/vr/ - Retro Games

Some progress. Replaced Darkman4.

https://www.youtube.com/watch?v=xnICjtIBi_w

I remember the first time I played MM5, I was excited to fight 'Protoman' but disappointed after it turned into "generic robot" and a shitty bossfight.
Wish come true.

I would add Metroid - Rogue Dawn in the recommendations list.

>>3989949
Why are you fighting yourself LOL

>>3990242
It's the story of MM5, but altered.
In the original game, Dr. Wily creates a robot that looks like Protoman to abuct Dr Light and frame Protoman.
In this fight, in the original game, when MM enters the room you see this fake Protoman, but the real Protoman appears and the fake Protoman changes form and turns into "Darkman4".

So since I'm making a hack where you play as Protoman, I had to change the story, and decided to keep "Darkman" as "Fake Protoman".

I want to disassemble Seiken Densetsu 3 but don't know where to start. I know the offsets, I found a SNES disassembler (Dispel) and looked up a SMW3 disassembly, but still don't know where to start.

>>3990509
Can you read 65816 assembly? If not, start reading this:
https://wiki.nesdev.com/w/images/7/76/Programmanual.pdf

The way to hack a video game throughly is to tie what you know to what you don't know, rinse, and repeat. So where do you find that initial traction? You look for code that manipulates hardware registers, because you have a manual dedicated to detailing how they work.
http://www.romhacking.net/documents/226/

To know where the best place to start, you need to understand how a game functions. It's really simple. A game is typically designed to compute a frame, and then wait for the chance to update video memory with the results. Video memory can only be updated by the CPU during blanking (https://en.wikipedia.org/wiki/Blanking_(video) ) because the GPU is accessing video memory to draw the screen otherwise. The vertical blank (V-blank) period is particularly important because it is relatively long, and so lots of data can be updated. In fact, a non-maskable interrupt (NMI) is dedicated to informing the CPU of V-blank. In this way, graphic updates are nearly entirely contained within the NMI routine.

Thus, the place to start is the NMI routine. You need to disassemble it. You can find it's address somewhere in the vector table at $00:FFE0.

Disassembling the NMI gives you lots of useful feedback. You can quickly identify where object data is, trace it to find what routines are used to draw sprites, and use this knowledge to start understanding game entity programming. You can see where updates to backgrounds take place, which may help you, say, tie a HUD drawing routine to the variables it is displaying, like health or score.

In addition, isolating the main game loop is easy. It should only be a few lines of code long. The simplest loop would just load the current game mode, execute a program to compute the frame for that mode, and wait for NMI. Being able to identify game mode programs is not only easy, but gives you access to the game engine.

>>3989831
mediafire requires to be logged in to download shit now? wtf

>>3991629
Everyone wants to track you, now.

>>3989831
i tried to play metroid genocide on my retropie and it's not working, i think it's because the hack is in a rar file... i presume i have to patch it myself. been a while since i played with hacked roms: late 90s

>tfw every promising hack of OoT is either abandoned or just cancelled

https://www.youtube.com/watch?v=1VuoJY7MnxM

>>3991949
There is a reason why 3D game staff rolls have dozens of artists; it takes a tremendous amount of effort to make an immersive 3D world.

>>3990760
Thanks! I really needed this.

>>3992008
You'll quickly learn even the smallest of games are a daunting to fully disassemble. If your goal is to modify specific aspects of the game, being able to locate memory and creatively trace and set breakpoints to find relevant code is an indispensable skill. If there is anything in particular you are hunting for, share it, and we might be able to locate it.

>>3992027
I know that modifying certain elements in the game can be easily done in a HEX Editor and by making breakpoints.

However, I'm not happy with small fixes. There is for instance a 3 player hack for Seiken Densetsu 3 that has some bugs which make the game freeze on certain moments. If the game is fully disassembled, such bugs can be found more easily and the game can be completely modified from source. It will also make it easier to make other games with the same engine.

Seiken Densetsu 3 is one of my favorite games of all time, but it has some small issues. I want to make it the perfect gem that it really should be.

A fully disassemble would also open the path for fan remakes in C or another higher level language which aren't constrained by the lack of space on a SNES cartridge. I even thought about remaking the whole game myself, but that would be too much effort and more prone to legal issues. I thought that a disassembly of the game would serve the communities' interest right now the best.

>>3991949
That looks like ass.
>shitty music
>incomprehensible area design
>a Link's Hat made entirely out of chainmail
>stabbing with a sabre

>>3992092
>shitty music

You know, that's what bothers me most in plenty of "big hacks". The musics are never good.

I say, don't make new musics for the sake of having new musics if it ends up worse. Another good example of that is the hack pictured in the OP.

Am I the only one who can't post on romhacking.net at the moment?

No "quote" or "reply" button, like I'm logged out even when I'm logged in.

>>3993327
It was like that a little while ago, but seems to be fixed now.

I thought maybe we all had our posting privileges revoked as discussion is not relevant to the future of romhacking, only streets of rage character swaps.

