raocow's talkhaus

alex2 wrote:There's so much to do and it's overwhelming me because I don't even know where to begin.

Well it is A2MT

You should have that Solid Snake final smash move from Hopeless Masquerade.

alex2 wrote:I cannot for the life of me remember how I made the boss not lag.

I overclocked the ROM using fastROM but that only fixed part of the problem.

I remember you mentioning calculating collision every other frame instead of every frame.

Well, there's my attempt at being useful.

Draw by hardware (after having $0200~$0460 assigned somehow by the routine) doesn't seem to be an issue for speedup, which is opposed to what I claimed before. And there's no the cost difference between drawing 8*8 and 16*16 tile I guess?
The cost of calculation for each bullet is rather crucial (col. detection, position updating, OAM things). It's pretty obvious that (assuming there are 60 bullets) optimizing one clock in bullet routine is 60 times more effective than a clock in normal routine. I stupidly overlooked it. But I have improved my bullet routine since realizing it, and the result is kinda satisfying.

How do you get an OAM slot for each bullets? From seeing your code I'm guessing that, for each frame, you somehow find the first OAM index then repeat adding 4 to the next index, don't you? The average cost per bullet is about only 4 INYs and I believe there's no better method than it. Hopefully NMSTL doesn't affect the average too much.
Other OAM calculations seems pretty lighter than mine, as it seems that there's only a byte for each x/y pos and doesn't deal with the case of partially horizontal offscreen. Having two bytes for each coordinate seems beneficial to performance, while mine's having three bytes - high/low/frac.

What I'm suggesting is to set the bank to #$7F(LDA #$7F : PHA : PLB) and then to replace $7Fxxxx with $xxxx. Its overhead is setting the bank to $7F and recovering the bank occasionally(to access ordinary ram in bank $7E, for example when JSLing $00f5b7). Accessing $7Fxxxx is everywhere in the bullet loop so it's worth trying.

alex2 wrote:Yeah that's correct. It's just a simple OAM search. It's actually pretty inefficient: it could be improved by treating OAM like a stack, but I don't really have any interest in doing so right now.

So, is the meaning of a stack is to draw a bullet in front of older bullets? I'm guessing like that because data in the OAM table isn't preserved - OAM is cleaned up at the beginning of every frame (except when you paused the game) - and you need to fill OAMs for all bullets every time anyway, which implies just doing INY 4 times would be the fastest way (or making each bullet's OAM index fixed). Correct me if there are something that I misunderstood.

alex2 wrote:IT SURE WOULD BE NICE IF THE SIX PEOPLE WHO VOTED FOR OFFENSIVE WOULD DEFEND THEIR CASE

they don't know what the choices mean
^ defense of their case-wait that's not at all what you meant

alex2 wrote:IT SURE WOULD BE NICE IF THE SIX PEOPLE WHO VOTED FOR OFFENSIVE WOULD DEFEND THEIR CASE

If you're going to take that attitude with people who hold a different opinion to yours, why even ask them for their opinion in the first place? It's not like you ever presented the case for survival, and it's not like as grand high supreme generalissimo of A2MT you're obliged to give a shit about anyone else's opinion in the first place.

Doctor Shemp wrote: If you're going to take that attitude with people who hold a different opinion to yours, why even ask them for their opinion in the first place?

Alex has no opinion in this. He could go either way. I presented my case in the previous page

viewtopic.php?f=11&t=14735&start=25#p225505

And he simply wants to hear a counter argument since offensive is winning. That's not too much to ask for if people do want offensive, is it?

Doctor Shemp wrote: It's not like you ever presented the case for survival, and it's not like as grand high supreme generalissimo of A2MT you're obliged to give a shit about anyone else's opinion in the first place.

Except that he does care about opinions in this.

STRANGE, ISN'T IT?

C'mon Shemp. Don't be a jerk.

[7/1/2014 11:01:28 PM] TheGreekBrit: offensive is winning
[7/1/2014 11:01:55 PM] Bwarch: lol why did you even have a poll if you had a side
[7/1/2014 11:02:09 PM] TheGreekBrit: I didn't
[7/1/2014 11:02:17 PM] TheGreekBrit: I'm just annoyed because nobody has actually said anything
[7/1/2014 11:02:23 PM] Bwarch: Ah.

It's hardly reasonable for me to know what alex said on IRC when I don't go on there, Bwarch. And when your request for opinions comes in ALL CAPS, it gives the impression that you're angry with the people you're asking, so what other conclusion am I meant to draw other than you hold the opposite side? From that IRC log, you yourself drew the same conclusion Bwarch.

I prefer an offensive fight. I mean, this is the final boss and it would be kind of lame to just survive its attacks without actually trying to damage it.
I really don't like the few games that have a survival final boss.
Plus it is called a boss FIGHT.

Question: Wasn't the Artist supposed to be the final boss?

Doctor Shemp wrote:From that IRC log, you yourself drew the same conclusion Bwarch.

Yes but I didn't approach Alex with a holier than thou attitude and instead asked him to explain his position calmly and collectively.

Doctor Shemp wrote:it's not like as grand high supreme generalissimo of A2MT you're obliged to give a shit about anyone else's opinion in the first place.

This really isn't productive to be doing things like this.

