Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
DeHackEd wrote:
So you slid the sled into the Bospider out of RNG necessity? Okay then That's the only thing that stands out as a possible improvement from me, but manipulating that guy is a bitch.
The ice sled sliding into Bospider bears no impact on the speed of the fight, since Bospider's eye is opened for a fixed duration each cycle and you merely need to hit him once during it(except for the final cycle which of course should connect on the first frame). Shooting the sled from a distance though lets you charge up earlier to break an extra sled between cycles, which is worth 12 dust sprites worth of manipulation.
The time loss from RNG on the fight though is a product of X's poor ability to manipulate RNG at a quick rate in this game, since ice sleds and dust sprites are his only two options during this fight, and the straight drop patterns being so few and far between in the RNG sequence. The manipulation on that fight is further exacerbated by the fact that you have to go into the stage at a point in the RNG sequence which will allow you to do the first phantom grab with an extra life(small ammo will cost 8 frames and small health even more), and you have very limited options for manipulating Vile's pattern, the second phantom grab, and the first pattern of the Bospider fight during the stage. I would be surprised if there exists a place in the RNG sequence where getting an 8 straight-down pattern fight is even possible.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Obvious yes vote for lag-reducing mechaniloid punches.
The turnaround kick discovery is pretty awesome. That will definitely be useful in all other X1-3 categories as well.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
lxx4xNx6xxl wrote:
how does everyone feel about 100% Glitched as a category?
I've been curious about this for a while myself. What are all of the major glitches which are forgone in the non-"game end glitch" categories, and how could they be used to save time in a 100% run?
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
I haven't had that same issue happen in a while, so I haven't been able to find out the reproduction steps. But it was happening quite frequently before and seemingly completely randomly with no rhyme or reason as to why.
I have been getting another error recently though that throws a fatal exception when I frequently switch back and fourth between loading different branches. Here are the details.
The most recent iteration of this error resulted in all of my branches and savestates being deleted and this error message every time I tried to make a branch or save the project. Now after restarting BizHawk, I am no longer able to open my project and get this error message when I try.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
I just had an issue where all of my automatic save states were erased and then it would only keep around 70 of them at a time, despite all of my capacities being set to 1gb in savestate history settings. I was able to fix it by restarting BizHawk, but I'm not exactly sure what caused it. I created a branch right before it happened, so it may have been related to that.
Edit: Actually it doesn't seem to be branch related at all. Just had it happen again without having saved or loaded any branches for a while.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Borukta/Al Hari done.
Link to video
Possible routing improvements:
1. The shop trip in between Cenktrich and Borukta can be eliminated conserving a little bit of Upward Vulcan ammo across those two operations. Since I still absolutely need to reach pilot-level 7 before the final mission of Borukta to pick up UVG-20, this will require nothing short of measuring the amount of frames and/or EXP I lose for each UVG ammo usage I can potentially omit, and taking out those which are the least costly until I have just enough ammo to last until level 7. I won't have to go into the optional mission at the end of Borukta to kill the one helicopter either since there are some enemies I didn't kill in the first 3 operations which yield enough XP net me level 7 without it.
2. Selling my old parts to buy a second UVG-20 in Borukta ended up being a net time loss. It saves you some resupplies but costs too much time in menuing. It will definitely be worth buying a second UVG-20 after Al Hari though if I don't need Flamethrower-20 for anything.