>>3989831
Do something not mega man

>>3997264
b-b-b-but Mega Man is great

i'm still stuck trying to make zelda2 hacks 1.5 years later .....lole..........

>>3998268
There is this new editor that came out few days ago for Zelda 2

http://www.romhacking.net/utilities/1300/

I don't know of they are of any use to anyone, but I still think this Genesis/SNES devtool from back in the days is worth sharing.
http://www.smspower.org/forums/15355-DeveloperResourcesConsoleDevTools

>>3998283
that's a good editor, i have tried it indeed. but it is too normie for my tastes as it has no asm .
all my hacks all literally 100% asm lol

>>3998283
i wish people would release the source code for things though.
There's been like 3-6 editors (in various states of release) in the past year , all trying to do the same thing but never really advancing on the editor frontier

What do you guys think of pic?

It was my first rom hack back in 2007, but I never got around to beating it.

>>3998472
It's shit, desu.
From a technical standpoint, it's impressive. But it's a bad game. Remodel isn't much better, since it just carves out huge swaths of the original hack and changes other rooms into simple corridors.

Dunno if you guys will appreciate this but I'm modding FFT's AI. I just taught it how to read reactions. My eventual goal is to make a patch that's nothing but AI changes.

https://www.youtube.com/watch?v=b6ZNr6_OaOk

some extremely good news from the DKC hacking front

>The Rom is now 8mb instead of 4mb, the editor no longer edits the Rom directly but instead you create a project with independent source files that get compiled into the Rom. The only down side of moving data around is not being able to use other editors anymore. On the plus side, you can break almost every original limitation for each level. Here are some of the new limits (some are subject to change so I post the lowest expected limits):
>- Levels can be resized up to 1024x16 (horizontal) or 64x256 (vertical).
>- Up to 8191 entities per level.
>- Up to 21845 entity scripts.

no word on when the editor comes out yet, but it supposedly includes an entity script editor so you can actually control the enemy/barrel/whatever AI

>>3991949
>https://www.youtube.com/watch?v=1VuoJY7MnxM
If only those people released what they already have done so other may finish it

Fancy

>>4002015
What is this? The palettes don't look like they could work on NES, is this a port of the game on SNES?

>>4002034
It's apparently a texture pack that an emulator can load in over top of the game's original graphics.

>>4002034
http://www.romhacking.net/forum/index.php?topic=24125

>>4002015
desu, I will never understand the obsession with (comparatively) HD graphics. old games were shaped by the limitations of the hardware. if those limitations irk you, then you're kinda missing the point.

>>4002015
I still do not get how this got onto the front page, but something like Rogue Dawn only gets a bullet point.

>>3989831
I want a super mario bros rom to kick my ass but not be impossible which should i get?

(Super mario bros frustation is out of the question).

>>4002104
Oh that's simple. No one submitted a news post for Rogue Dawn.
RHDN news posts are written and submitted by the users.

I'm trying to make my Protoman hack for MM5 as accurate as possible with how Protoman plays in MM9/10, but it's pretty hard considering I can't look at MM9/10 code.

Anyone knows of a way I could measure things like length of frames in an animation, or the speed of the charge shot in MM9/10 ?

>>3998472
Beat it and enjoyed it even with the bullshit. I'm surprised no one has even fixed the dungeons and bugs. It would be perfect if that were true.

>>4002535
why can't u look at mm9 code

>>4002065
i disagree. i don't play game for graphics alone

>>4002710
How would you do that? It's a Wiiware game. Besides, I can only read assembly.

>>4002718
can't you download it? isn't it in assembly?

red pill me on wiiware games

>The game development suite Unity can be used to create official Wii games

^this suggests unity is used to create wii games and unity would be a point to look for nintendo related tools

also being in unity has other implications

>>4002535
Get some screen capture recording and count the frames and measure the pixels frame by frame.

>>4002953
You actually would need to measure the distances traveled over several frames to account for sub-pixel accuracy. If you tried to just compare adjacent frames, you'd probably get inconsistent results.

>>4002979
If you can't tell, then you can't tell.
But you are right, you'd need to take a couple of measurements. You'd need to figure out the max speed, how long it takes to reach that speed, and so on, but you should be able to get it close enough to be more accurate than what anyone would be able to notice.

Is there any rom hack that you think should be used in the first playthrough of the game ? Except retranslations, of course.

>>4003557
Automap plus for Legend of Zelda.

Metal Gear Personal Area Network.

>>4003521
What I'm trying to get at is if you restrict velocities to integral pixel/frame values, that is FAR too course. If you only measure the difference between adjacent frames, then an object may move 1 pixel one frame, and then 2 pixels the next, then 1 pixel the next, etc.

Instead, you need to measure how far something goes at constant velocity over many frames. For instance, let's say an object covers 10 pixels in 20 frames, that means it has a velocity of 0.5 pixels/frame.

Fortunately for not-anon, megaman has no acceleration, so it should be easy to determine velocities using the method described. In other words, velocities don't really smoothly vary in megaman, except during jumps.