Doctor Shemp wrote:And when your request for opinions comes in ALL CAPS, it gives the impression that you're angry with the people you're asking,

Just simply an impulsive post, goodness. He's allowed to be frustrated when there's a faceless majority edge going on here.

P.S. If that faceless majority would like to come forward and explain their position we're still waiting to hear from anybody who wants an offense-intensive boss fight.

Edit: Thank you Slit.

Hey, if someone sets up a poll, the working assumption is that every option on the poll is a feasible one. If you think an option won't work, and no-one has presented a case saying it will work, don't make it an option in the first place. And of course a secret ballot will produce a faceless majority. That's the point of a secret ballot.

Alright, lemme give my reasoning for choosing "offensive."

It all boils down to the player's want to feel like they're actually doing something. In situations where it's simply "just, I dunno, dodge shit or something" it doesn't feel like the player is really doing anything to progress the situation, removing a bit of fun from the gameplay experience.

You're not defeating the boss, you're letting the boss defeat itself while you're just kinda along for the ride, unable to progress the situation until the game says "alright, lets go on to the next part." With an "offensive" type boss, your actions have a direct effect on the battle, making it feel like you're actually changing the flow of it, giving you the power to progress the battle at whatever speed is equivalent to your quality of play. It's fulfilling, and really makes it feel like you're beating something, rather than reacting to it until it just throws up it's hands and says "whatever."

You'll have to balance things so that it doesn't become completely overwhelming to focus on attacking and dodging at the same time, but if you do, I believe it will make for a much more enjoyable final battle to cap off the game overall.

alex2 wrote:

Slit08 wrote:Question: Wasn't the Artist supposed to be the final boss?

Nitori is the artist, canon

holy crap mind=blown

Hmm why not just have a mix of offensive and timeout based attacks?

But yeah a final boss may fit better if you are more actively involved.

I prefer an offensive fight because on a practical level, fighting is more fun than dodging. Especially in Super Mario World where Mario's defences are crap and where the physics mostly work better for offensive, fairly short boss battles than the more drawn out ones in your average shoot em up.

And hey, I've seen too many 'dodge everything' bosses in Mario games lately. Just look at Nintendo's own work; New Super Mario Bros Wii, New Super Mario Bros 2, Mario 3D Land, Mario 3D World... all just dodging enemy attacks for the final battle. Same with a lot of fan games and ROM hacks, since apparently making a boss that floods the screen with bullets or chases you or whatever else is easier than making one you actually fight back against.

TDLR: I'm bored of dodging based final boss battles, others are, the physics for Super Mario World work better for battles built around offense than defence and put simply, it's just more fun to actually fight your opponent than merely dodge their attacks.

alex2 wrote:Okay I should clarify one thing:

the boss is still going to be toho even if it's offensive. Mario does not have physics. You'd be shooting bullets just like a toho game, moving left, right, up, and down.

I thought that would've been obvious but whatever.

That's fine, because it still makes the battle feel more interesting than just 'dodge the enemy for a certain amount of time'. Or 'dodge the enemy for a certain amount of time, then lob a green shell at it' like in Scarlet Devil Mario.

If it's survival, will Mario have touhou physics (i.e. none) or platformer physics? I think you should do whichever one doesn't mix touhou with platformer physics since that has an abysmal track record.

alex2 wrote:
the boss is still going to be toho even if it's offensive. Mario does not have physics. You'd be shooting bullets just like a toho game, moving left, right, up, and down.

Well that changes everything or does it

alex2 wrote:For the people who might read this thread for whatever reason and have no context: this is the concept I'm going after for the boss.

I think that people thinking that this would possibly be platforming-based movement and not touhou-based movement didn't watch this video!!


I didn't drop a vote because I don't really have a preference I guess. My assumption would be that survival-only would be easier to code because you need only program the boss' shooting behavior and not need to make any for the player, but of course the 'offensive' option makes for a better 'boss'... Although, you could probably justify the former if the 'story' is something like "you've sabotaged The Artist's Nitori's base/plans and you just need to stall for time while they throw everything they can at you in a last-ditch effort thing," or alternatively stalling for time until everything crashes down on them. Hopefully something more coherent than that though.

So your whole motivation for doing it was "I want raocow to see a touhou character"

To be honest, it seemed like a waste of effort to create something like that for a dead unfinished hack. Better to have its "primary purpose" be for something where it'll get more play. Seems like a better fit, too.

Super Mario World Central? Blech!
Now I'm almost tempted to revamp the Nitori sprite to be ever-so-slightly bigger, so as to clip out of her 64x64 restriction, creating forced cut-off. Haha.

Aw, well. I'll be in my crate, if anybody needs a sprite or two.

so now a2mts gonna be super anticlimatic thats great

fitting but still

alex2 wrote:why you gotta be like that yo

A number of reasons. Namely, an unpleasant experience with a member of the community, who will go unnamed. It's not something that I'm going to talk about here, but it IS the reason I quit rom-hacking.

I don't give a damn about recognition. That sprite was a gift to you, so do with it what you will.

Icelink256 wrote:A number of reasons. Namely, an unpleasant experience with a member of the community, who will go unnamed. It's not something that I'm going to talk about here, but it IS the reason I quit rom-hacking.

wow, one user sure equals a number of reasons for hating an entire community!

(also I have no idea who you were on there, but I probably could have done something about said user if you had reported it.)