I don't know if this was mentioned, but I always thought fan translations were taboo because the patch being used makes obsoletions more difficult.
What I mean is, if someone uses a 90% complete patch to make a TAS, and someone wants to obsolete that TAS while a more updated patch exists, what should they do? Similarly, what if there are two or more complete fan translations of the game?
And, after reading Tootache's post, you have to consider the possibility of other changes the patch might incur. For instance, the patch might lower the hp of every enemy in the game by 1, which is a hard thing to notice. I doubt a fan translator would do these types of things, but who knows.
At the same time, I'm all for translation patches if people can get past these challenges they create. Even if you can't read text, it's always nice to know what's happening in terms of menus and items.
If you put Pent and Nils in slots 4 and 5, and put dummy units in the remainder slots, then they will end up in position. I just tested it a moment ago.
If only this game had a dash button :P
That was pretty entertaining regardless. I liked how you often unexpectedly damage boosted onto a platform, that every boss died in glitchy rainbow colored fashion, and of course the wall clipping glitches.
I wasn't expecting much but I found that surprisingly entertaining. As long as you can keep some variety in the combat a full run should be worthwhile. How long would you estimate a TAS would be?
Don't expect it any time soon. With the level of precision they're aiming for it will likely take year(s) to finish, and that's if the team doesn't lose motivation, which is very common in large, ambitious TAS projects.
SoulCal- It would probably be best to wait a while before trying to continue TASing the game.
Problem list:
-I doubt there's an input plugin for dolphin, which is essential for optimizing 3D games.
-You haven't yet cracked the RNG, meaning getting a ganado to drop a grenade is a complete and very futile guessing game. I doubt dolphin has a functional RAM search, but if it does then you can ignore this point.
-No real discussion of new tricks/techniques.
-Most importantly, massive problems with synch/playback/encode rendering, and even nastier things like save corruption.
Diggidyo's Megaman 10 run's got us all excited about the reality of Wii TASing, but keep in mind that was a relatively simple 2D game. The 3D realm is a whole new can of worms.
I don't want to discourage you from experimenting, though. It's great to see short concept demo type TASes to get excited for what the future has to offer.
I always preferred the wii version for the superior aiming, and I'm also pretty sure it has less loading times. It would probably be easier to TAS because you can precisely aim in fewer frames, assuming wii remote movement is supported.
Absolute yes vote. This run was sensational. I don’t understand how people can vote yes on snorefest GBC and NES TASes while voting no on a gem like this. With this new speedbooster exploit, this game is starting to look more like Sonic than Super Metroid. It was truly a pleasure seeing room after room zipped through at a furious pace. I just wish this game weren’t so faithful to Zero Mission as to take out bosses like Phantoon, Spore Spawn, 2nd half of the Mother Brain battle, etc.
Wow, nice! I'll be watching this shortly.
I know that at some point, somebody wrote a LUA script which fastforwards door transitions and item-collection music. Could somebody find that script? Thanks! (In fact, if someone had a script that only skipped item-collection music, that would be even better.)
while true do
istop = memory.readbyte(0x7E0009)
dstop = memory.readbyte(0x7E05F5)
estop = memory.readbyte(0x7E0E18)
-- Items, Doors, Elevators
if istop+dstop+estop > 0 then
snes9x.speedmode("maximum")
else
if istop+dstop+estop < 1 then
snes9x.speedmode("normal")
end
end
snes9x.frameadvance()
end
Where it first says snes9x.speedmode, you can change the settings to “turbo” or “nothrottle” if “maximum” is too fast for you.
I tried to make an encode running this script, but the act of encoding seems to disable the effect of the script :( Any ideas why this might be?
I don't know, every room looks fast but it's overall quite long and tedious. I had trouble staying awake. Like other hacks I've seen, the map layout is neat but there aren't many enemies in your way and so it's really just an orgy of running and jumping really fast. Meh vote.