>>4003557
Sonic 3 Complete, for sure.
Mother 25th Anniversary Edition, MAYBE. I like it personally but the rest of /vr/ spergs out whenever it gets brought up.
MaternalBound Omega is objectively great though.
Pokemon Red++ / Pokemon Crystal Anniversary Edition are way more fun than vanilla Pokemon games but again, your opinion (and /vr/'s) may vary.

>>4003634
Red++ is great but I wouldn't suggest having it as your first playthrough just because it adds things like Pokémon and moves from later games. TPP Anniversary Crystal (I assume that's the hack you're referring to) is a bit better as a first game because it sticks to the Pokémon from Gen 2, but the changed moves, higher level curve (not that it's necessarily hard but still), and the fact that the first couple badges of the game are completely different keep me from recommending it as someone's first experience with Crystal. I'd say both of those games are made for people who have already played the vanilla games and want to revisit them with the improvements by later generations of games. Polished Crystal is another Crystal hack that fits said niche, but I might be biased in saying that since I helped with the music for it.

>>3989831
Are there any decent hacks of Super Mario RPG out there? Pic related was pretty well done but it was also short.

>>4003634
>but the rest of /vr/ spergs out whenever it gets brought up.
What? Why?
What's wrong with it?

>>4004050
>Mother: 25th Anniversary Edition aims to make Mother much more enjoyable for both Earthbound fans and those playing it for the first time.

>All enemy graphics, NPC sprites and tiles are be redrawn from scratch to be more faithful to the clay models. The goofy enemies are goofier and the creepy enemies are creepier. A shiny new font and title screen was added, too!

>The overworld has been tweaked to be much less bleak and repetitive, so it’s harder to get lost. There is more scenery, areas are more colorful and confusing areas like Duncan’s factory have been simplified. On the other hand, tiny areas like the Cemetery or Snowman have been expanded.

>There are also a lot of changes under the hood. Enemies will yield slightly more exp and the encounter rate has been readjusted. The “Dangerous Foe” battle theme is used less often and party members will be met at a higher level. DragonDePlatino’s goal in creating this hack was to make Mother less challenging, but still quite difficult.

>And last but not least, this hack uses Tomato’s official Mother 1 + 2 translation, adapted for the NES. Over 1400 lines of text have been altered to create connections with the rest of the series, improve clarity and shorten reading time. This hack will also feature much quicker battle text courtesy of vince94. Battles progress almost twice as quickly!

If grinding = difficulty for you, I can see why you might not like it. Conversely, I prefer the Pokémon hacks I listed because they're slightly more of a challenge. Different strokes, I guess.

>>4003557
The lag reduction patches for Mega Man 3,4,5 and 6.

The "sliding" fix for MM6.

I'm tempted to say "Zero project" for MMX3 as well but it's good to play the original first to realize the flaws by yourself.

Bump

>>4002065
Its a fresh coat of paint on something you know. Can be fun to see things in a different light once in a while. Its not meant to be a replacement.

who here /6502/ ? yeah i was like programming 6502 and then the pickle matrix got damp from the ketchup

btw whats up with this zelda 2 code? do u see the optimization which could be made?

ldx $10 ; 0x1e8cf $E8BF A6 10 ;; used as monster x register ;draw boss hp bar
lda #$25 ; 0x1e8d1 $E8C1 A9 25 ; A = 25 (timer for kill animation)
sta $0504,x ; 0x1e8d3 $E8C3 9D 04 05 ;; Timer for Enemy
lda #$00 ; 0x1e8d6 $E8C6 A9 00 ; A = 00
sta $043E,x ; 0x1e8d8 $E8C8 9D 3E 04 ;
lda #$02 ; 0x1e8db $E8CB A9 02 ; A = 02
sta $B6,x ; 0x1e8dd $E8CD 95 B6 ; Generated Enemy Slot
lda #$04 ; 0x1e8df $E8CF A9 04 ; A = 04
sta $EF ; 0x1e8e1 $E8D1 85 EF ; Sound Effects Type 4
sec ; 0x1e8e3 $E8D3 38 ;
rts ; 0x1e8e4 $E8D4 60
they wasted a fugging byte how could they waste a byte

>>4007113
>X = $10;
>$0504[X] = 0x25;
>$043E[X] = 0x00;
>$B6[X] = 0x02;
>$EF = 0x04;
>return 1;

I see nothing wrong about this program, nor a way to improve it. Are you complaining about the SEC? Using the carry is a common way to return a boolean value from a subroutine.

>>4003557
what This Anon said,

Zero Project for Megaman X3, IT's much much Better than the original. Zero totally Playable, re-balancances for the entire game, Updated Script

It's much much better (not to mention Zero can grab Upgrades)

>>4007290
no complaints about the SEC! that's needed to signify sword hit something

>>4007971
There is literally no way to improve it based on the given information.