3. The Al Hari shop trip can be omitted as well if I conserve enough UVG ammo to last me until the end of the operation, which should be easy since I used a lot of it in places where it wasn't necessary so I could level it up higher. The reason I would want to omit this shop trip is because after upgrading to the 2nd tier carrier and shopping in New York, I can guarantee I will have enough ammo reserves to last me until the Zambola shop trip, and may even be able to reach 120% UVG-20 xp before the final Ark encounter, which instantly increases weapon dps by 50% with no incremental gains in between 100-120%. If I end up not being able to reach 120% xp by Ark, or if I end up needing FBG-20 for anything, then it will probably be better to use a route that uses UVG-20 much more liberally throughout Al Hari to level it up higher, and shop before the final boss to buy VG-20 and FBG-20. It's impossible to know which routes are viable at this point, since the former might run into issues by not having FBG-20, and the latter might run out of UVG ammo, so I'm just going to pick one at this point and see if it works out. If both are viable though, then both will need to be tested up to the Zambola shop trip to find out which is faster.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Details
I already saved over the project and erased the branch that caused it. This time I got the error message by trying to paste an amount of frames that would have exceeded the movie's frame count.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
I just had another bug happen where I saved my .tasproj, loaded a branch, repeatedly got spammed with this error message so that I had to close BizHawk in task manager, and BizHawk automatically saved over my .tasproj in the state of the branch I had loaded causing me to lose another couple hours of progress.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
I can't remember what the exact steps were that made it occur the last time, other than the final thing being me trying to load a branch. It happened again though, this time by trying to go back to a point in the TAS where there were no were no close savestates, causing it to start from frame 0, and then adding input into that section.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
I've had this happen twice to now where BizHawk crashes after trying to load a branch, then the size of the .tasproj file I was working on is drastically reduced and I get this error message upon trying to reopen it. I've been backing up my files fairly regularly, but it has caused me to lose nearly a full day worth of work twice now. Is there any way to restore the .tasproj I was working on, or use branches without this happening in the future?
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
For the mission where you have to use Luven, there is a 3rd-level wanzer for him somewhere on the ground in a previous mission that takes a small amount of time to grab. I think my most recent shop trip before that mission will be the one coming out of Al Hari to pick up FBG-20, so if I can pick up a 2nd-level Luven wanzer then(and can fit it into my money route) that might be the only potentially viable alternative to getting the 3rd-level wanzer on the ground. It might even be possible to keep him alive with a 1st-level wanzer, but I'll look into it when I get there. I'm not sure if it's faster to control him over letting his AI set the bombs, but I'll definitely look into that as well.
In regards to enemy spawns, I found out that(at least as far as the enemy meter encounters are concerned) they are strictly based on camera positioning and timing and not actual spawn positions in the stage, which is why I approached most of them backwards on the 2nd Cenktrich mission to bring the edge of the camera closer to Albert(I couldn't do it on all of them since that require an extra UVG shot and I had to conserve some ammo). The minimum spawn time isn't the same for each of these encounters, but it seems like when the spawn time varies it varies by a significantly large amount, so it's generally pretty easy to find the fastest pattern. Some spawns outside of these encounters seem to work in a similar fashion since they can be moved around a bit, but not all.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Bergen 1/Cenktrich done.
Link to video
After hours of hex editing I was able to adjust my Upward Vulcan ammo/experience route into having one consolidated Cenktrich resupply where I reach 0 ammo just before and after the final mission. Since doing no Cenktrich resupply would leave me with too little UVG experience to kill the final boss before I run out, and I have 74 remaining UVG ammo to last me the first 4 Borukta missions before the shop, I am confident that this secondary weapon route is optimal up to this point.
The current plan for the remainder of the secondary weapon route is to go to Borukta and pick up UVG-20 next so that I am a high enough level to purchase Flamethrower-20 on my way out of Al Hari, skipping Flamethrower-10 entirely(my UVG-20 will be high enough level at that point to surpass it in DPS). If necessary, I can use the main base mission in Al Hari to overwrite my FBG-20 ammo since it does not have a score tally screen. I am not sure if it will be worth picking up FBG-30 on the mandatory backtrack to Al Hari since it would be an extra shop trip, potentially an extra overwrite glitch, and would only be used in 2 continents.
As for primary weapon route, after all the Vulcan experience farming I have done it has only increased by a whopping 2 damage per shot, so in the v2 TAS I am definitely not going to be wasting any frames to turn around and shoot stuff behind me(at least not with VG-10). I will pick up Vulcan-20 in Borukta, and after that I am not sure what the route will be. I believe Agwawaf's TAS used VG-20 until getting a Shotgun-30 from the carrier obtained in Sibirska, but I am not sure if there is anything worth getting in between.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Thank you Omnigamer, your script has been a tremendous help so far in the run.
Is it possible to show where potential enemy spawns are via the script? Also, can you shed any light on how the timing of spawns, particularly in the "enemy meter" encounters, works?
