in the stage up to the boss fight would be great.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Thank you for putting up an SMV. Now I can see exactly what you're doing. I looked at it, and I like your idea, since it would allow us to get out of Fillmore one phase earlier, and leave with 114 people instead of 106 people like the other strategy for leaving Fillmore earlier. There are just a few things I'm concerned about: how much time is used up moving the construction pointer? Whenever you move the construction pointer, the construction timer stops. I'll try to find out how many frames are spent moving the construction pointer in comparison with my strategy. It probably won't be a huge difference, and the fact that you skip that message about the rock will reduce the difference in timer stoppage, but it's worth a check. Also, (and I don't think this is at all your fault) for some reason when this strategy is used Fillmore's first out-of-town construction phase occurs after two phases in Bloodpool instead of one like in my strategy. If you look at my population growth table for my second sim test, you'll see that Fillmore grows yet again on the same phase that the population of Bloodpool and Fillmore combined goes over 400. In my strategy, that second out-of-town growth in Fillmore isn't necessary to get out of Bloodpool by it's 8th phase (14th overall), but in your strategy it is necessary in order to get out by Bloodpool's 8th phase (13th overall), because you leave Fillmore with less people. But in your strategy, Fillmore's second population increase in Bloodpool happens in Bloodpool's 9th phase, not it's 8th. So you end up having to wait one more period in Bloodpool, which cancels out leaving Fillmore one phase early. It's just that this time you leave Bloodpool with 16 more people, which according to my sim tests won't be enough to save construction phases anywhere else. Given that, I think which strategy we choose will probably depend on which one delays the timer less. That reminds me that I forgot to put up my document listing memory locations (well, actually there's more than memory locations, but I haven't bothered to cleanly separate stuff). I'm really sorry that I didn't think to put it up earlier, because it could really come in handy while doing testing sim stuff, if you haven't already found the memory locations. ActRaiser Memory Locations. EDIT: If we're going to have a fair comparison of the two strategies, though, we'll need to make sure things like road placement are optimized. One thing that comes to mind for your strategy is that the road just to the left of the southern bat lair is not necessary, because Fillmore won't grow big enough to use that, judging by my sim tests.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I'm not sure I'm going to be able to shave off any more construction phases reasonably. Getting out of Fillmore one phase earlier would get us 20 more people due to spending that phase elsewhere with highest-civilization-level construction, but that still leaves 48 more extra people to get by phase 40. We could try to build simultaneously towards both lairs in Marahna, thus bypassing the middle civilization level. But that would get us, max, about 36 more people (and probably not that much due to food), not to mention it would take a lot of time to move the construction pointer that much (remember that the construction timer stops while you have the angel direct the construction pointer). We might be able to get more houses built at the beginning of Bloodpool and/or Fillmore and/or Kasandora, but to do so would require taking advantage of methods of movement that stop the construction timer in order to kill monsters as soon as they get out of their lairs. Getting three more houses this way would, like skipping the middle civilization level in Marahna, suck up a lot of time that could've been spent letting the construction timer increment. What about out-of-town construction? By the time the population has reached 1900 in Marahna in my 2nd sim test, every town (other than Marahna, of course) has gone through out-of-town construction no more than three turns ago. But towns take *6* phases to build while you're not in them, unless you've just left the town. So out-of-town construction will not save the day. In short, I think we might be able to shave off one more phase, but the tactics involved would require enough stopping of the construction timer to nullify the benefit. I think I may proceed with the current basic sim strategy, but of course make sure to optimize things like road building.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
DarkKobold: I'm a bit confused. It doesn't matter what direction I point the cursor if I want to build roads in both directions in one turn. As long as there are houses/fields on that second up road, I could build a road to either the right or left, then turn around and build one in the other direction, all in one turn. Sealing the lair simply freezes the timer until sealing is done, if you're wondering. So I build a road to the left, seal the lair, then go to the right, all in one turn. I could've first built a road to the right then gone to the left to seal the lair, then go back to the right again, but that would've been a waste of time because I would still have to wait another turn before I could build another square to the right, just like before. Am I making any sense?