>>4004176
Damn, that sounds great

>>4008445
>>4007113
this is 6502 nes .

it can be changed to be 1 byte shorter but with the same outcome.
there is enough informations...

>>4009406
You're full of shit.

>>4009406
Alright, after thinking about it more, you could put an ASL A in place of the LDA #$04, but that's unmaintainable. If you wanted to assign different values, your cleverness would break the program. That's untenable when you are developing in ASM.

Now fuck off.

>>4003932
Far as I can tell, none. Shame cause it has enough tools too overly reshape the game. Only limitations are the hackers abilities.If one were to somehow have the software to make prerendered graphics back in the day it would greatly help I would imagine.

>>4011358
someone i knew did cinema 4D stuff and managed to nail down the SMRPG style down pretty well
http://peardian.tumblr.com/

>>4011707
>>4011358
It's just normal 3D modeling with a little touch up. Blender could do any of this.

>>4004176
I need a patch that makes Earthbound more enjoyable.

>>3990226
I would recommend DKC 2 The Lost Levels

Aparrently someone saved parts of the Ura zelda:
https://www.youtube.com/watch?v=vlDICKqd52A&t=141s

There is also a hack called Project Beta Triforce that has released some dungeons.Appently is on hiatus but its worth to check it out.:
https://www.youtube.com/user/areoblast43o9

This guy is making a BK hack. He is releasing periodic roms instead of full release in case that if he gives up, at least he already released content for us:
https://www.youtube.com/channel/UC2P5WqJx5ttZTnK7l-caIXQ

Bump

>>4012114
Yeah, but it still takes some knowledge and work to get it right. It's not as intuitive as drawing in pixles.

>>4011707
Looks great, I wish he offered his sprites on a romhack community, it may motivate people to add content for new hacks.

I know C4D, but don't have much experience modeling characters (have done mostly motion graphics stuff). But I'd like to try making stuff for in the style of old games I like, I had Digimon World 1 in mind, Mario RPG would be cool too.

Do Mario RPG editing tools cover stage editing? If someone where to create graphics for new zones and enemies, like the guy on this tumblr, would it be feasible to create a romhack with new zones and a new story using existing tools?

(Would probably be better as a standalone fangame at that point)

I noticed Final Fantasy 3 had a few cut enemies, such as pic related, and I was wondering if anybody was working on or if there is a hack that finished them up and reintroduced them, because I can't find anything.

I love hacks that deal with fixing up and adding cut encounters to fight. New enemies are also alright if done well and are somewhat in the vein of what the devs would make.

>>4014720
>>4014745
Compared to 2D, 3D is a whole different can of worms.

The most challenging part about 3D is the reliance on tools to generate almost everything. Good 3D is as much as a programming task as it is an artistic endeavor. The sheer amount of data that is required to define even modest sized geometry completely swamps manual labor.

This is in stark contrast to 2D, were you fill up a relatively small grid.

The trick with old graphics is to realize that tools we take for granted now didn't exist then. Simple shapes and light models were used because that's what is easy to program. I think the primitive tools are what imparts that certain quality you guys are referring to. Modern Blender et al can reproduce anything from the 90s, no questions asked, but they make it really easy to break the feeling of early 3D because of how subtly powerful they are.

>>4014992
Well I was talking abour Mario RPG's maps, which are pre-rendered graphics, but thanks for the input anyway

Does anyone know how one would go about hacking a game to skip the title screen? I want the game to boot straight to the main menu. I'm pretty inexperienced with ROM hacking, and know 65816 assembly is somewhat hard to hack. Anyone know where I could start, or would anyone be interested on working on this hack with me (possibly for some money)?

>>4015207
which game is it?

>>4015207
look at what happens when you press new game or whatever in a debugger and make it do that as it boots?

>>4015207
Read >>3990760, especially the last bit.

This would probably be a very simple hack. The key is to isolate the main game loop, which is usually very simple. The loop will most likely load the current game mode, then execute the corresponding program, then wait for NMI. Next is the simple task of mapping out the game modes, some of which will likely be the parts you want to skip (initialization, splash, licensing screen, title, etc.). To complete the hack you will edit one of the early game mode programs to set the next game mode to be the main game mode, rather than the splash/license/title/etc. The only possible complication will be the need to run any initialization code the main game engine relies on, which will be slightly more involved.

What game do you want to hack?

Does anyone have any clue how Sonic 3 did some of its backgrounds? The MD is limited to 2 layers as far as I know, and in some places there appear to be more.

Take for instance Hydrocity 2 and Ice Cap 1, both of which have some very interesting effects. The lower half of Hydrocity has horizontally parallaxed columns that move over each other:

https://youtu.be/D6atWDafyEU?list=PL57EDB0404BAFDFE7&t=113

Ice Cap has a background with both horizontal and vertical effects. The vertical effects happen in tall shafts, where the bg center moves slower than the edges of the shaft, giving the illusion of depth. The horizontal parts are similar to the column parallax in Hydrocity.

https://youtu.be/lGsc6SgcpTw?list=PL57EDB0404BAFDFE7&t=48

I was pretty sure the vertical effects were implemented with a per-character background offset mode, but this doesn't explain the horizontal parallax effects. Nor does it explain how the game goes between the two effects without any subtle transitions (like the foreground momentarily hiding the background so the change could occur).