I have noticed that the timing of the spawns can be manipulated simply advancing RNG, but I am not sure exactly what is happening. Do the actual spawn intervals themselves vary based on RNG? If so, what is the fastest spawn time(assuming it is consistent across different encounters) and when does the spawn timer for each enemy start?
Or are the spawn intervals fixed with RNG determining which spawn location is used, and the variance in spawn timing occurs when I block the chosen location by having it on screen? If this is the case, does an enemy spawn immediately at a random off-screen spawn point, or does the spawn timer get reset?
Here's the bk2 and savestates before the first two enemy meter encounters in Cenktrich in case you want to look at it. I'm using Aeon Genesis English patch for this run since I plan to redo it anyway before I submit.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
http://www.twitch.tv/huurtheduur/v/52279424
New glitch that appears to freeze X's Y-axis speed while inside the ride armor. I don't know how to reproduce it(and apparently neither do the people in the video), but in the grand scenario it could perhaps be activated while moving upward to get the mech up on top of the roof to punch the igloo holding the heart tank and eliminate the Penguin revisit in 100%.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Ah perhaps my wording was a bit ambiguous. I am in fact getting the Power Raiser Omega whereas real-time runs do not. The reason they don't is because on the N64 version, the more common version for real-time runs, the auto aim is too aggressive which causes the larger hitbox of the omega shots to hit the spikey boss's armor while aiming at his head.
Airborne movement is faster than running but halts you for a few frames each time you hit the ground. Pre-spring boots, chaining jumps together vs simply running breaks even over long distances on flat ground, but jumping allows you to adjust your directions with much greater precision since turning has an acceleration period while in the air but not on the ground. Since diagonal jumping has a different default angle than diagonal running(relative to the camera), once you jump once to adjust your angle you have to commit to chaining jumps together until the next time you make a turn, otherwise you will have to compromise your straight-line trajectory more than once.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Actually, 60.0984755 fps would result in a 31:21.84. I'm not sure why Bsnes says it runs at 60.0984755 fps but then says the run is a 31:21.83. I guess it must be running at framerate of the SNES console, 60.098814?
I'm not sure how you're getting 60.05; there must be an error somewhere on your part.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
WIP 1: Intro-snake pit
Link to videobk2
I've finally finished planning out the zenny/upgrade routes. The only deviations from the RTA route will be forgoing the 2600 zenny hole before the snake pit as well as the Power Raiser Alpha before Tronne, and picking up the Power Raiser Omega from the shop trip after Cardon. I should be able to easily farm the 21460 zenny for that trip from manipulating high zenny drops from the 6 tanks that you fight during downtime as well as the 9240 zenny hole and 1 wasp nest. I'm almost certain it is impossible to forgo the wasp nest nor the zenny hole even with max zenny drops from every enemy in the game up to that point. Likewise, doing the bank robbery sidequest later on to steal 200k zenny in order to upgrade the Grand Grenade will be a necessity.
There are a few problems I have encountered in the run so far. One is that, according to some speedrunners, there is apparently a rare glitch that will sometimes cause you to get a 4th shot in your volley with a minimum energy stat. There is no video evidence of this glitch, and I could not figure out how to reproduce it myself, but if possible it could save some time on the first boss as well as other parts of the run before upgrading the energy stat.
Another problem I have been having is manipulating the positions of multiple enemies at once. All 8 snakes in the snake pit have random positions that cannot be manipulated individually, so I'm not sure if there is any methodical way I can determine the optimal pattern for them. This is further complicated by the fact that you can kill multiple snakes with one shot if they stack up perfectly on top of eachother. I suppose you could even kill all 7 in the 2nd room with 1 stutter-shot volley if such a pattern exists to facilitate this. This is the best pattern I have been able to find so far via trial and error, but you have plenty of opportunity to manipulate RNG prior by shooting walls with buster shots. Getting the 4 ground units in City Hall into optimal positions will probably be more difficult to manipulate than the snake pit since you have to kill the carriers beforehand, but I think it will be simpler to determine what the optimal locations for them actually are.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
I think you got the 31:24.93 figure by dividing the frame count by 60.0? Bsnes core runs at 60.0984755 fps so the 31:21.83 figure is more accurate.