This is why the turbo script should be attached to the submission notes :P To each their own, but this is one of the few runs I got through in one sitting without getting bored, which is saying a lot because many submission bore me in the first few minutes.
I'm not really in a position to comment on the technical aspects of this run, but some of your comments allude to your own thoughts on the matter:
Run comments wrote:
If this run's fall into Cpadolf's or Kriole's hand, I'm sure it will be better ;)
This is basically a big red flag saying "I think this run can be improved". Why would you not have run this by one or the other of them before submitting it?
(Incidentally, looking into the thread for the hack turns up "it's not optimized at all". I don't think there's any real excuse to submit a run you don't think is optimised.)
As a TASer, I know that authors are always their own worst critics, and Hoandizi is clearly a particularly humble guy based on his posting nature. He’s simply not giving his own work enough credit, and there’s absolutely no reason his run should be rejected because of his own modesty.
Yes, I don’t doubt that if we had some superteam of Taco/Kriole/Cpadolf/HeroOfTheDay/NameSpoofer/Saturn working on a 100% run it would be better. But here’s the problem:
-As someone who’s watched a lot of Super Metroid runs, I believe I am in a position to say this is sufficiently optimized for publication. You don’t need a complex run of a complex hack to be subpixel optimized on a first submission!
-Who’s to say anyone will ever develop the motivation to obsolete this? As technical quality increases, the amount of time that goes into a run increases exponentially, and often this has a paralysing effect on run progress. Far too many popular games suffer this affliction. Why spare viewers of amazing runs because a slightly better one could be made, but never will?
-If the author were arrogant and boasted about the optimization level of his run, or didn’t comment at all about the quality of his work, would you be prompted to vote differently?
-As other people have pointed out, rejecting a run for having known improvements doesn’t hold water anymore, nor do I think it should. Nobody should be expected to re-do an hour long run to save a tiny amount of frames at the beginning of the game, for instance. Such a rule should only apply if the improvement is sufficiently significant or easily editable, and I don’t think either applies to the 14 frame countdown trick.
Turn 1: Have Pent use the rescue staff, then Marcus can rescue him. That might take him out of dancing range, though. I'll play around with it (made a savestate before the chapter starts) and report back if I find anything interesting.
Edit: It works from a spacing standpoint, as seen in the following screen:
Someone would have to rescue Ninian, obviously.
Problem is, even if rescue staffing saves a turn, it would mean Florina can't reach the village with the warp staff, so scrap the idea altogether.
Thanks for the info, though I already knew most of that ;) I was just considering hexing that movement trick into my FE8 run, but at 4-9 frames per use, it hardly seems worth the effort. Chalk it up to "known improvements".
I rewatched the last chapter of the WIP, and I'm wondering why Pent doesn't use the rescue staff on Marcus, or maybe some other useful unit (Ninian?). It seems to me that it might save a turn.
I decided to have a detailed look at both Roland’s run and the Nico Video one, and compare them. The bolded parts are where Roland might be able to improve.
Lyn Mode
Prologue: Not really comparable because Roland has to go through some tutorial nonsense, whereas the Nico run does not because they’re playing from a 2nd playthrough.
Chapter 1: Nico run gets 2 criticals, while Roland gets 3. Again, it is hard to compare because of tutorial silliness.
Chapter 2: Roland slaughters the nico run here. He not only saves a turn but also avoids facing one of the bandits.
Chapter 3: Roland’s strategy results in more player phase kills, which means more enemies are 1HKOed without a chance to counter. It’s almost undoubtedly better.
Chapter 4: Roland breaks the wall, which allows the enemies to crowd the chokepoint sooner. The nico run also fails to use the glitch that lets multiple units stand on one tile.
Chapter 5: Roland removes Sain and Erk from the strategy, which works better with proper RN manipulation. He wouldn’t be able to include Erk anyway because of tutorial restrictions.