My best guess is that the horizontal effects are done by manually compositing the character graphics in real time and drawing the result. These effects seem to be limited to smallish areas of the background, so this may be possible (but seems very unlikely). Vertical effects are just a per-character offset mode. Everything else is heavy use of HDMA banging on the scroll registers.

Any thoughts?

>>4015610
I should add that the effects are clearly not done with objects (sprites), because they do not suffer from too-many-objects-on-a-scanline flickering.

Any news from that guy who was making a nude hack of some cute anime girl game on the Famicom?

what are the most impressive romhacks you people have seen

i'm torn between minus infinity and bisqwit's castlevania 2

>>4015610
Shit, I'm surprised I never really thought about that.
too lazy to go pore over the s3&k disassembly though

Here's an interesting site I found about Russian romhacks being catalogued that seemed to make it's way through markets and the sorts. Some of these are actually pretty impressive, including re-dubbing the voice acting, and then adding it back in.

http://rgdb.info/

What do you guys use for PS1 games memory viewing and seaching?

Also progress on the MM5 Protoman hack is going well. I think I'm about 90-95% done unless new glitches or ideas arise.

To be accurate with MM9/10 I've added the possibility to shoot a 2nd shoot, a regular buster shot, while a "half charge" shot is on screen, and each have their own hit detection size. This wasn't as simple as is it may sound considering the entire game system is based on only having one type of weapons/shots at the same time on screen, which actually causes a glitch when you use Beat (when you fire a buster shot when Beat is on screen, his hit detection gets lowered by 2 pixels on the Y axis).

Also in the original MM5, the charge counter would freeze during a slide, or during the "rotation" anim on those moving platforms, meaning that the charging of your charge shots gets stopped during those movements. I've fixed that as well.

https://www.youtube.com/watch?v=x4vMp_CO0rY

>>4018674
does th second smol shot still work

>>4018764
Yes, the reason why it doesn't hit in the video is because it has a smaller size, which is what I wanted to show.

When there is a "half charge" on screen, the regular shot actually uses the dmg table of the "half charge" shot, rather than its own, but considering that both are the same...

>>4015241
>>4015503
The game is RPG Tsukuru 2 for Super Famicom. I essentially want to remove the RPG Maker tools from the gamer so I can distribute my game as a standalone ROM. I have a bit of programming experience, but it's all upper level stuff. I'll take a look at what's been posted, hopefully you guys will be able to help out a bit too ;)

How would you do a multiplication in 6502 ASM?

>>4020674
Add in a loop.
Bitshift to multiply or divide by two.
Create a lookup table.

>>4020674
Shift-add multiplication is a pretty good solution.

https://en.wikipedia.org/wiki/Multiplication_algorithm#Shift_and_add

is it possible to implement saving for PCE games like Neutopia using the CD hardware?

>>4021742
doesn't neutopia already have save feature

>>4004691
>The "sliding" fix for MM6.

Wow, looks like 6 is playable again for me.

>>4022229
I recommend this one:

http://www.geocities.jp/borokobo/zip/Rock6FixSlide2.zip

However it doesn't work with MM6, only with RM6. Well it does, but it'll glitch in two particular areas in the fortress which can get you killed.

If couple it with the patch which does lag reduction and speed up of other things (like being able to skip the anim when switching power armors, speed up of scrollling, etc) :

http://www.geocities.jp/borokobo/zip/Rock6Optimization101.zip

and then play it on real hardware... it makes the game 100% more fun. it's amazing how good the game becomes with these 2 patches.

>>4015610
Easy, you directly overwrite VRAM to replace certain patterns in the background with other patterns based on the camera's x coordinate. The patterns are visually identical but shifted horizontally to create the illusion of multiple layers. That's how you create multilayered backgrounds with just one background layer.

I finished Brave New World the other day which was excellent.

>>4022481
>then play it on real hardware
is it still fun on emulator?

>>4022570
This is what I was thinking, but if this is how it was done, the question then is are the graphics saved in ROM somewhere, or are they composited on the fly?

>>4022707
>emulating

>>4022962
Not sure. You could maybe take a look for yourself because most of Sonic 3 (& Knuckles)'s graphics are uncompressed.

I've spent half of the week-end moving graphics definitions around to save and optimize space.

I used to put them in the "main PRG bank" which had tons of free space, but now I need that free space for code

i just shared my romhack with someone and it bugged out on their nes (everdrive) even though it worked on my machine (fceux)

what do i do wrong ?? i cleared all the memory before use this time.

did i mess up the nes header when i tried to expand the file size?

oh god FAILURE WHY

>>4024854
o i shared the wrong rom with them

thank u everyone i bet that's whats wrong

>>4024813
when are you going to start writing the disassembly for rockmans?

it's not that hard...........
-identify nmi
-identify jump tables
-identify byte tables vs code
it all starts to become obvious

>>4024854
I hope you figure this out. It breaks my heart when a hack, or especially a homebrew game only works on emulators.