Although the time save over FractalFusion/Graveworm's run is probably closer to ~25 seconds after discounting the emulation and framerate disparities between Bsnes and Snes9x 1.43. (This comparison was between different routes so some things are different but it should still be reasonably relevant)
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
Well this run needs a branch name to be published since it's slower than the currently published branchless run. I think something like "all upgrades" makes the most sense since, like X3, I think it's contradictory to include "100%" in the title of a category but then have a caveat of forgoing certain items.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
hidaigai wrote:
>Hetfield90
You said "100%" run could obsolete both the "all upgrades" and "best ending, no upgrades". I agree "all upgrades" run is obsoleted because of redundancies. But, why is "best ending, no upgrades" run? I don't think "all upgrades" run can obsolete "no upgrades" run.
The SNES Mega Man X series games don't really have many options for having more than two branches without significant overlap between them, and have generally had one "any%" category and one other category that is as far of a departure from "any%" as possible. The rationale behind a "100%" run obsoleting the "best ending, no upgrades" run is because both are/were designed to be the antithesis category to "any%". Since, at the time, the "any%" run collected all of X's parts and none of Zero's, FractalFusion made a run that collected none of X's parts and all of Zero's. Now that "any%" collects neither(aside from just the buster upgrade), a more appropriate antithesis would be a run that collects everything. "Best ending, no upgrades" overlaps with "100%" in that they both collect the Zero parts, and overlaps with "any%" in that they both forgo the vast majority of X's upgrades(with only one exception in the arm parts in "any%"). The only interesting unique content in the "best ending, no upgrades" branch at that point would be the Gator elevator skip and more strike chain usage at the cost of less interesting boss fights and less varied movement tech without the buster upgrade, which I don't believe is enough to merit a separate branch.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
First of all, this TAS is very optimized. According to route tests that FractalFusion and I did, it is about 15 seconds slower to obtain the shoryuken, so this run's strats and optimizations are consistent with my current any% run. Hidaigai also sent us his WIPs to check after each stage, so in terms of improvements, there are none which I am aware of.
As for branches, I pushed for my v1 run to obsolete the shoryuken run to show the history of any% routes since obtaining the shoryuken was previously thought to be faster. But this was also done under the premise that FractalFusion and I were working on a true 100% run that collects both X's and Zero's parts to obsolete the current "best ending, no upgrades" run. That way all of the game's options for item acquisition(minimal X parts, all X parts, no Zero parts, all Zero parts) would still be represented in as few as two branches.
After having to restart the run multiple times due to optimizations being found in the earlier stages, motivation decreased from a significant increase in difficulty to maintain the frames being saved, and progress eventually came to a standstill. I intend to pick it back up at some point, but there are two other projects I am working on at the moment so I wouldn't anticipate it being finished any time soon.
In my v1 thread, Mothrayas suggested having 3 branches("any%", "all upgrades", and "best ending, no upgrades") until such a run was completed, and that the "100%" run could obsolete both the "all upgrades" and "best ending, no upgrades" runs since it alone covers the two options for item acquisition in this game which "any%" does not. In agreement with this, I think that this run should be published as a separate "all upgrades" branch so that the site is not without a shoryuken run for the time being, but under the premise that it and the "best ending, no upgrades" run can be later obsoleted by a true "100%" run to eliminate redundancies between branches.
Experienced Forum User, Published Author, Skilled player
(1221)
Joined: 8/29/2014
Posts: 301
arandomgameTASer wrote:
Config > Customize > Advanced > check 'Frame advance buttons skips non-input frames'. Should solve your problem.
This only works in games where every other frame is actually recognized as a non-input frame. Unfortunately they aren't in the game I'm TASing at the moment(I only checked 1 N64 and 1 PSX game, so maybe it's system-specific and not game-specific?).
I just had another bug happen where I saved my .tasproj, loaded a branch, repeatedly got spammed with this error message so that I had to close BizHawk in task manager, and BizHawk automatically saved over my .tasproj in the state of the branch I had loaded causing me to lose another couple hours of progress.
I've had this happen twice to now where BizHawk crashes after trying to load a branch, then the size of the .tasproj file I was working on is drastically reduced and I get this error message upon trying to reopen it. I've been backing up my files fairly regularly, but it has caused me to lose nearly a full day worth of work twice now. Is there any way to restore the .tasproj I was working on, or use branches without this happening in the future?