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Oh, I'm sorry, I probably caused some confusion by forgetting to mention that what's to the right of each slash is not population, but food. Keeping track of that makes it easier to tell why the population grew by a certain amount.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I'm not sure that will make much of a difference. For each stage, I will get at least 3000 points (because I've got 3 lives) plus whatever time i've got left (10 points per second). Even if you don't take time into account, I get 120 souls for playing two Acts. Fillmore's minimum number of souls is 500, but its population isn't that high by the end of Marahna. Bloodpool's minimum is 330, but 6000 points = 120 souls. 330+120=450, and the max population Bloodpool reaches in this WIP is 422. Kasandora's min of 450+120=570, and I reach 506 in this WIP. Aito's min of 320 +120 =440, and I reach 402 in this WIP. Marahna's min in 210, and I reach 198 in this WIP. And of course, the number of souls I earn is actually somewhat more than that due to points of earned from time left.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Dunnius: I would really appreciate if you would post an SMV--it's much easier to understand your strategy when I can see it in action! On another note, I haven't changed my Fillmore strategy yet, but I realized that I could get to the upper-left lair in Bloodpool earlier without losing much time to moving the construction square or whatever you call it. Thus Bloodpool reaches its highest civilization level earlier. That makes the population grow more quickly enough in Bloodpool that, in combination with managing to get one more house in the first construction phase in Kasandora, I managed to leave Kasandora one turn earlier. I also managed to leave Marahna one turn earlier, for a total of two turns saved over my 1st WIP. So that you guys can see what I'm doing, here's my 2nd WIP: Note that I didn't bother to fine tune some little things like how many roads actually need to be built to get the growth we want without making excess roads. You'll notice some missed arrow shots. The action sequences are played in real time. The point is to figure out the best general strategy for each sim, since what you do in each sim has a "ripple effect" on how things go in the later sims, which makes it a pain. And a spreadsheet of this 2nd WIP.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
dunnius: I can see how getting out of Fillmore quicker might be good, since the extra turn spent in Bloodpool gets more population growth since Bloodpool is at it's highest civilization level at that point instead of just the 2nd. But in order to get the people to build a house by the river in Fillmore, I had to manage to collect 24 souls on the second turn. I did it, but I ended up taking advantage of the fact that moving while moving the construction square around doesn't drain the timer. Not exactly the most efficient way to do it, and I'm not sure the overall population gain will save a phase. Have you found a good way to get through Fillmore with the minimum of turns? Maybe you should upload a WIP if you've got one? WIPs can be uploaded to http://dehacked.2y.net/microstorage.php
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
System Error: Look again at the rules: "However, there are certain games with unlockable modes that can only be accessed if a save file (or an otherwise "dirty" SRAM) is present. If you really wish to submit a movie made on such mode, you will need a verification movie first. Any input file that starts from power-on (for example, a previously submitted or published movie for that game) and creates the exact circumstances for your submission to sync will generally do. Note that you don't have to optimize the verification movie: it only serves as a save or SRAM generator that makes it possible to claim the legitimacy of your effort." So it looks like someone could submit a run of Special or Professional mode if they also submitted a movie that plays through the game to unlock the mode. Runs that start from saves have been accepted before (e.g., the Chrono Trigger New Game+ run). EDIT: in any case, I'm on a roll and would like to see just how fast the sims can be done, even if we eventually just end up with a run that does Special mode. EDIT: A copy of the growth spreadsheet for my first sim test run http://spreadsheets.google.com/pub?key=p9d2yHwYpAENushcg3Dv-dQ That last phase will be gone in this next test run now that I have more people in Bloodpool when I leave on the same phase as before. Looking at the spreadsheet, it doesn't look like leaving Fillmore earlier will help, not that I think it would be harmful either. Or maybe I'm missing something.