>>4024902
I'm far from being that good yet

Is BTB autistic?

>>4004050
Graphic overhaul is garbage, gameplay changes are fine, though personally other than Duncan's Factory I think the vanilla game is already well off.

So I want enemies shots to be x2 speed when deflected.

On top of the direction, shots use 4 values for speed:
X direction speed low
X direction speed high
Y direction speed low
Y direction speed high


I used a routine to double these values.

It works great except for certain types of diagonal shots, that is to say, when the enemies shots go in diagonal and are programmed to go where the player stood at the moment the shot is fired.

In some of these cases, the diagonal shots go slightly off course when deflected even if the speed values are perfectly x2.

What can I do to remedy this? Should I just try to write some exceptions, writing the values manually when it's wrong, or is there a better way?

I know that this issue also happens in MM10, but it's much more rare and less noticeable.

>>4027178
On an nes i don't know shit but would guess the vector math is too hard to get right in the cycles you have, so yeah, hardcode it and tell yourself it's an optimization.

>>4027178
why is there a "low" speed and "high" speed?
le what does it mean?

>>4027194
When the low speed goes higher than FF, the high gets +1

That way instead of being limited to FF different speed values, you have FFxFF

>>4027178
Because vectors.
To properly scale a vector you need to normalize it first, then scale it to the required magnitude. You won't notice it when dealing with orthogonal vectors, but moving off the x or y, or a 1 to 1 slope will show errors. If you don't understand that, don't worry because you won't be normalizing vectors in the NES (multiplication, powers, divisions, and a square root)

>>3989949
Newfag here. How do I play this?

>>4027195
a.k.a. fixed-point arithmetic. The MSBs are the integral (pixel) part, and the LSBs are the fractional (subpixel) part.

>>4027178
> the diagonal shots go slightly off course when deflected
I don't understand what your issue is. A deflected shot isn't aimed, so it can't go off course, correct?

Do you just want to correct the aimed shots that are slightly off? >>4027181 is likely right, in that the code responsible for computing the velocity of the target is an approximation. The correct function would look like this

D = (target location - source location) // vector from source to target
N = D / sqrt(D_x * D_x + D_y * D_y) // direction from source to target
V = speed * N // Projectile velocity

but the multiplications, sqrt, and division are expensive on the 6502 because they must be implemented in software (or auxiliary hardware). Lookup tables are probably the way to go (and probably what the devs used) because they are fast, however their quality depends on their size. It's an example of space-time complexity trade off; use up a lot of space to save time, or use up a lot of time to save space.

For example, I programmed a 2-player asteroids game for the SNES ages ago. To compute projectile velocities, I used the ship velocity as the input to a function that computed the angle of a vector. To be fast, this function worked by repeatedly dividing the velocity components by 2 (implemented by shifting) until both of the components' highest set bits fit within 4 bits. I then used these values to index into a 2d array of angles.

>>4027605
>To properly scale a vector you need to normalize it first
To properly use vectors you need to read the first 10 pages of a elementary linear algebra book before uttering such trash.

R = x X + y Y
s R = s (x X + y Y) = s x X + s y Y

>>4028276
You are correct,
>>4027605
Is full of shit, please ignore it.

Remember children, just because you never forget how to ride a bicycle doesn't mean you won't fall off and tumble down a hill.

>>4028276
>A deflected shot isn't aimed, so it can't go off course, correct?

When you play as Protoman, you can use his shield to deflect (some) enemies shot and they're supposed to go back exactly the way they came.
It's not like when your shots get deflected on a shielded enemy, in this case you always get the same speed/direction values.
I got that part right long ago, but the thing is, shots deflected that way also gain 2x speed. Which is when this becomes an issue with some diagonal shots. I was assuming the issue came from the fact that the NES resolution is non square.

I've thought of using a lookup table indeed. I do have several hundred bytes of free space in what I call the "main PRG bank", the one that's on at all times, after I spent 2 days cleaning it. Maybe I should try that indeed.

>>4027809
It's not released yet. At this point it's about 90-95% done but I can't give an estimate of release.

Super Mario World, How can I set all carryable sprites to universally appear on top of Mario's sprite?
Right now carrying a shell or a P-switch will put the sprite in question behind Mario

>>4028625
>1) Isolate object memory in RAM (the 512 + 16 bytes DMAed to OAM every frame)
>2) Isolate the program(s) that write to this memory. These are the "sprite" (collection of objects) drawing routines
>3) Isolate the call that draws Mario, and the call that draws the carried item
>4) Reorder the calls

Step 1 is trivial if you trace DMA activity. Step 2 requires you to set breakpoints to watch for when the relevant RAM is written to, then trace the subroutine that stops the debugger, and finally figure out the subroutine's entry point. Step 3 is accomplished by searching for calls to the now isolated sprite drawing program, and which sprite is drawn is figured out by NOPing the calls out to see what didn't get drawn. Step 4 is then to swap the Mario and item drawing calls.