Chapter 6: The nico run has Serra get the door key, which one of the cavs trades for. It’s faster to just have a cav get the door key directly. I also have no idea why Wil is used in the strategy. On the other hand, they handle the southern door better by killing the soldier on the blue tile on player phase.
Chapter 7: The nico run has Rath kill the boss on enemy phase, but Florina is perfectly capable of killing the boss herself on player phase.
Chapter 8: The nico run has Florina drop Lyn instead of Kent, and this is superior because Lyn 1HKOes cavaliers normally and has a higher crit rate thanks to her mani katti sword. Otherwise Roland’s strat is better because he saves a turn by having Florina attack twice on turn 2. It might be worth testing to see if 3 turning is better since dancing on Florina takes quite a long time, possibly more than seeing a ballista attack + extra player phase screen.
Chapter 9: Roland repositions 2 units whereas the nico run only repositions 1.
Chapter 10: Basically identical.
On a more general note, there seems to be a lot of effort put into raising Lyn, which results in significant time lost in manipulating quality level-ups for her, seeing those extra stat-ups on screen, and using inferior strategies that maximize her killing potential. It’s also clear that the RN manipulation is worse, which is especially apparent on enemy phase crits that Roland gets and the Nico run doesn’t. They also don’t burn RNs as efficiently as they can.
Eliwood Mode
Chapter 11: The nico run does a lot of wrong on the first 2 turns. First off, they rescue-drop Eliwood on turn 1, but there’s no reason to drop him at all because Marcus can easily kill the bandit he faces. Then on turn 2, they get an unnecessary stat booster, and use 2 regular hits to take out an archer instead of a crit. The rest of the chapter is worse as a by-product of Marcus’ inferior positioning.
Chapter 12: Unnecessary stat-booster acquired by Rebecca. However, the Nico run uses a rescue chain to bring Marcus ahead 2 spaces, which seems to save a turn.
Chapter 13: Roland gets Marcus and Eliwood across the river with a rescue chain, whereas the Nico run takes the longer route that involves destroying both snags. Roland also saves time by not recruiting Guy.
Chapter 14: Roland repositions 4 units, whereas Nico repositions 1. There’s probably a good reason for this, though. The nico run clearly uses more units than necessary, doing things like dropping Hector in front of an armor for a wolf beil 1HKO when Lowen can just 1HKO with a crit.
Chapter 15: Mine glitch.
Chapter 16: The Nico run has this nasty habit of using more units than is strictly necessary. There’s no reason Priscilla should be following Eliwood and Marcus to the throne, for instance. Despite only using Marcus for the majority of the throne charge, Roland saves at least 1 turn and massive amounts of time.
Chapter 17: The Nico run saves a turn, but loses huge amounts of realtime by having to fight a bunch of enemies.
Chapter 18: Roland has Florina kill the boss directly instead of dropping Marcus.
Chapter 19: Again, Roland uses fewer characters yet manages to achieve a more efficient clear.
Chapter 20: See above. The nico run loses a turn by not critting the boss on enemy phase.
Honestly, I don’t feel like comparing the runs anymore. There’s a lot of signs to me that the runner doesn’t plan out his run carefully. He gets a lot of unnecessary stat boosters, he often feeds units like Lyn and Florina kills that waste time just to give them perfect level-ups (which in itself wastes time in excessive manipulation), and he seems to want to use every last character he deploys, even though that often wastes time. Although I did spot some instances where the Nico run might be better in Lyn mode, Roland is just destroying their run in Eliwood mode.
I mean no disrespect to the Nico runner. Obviously I applaud him for creating an FE TAS, but from a purely efficient standpoint, Roland’s run is much better.
As for the TAS itself, there are some strategies that look to be faster than what Rolanmen has done, but the majority of it looks improveable […] There are some lessons to be learned from this TAS though.
I think it would be better if you pointed them out, though I guess you did spot 3 of them in a later post. Is there anywhere they improve from C21 onwards?