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Normally when building towards a monster lair, one's instinct is to not go off on any tangents but instead go straight to that lair. But sometimes you can go on a brief tangent and build a piece of road somewhere else in order to reach another lair faster once you're finished with the one you're heading toward. Look at the beginning of Fillmore in my Sim test. Notice that I build pieces of road both up and down right at the beginning even though I'm heading towards the two North lairs. That way, when I head towards the south lair later, it'll take one less turn because by then, someone will have built a house or field on that square which I put a road on earlier.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I'm not sure how much the sim improvements will be applicable to a real-time run. Some of the improvement stems from cheat roads, but a fair bit also stems from being able to gather souls more quickly in the beginning of each sim. I tested Bloodpool a little while ago, and was able to get to the last civilization level in one less turn. Unfortunately, that doesn't mean that I get to leave Bloodpool sooner, but it does mean that when I leave Bloodpool has 16 more people, which is enough to save one phase later on. EDIT: here's what I've got so far for my second sim WIP. It's up to the end of Bloodpool, but I'm going to redo Bloodpool and see if I can get away with removing any roads. Go ahead and fast forward through the action sequences, unless you'd like to see me play really badly :). Sim WIP 2 Bloodpool
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
I'm guessing you went ahead and built another "cheat road" to the south after the first one. I tried that a bit earlier and was able to leave Fillmore on the same phase that the Hero reached Level 2, the bare minimum to leave Fillmore. I haven't tested going through Bloodpool after leaving Fillmore like that. Maybe you could put up a WIP when you're done?
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
A source of anxiety for me has been how to handle the sims, since what you do in one sim depends on what you do/did in another. I remembered that the only part of the sim that the action sequence influences is the max population. Indirectly, of course, it also influences the sims because you have to decide which sources of magic to pick up. In any case, it occurred to me that I could just play through the action sequences in real time and then play the sims precisely in order to more quickly go through the sims and test certain strategies. I finished a test run of the sims up to becoming Level 10 (high enough to enter Northwall). I managed to save four construction phases over Sir VG's run, and I may be able to save another if I can build two more houses in the beginnning of Bloodpool and/or Kasandora. Also, it looks I may not need that extra square in Fillmore after all. When I visited it after finishing Marahna, Fillmore was still able to go through one good-sized construction phase. I can also see where I may be able to eliminate unnecessary roads in Aitos, Kasandora, and Bloodpool (not that much of what I build in Bloodpool is very out of the way, so it wouldn't save much time in that case). I'm gonna do a second test run with some changes, based on the first test.
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
You worked on your own SMV editor? There are already several editors available at http://www.bluetoaster.net/emu/tools.htm Or can your editor do things that the other editors can't? If so, I (and other SNES TASers) might be interested in seeing it! By the way, I think I did move down before casting lightning in WIP 5. EDIT: What problems were you having with doing the action segments? One thing that would help is memory watching, if you aren't already using that. That way you can keep close track of thing's like the hero's position. Another one is to make a table marking the number of frames it took you to reach certain points in the game, and the number of frames it took you to go from the previous point to the current one. That helps you keep track of how quickly you're going through parts of the level. Keeping a table has been very helpful in trying to match Ogreslayer's level times, which are very good (he said he put a lot of effort into the action sequences).
Experienced Forum User, Player, Published Author (215)
Joined: 2/12/2006
Posts: 373
Location: Oregon
Experimented with lua for manipulation of stardust fall. I haven't used lua before, so to get my feet wet, I tried a script that made the hero wait in a spot a varying amount of time before casting, and another that makes him move a varying number of positions to the right before casting. And I can see that writing an optimal script (if I even end up using lua) won't be easy, due to the sheer complexity. The boss teleports all the time, there are both different possible positions for the hero and different amounts of time he could wait, and sometimes slashing may actually be better. More testing of stardust's patterns definitely needed.
1 2
8 9 10
14 15