I make it sound simple, but there may be a few technicalities. But this is more or less how to go about doing it.

Why this works is because object layer is determined by the order objects appear in OAM. In the simplest case, the 0-th object is the highest priority, and subsequent objects have lower priority. Higher priority objects are drawn on top of lower priority objects when there is overlap. You can also set an offset for the highest priority object, which allows "rotating" the objects in the case of too many on the same scan line, which results in flickering, but (mostly) visible objects. You could perhaps just change the highest priority object, however this would probably serious fuck with rendering. So by changing the order of the calls, you change how the objects appear in OAM, so you get yo item on top of da wop.

bump

I was recently informed of a Super Mario World hack in which two players play and the game switches who is in control of Mario every 0.5, 3 or 5 seconds. He claimed it was called "One Mind", but I can't find any such hack. Does anyone know what he might be referring to and where I could get it?

>>4030576
google 3rd result

>>4030660
No such luck. If it was that easy I wouldn't be here.

>>4030661
http://www.dotsarecool.com/twitch/hacks.html

Are you blind? I googled "One Mind Mario World" and got this as the third result just as Anon said.

Bump

Almost done.

>>4031201
Hold on a minute, I thought you quit

>>4031270
Who the fuck told you that?

>>4031318
Someone in these very threads apparently posing as you saying the project was cancelled.
Fooled a lot of people.

>>4031332
What? Is it archived?

Well at any rate, its still going on. I was kinda perma banned because mods are shit and I had to post in the /vp/ romhack discord instead.

But that's just plain ol wrong as you can see. The front sprites are almost done. I just need a little help with the back sprites and I was kinda wondering if some of you would be willing to help me with that.

And no. Certainly not dead.

>>4031404
It probably is, but I don't remember which edition it was

>>4031413
http://desuarchive.org/vr/thread/3884525/#3902998
It was just last month, and last thread, moron.
The post and all of the replies to it were also deleted because a shitstorm broke out.

>>4031201
Is it too late to make a request?

SNES ASM, help. Am I understanding this correctly?
I don't get some branch stuff. For example, take this:

$00/8D5F: B0 11 - BCS $11 [$8D72] - A:0000 X:F180 Y:000E P:envmxdizc

B0 11 translates to BCS $11, or "Branch to $11 if carry flag is set"
Where does the [$8D72] come from? Obviously it means "If carry flag is set, jump to $00/8D72" but what does the $11 have to do with this? I'd get it if the hex was B0 72 8D or something, but I don't see how that $11 implies the particular address of 8D72.
I went to see the RAM address of $11 too (the direct page) but it's just 00s

>>4003557
Metroid + Saving. Saves instead of passwords, gives a small minimap when paused I think, and you no longer lose wave or ice beam when picking up the other.

Castlevania 2 Redacted. The text in the game is written faster, and the transitions from day and night are quicker. The dialogue actually having value is nice too. Not just a re-translation.

There's a bunch of bugfix patches for FF6. Roughly around 80 or so. There's also a few major bugfix patches that compile many of the smaller ones together, but most of them also make gameplay changes or otherwise modify some shit that they shouldn't.

I might get some flak for this, but Project Base is a good romhack for people who haven't played Super Metroid before. Make sure you get the one that just does minor gameplay tweaks, and not the one that redoes the level layouts. Movement in Super Metroid just feels clunky after trying Project Base.

Here's the deal. I'm doing three things with Vidyamon.
-A simple graphics hack of Crystal version without the memes and the 4chan tans. Just the video game characters. All trade evolutions are level ups now and every mon is available in game. Simple stuff. I've realized that some anons don't like meme stuff so I'm making a version without it.
-a randomized version of the above. So you don't have to wait to get a certain vidyamon. I know there's tools online to do this but I figured it'd be easier on a lot of folks to just have one ready for shits and giggles.
-My original plan that'll come way later that includes its own storyline, the 4chan tans, maybe even some original songs, all that good stuff.

>>4031487
Not at all.

>>4031468
I know I've been taking long as shit breaks but this isn't my full time job. I feel bad about leaving you guys like that because shit like that happens. But the thing is, sometimes I'm burned out, not in the mood or just plain old busy with more important matters. The important thing to know is that I'm not giving up on this project. There's too many contributions from you lot to just flat out quit like that.

>>4031520
Looking at some 65816 references, it seems like the addressing mode for BCS is program counter relative.

So basically it's saying skip ahead 0x11 bytes.

The PC is a register that tells the processor what command to do next, so the snes will load whatever command it finds at the PC +1 address then increment PC by the size of that command. Then it executes the command, in which a branch that does get activated will add the specified value to the PC, changing which line of code gets executed next.

So PC = 8D5E.
Plus 2 byte size: 8D60.
Plus 11 to branch: 8D71.
Thus 8D72 is the address of the next command.