For those that weren't on IRC, all three of the GBA games have the camera scroll on where the cursor icon is located, but the arrow while moving can move beyond that location. Normally, holding B while moving keeps the cursor moving fast enough to keep up with the arrow. The problem with this is that once a move location is confirmed, normally the camera scrolls back to the unit before moving if you go too far from it.
With this new trick, by not holding B, or only holding for a few frames, you let the camera move slow enough that, when you confirm the unit movement the camera does not scroll back and the move happens straight away rather than waiting. It doesn't work in every instance, but it will save a good few frames when it does.
Crap, I can’t believe I didn’t find this myself! I’ve even used this trick in my FE8 TAS when killing Aias (the C14 boss that Seth double crits on turn 1), thinking it was just a neat entertainment trick. How many frames does this save per character movement, approximately?
Carl, that was truly amazing. Some highlights for me:
-Juggling 4 eggs
-Shooting all the ghosts to form a vertical line
-Eating the circular arrow block and glitching on top of it
-Wall jumping antics in general
It's great to see a newcome tackle such a difficult game, no less do it this well. My only concern is that you go overboard with shooting Yoshi's tongue vertically, but otherwise keep up the good work!
I don't know if I'm just crazy, but the run also kind of felt unoptimized to me. Whether that's true or not is up to SM64 experts to figure out I guess, but I didn't see the polish I'd normally expect from a SM64 run.
I agree with the general sentiment that the run doesn't showcase enough of the game, it's just bowser fight -> blj -> bowser fight, and it's pretty repetitive to watch. Voting no.
If you're using firefox then NicoFox is a good way to rip videos from Nico. It's also worthwhile just to make an account with the help of google translate.
The link here explains all the lua commands that VBA supports. However, I noticed there isn't any sort of command that messes with emulation speed, whereas SNES9x does ( snes9x.speedmode("") ).
Is there simply no such command programmed yet or is it missing from the guide?
@Darkkobold: What you say sounds pretty reasonable.
Still, it scares me a little the way the entertainment industry is evolving. It seems that large record labels make very shitty music videos on purpose just to see what they can get away with, and when it works, they continue to lower the bar, thus creating videos that insult the audience at every turn.
For those who don't believe me, watch Todd's pop song reviews (a guy on ThatGuyWithTheGlasses, very funny).
Songs like "Whip my Hair" and "Dirty Bit" show that music labels are intentionally trolling their audience, and making huge profits because people seem to only care about a catchy 3 second loop, while lyrics and subject matter get tossed to the wasteland. Has there ever been a time where songs without verses and 3 word choruses were considered acceptable music? When voices are so warped they don't sound human anymore? When there's clear evidence that a 6 year old's brainfart could produce better lyrics than grown songwriters? This isn't just true of the music industry either. For instance, look at the writing and acting quality in a movie like Transformers. It's as though filmmakers will sink to the lowest common denominator possible so long as there's flashy CGI and a cool trailer to back up their big budget blockbuster.
Despite sounding like a rant, I'm not really angry about any of this, it's just a reality of life, and I'm slowly coming to terms with the sheer absurdity of the world.
I don't think you guys are following. Rebecca Black got signed to Vevo, a major record label. She is probably going to make millions... for making a music video so bad it went furiously viral. In essence, she's going to make more than a surgeon...for trolling the internet.
Doesn't that make your blood boil with rage? The fact that pure irony is making people rich and famous instead of, you know, talent?
Proof
She's actually going to become a pop star... FOR TROLLING THE INTERNET.
I want to become a pop star by trolling the internet now. Someone write me some lyrics.
It's not the glitching that impressed me, the pace of the run just felt overwhelmingly fast. I tried watching Taco and Kriole's any% run after to see if I just had SM fever, but I found that boring in comparison, probably because I've seen regular SM TASed like 20 times.
Someone would have to rescue Ninian, obviously.
Problem is, even if rescue staffing saves a turn, it would mean Florina can't reach the village with the warp staff, so scrap the idea altogether.