>>4031520
65xx branches save space by using 1-byte relative addressing (except the 65816 BRL which uses two). Branches CANNOT cross bank boundaries; the target of a branch is always in the same bank as the branching instruction.

If the branch test fails, a branch is not taken, and the PC is incremented to the next in-bank instruction (in your case the bank is $00 and the instruction after the branch at $00:8D5F is located at $00:8D61). If the test passes, then the PC is set to the sum of the next in-bank instruction ($00:8D61) and the sign-extended branch operand ($11 => $000011) with the bank ($00) unchanged (in your case this results in $00:8D61 + $000011 = $00:8D72). More precisely, the formula is:

(current bank << 16) + (address of next instruction + branch operand) & 0x00FFFF

Read Eye's book, specifically the chapter on addressing modes.

https://wiki.nesdev.com/w/images/7/76/Programmanual.pdf

>>4031528
I completely understand, VidyAnon.
Take your time. It doesn't have to be done tomorrow or anything.

>>4031201
About the only thing that looks weird to me is the backsprite of the Metroid. It's kinda low, and small.

There is one thing I don't remember.

I'm doing an ADC operation, and I want to branch out if the result DOESN'T go over FF

should I do BCS, BCC or BVS, BVC ?

The carry flag is cleared if it goes over FF right?

>>4031520
Now that you understand this, try to branch with a value between 80 and FF to see what happens

NVM that, I think I got it right. I got confused for a while because I had forgotten that if you want +01 you have to do "ADC 00" and not "ADC 01".

As for this
>>4027178

I tried with a x2 multiplication routine and it worked well, the problem being that x2 speed for the deflected shots didn't feel right at all.
Way too much. x1.5 is more like it and feels MUCH closer to how it is in MM9/10

Sadly the multiplication routine I found could only do whole numbers, so I ended up ditching it. Instead I went for a double solution: for shots with fixed values: simple LDA/STA, and for shots which speed value vary because they're homed at the player's direction, I used a table coupled with a bunch of checks and ADC to make sure the "high" speed values are correct.
This eats up a lot of space, but thankfully I'm nearing project completion and I still have free space.

>>4032824
ADC 00 always confused me.
are you considering the value of the carry?

>>4032824
ASL <=> Multiplication by 2
LSR <=> Division by 2

a * 1.5 = a + a>>1

It's really fucking easy

>>4032772
heres my notes on the overflow flag but i've never played with it

V = OVERFLOW. Set if the addition of two like-signed numbers or the subtraction of two unlike-signed numbers produces a result
greater than +127 or less than -128.
Bit 6 - V - Overflow flag: when an arithmetic operation produces a result too large to be represented in a byte, V is set.
The Overflow flag is generally misunderstood and therefore under-utilised. After an ADC or SBC instruction,
the overflow flag will be set if the twos complement result is less than -80 (-128dec) or greater than 7F (+127dec), and it will cleared otherwise.
In twos complement:
$80 : $FF = -128 : -1
$00 : $7F = 0 : +127
Thus, after:
CLC
LDA #$7F ; +127
ADC #$01 ; + +1
the overflow flag is 1 (+127 + +1 = +128), and after:
CLC
LDA #$81 ; -127
ADC #$FF ; + -1
the overflow flag is 0 (-127 + -1 = -128). The overflow flag is not affected by increments, decrements, shifts and logical operations
i.e. only ADC, BIT, CLV, PLP, RTI and SBC affect it. There is no op code to set the overflow but a BIT test on an RTS instruction will do the trick.
The overflow flag is set during arithmetic operations if the result has yielded an invalid 2's complement result (e.g. adding to positive numbers and ending up with a negative result: 64 + 64 => -128). It is determined by looking at the carry between bits 6 and 7 and between bit 7 and the carry flag.

>>4033130
Are you saying there is a much easier way to do x1,5 ? I'll look into this

>>4033132
Thanks

anyway here is how it looks in action with what I did

https://www.youtube.com/watch?v=cRgu3Btk35w

Hey /vr/, it's /co/. I made this dumb rec list, and would like help expanding it and/or making it less dumb.

Thanks in advance.

>>4033852
I'd recommend deleting that. As for /co/ related, there's one good mod I can think of, and it's not /vr/. Punisher mod for Max Payne 2. Massive improvement.

>>4033852
add Batman and Flash :^^)

Updated

>>4033132
>It is determined by looking at the carry between bits 6 and 7 and between bit 7 and the carry flag
More precisely, when adding/subtracting A and B overflow flag V is set by XORing the carry C and penultimate carry C' of the computation.

V = C XOR C'

case 1) if A and B are positive, then there is no carry C since the sign bits are both 0. If there is no penultimate carry C', the result did not underflow, and thus V = 0 XOR 0 = 0. If there is a penultimate carry C' that means overflow (because the result is negative), and thus V = 0 XOR 1 = 1.

case 2) if A and B are different signs, the result will never overflow, so V = 0. Now, there is a carry C only if there is a penultimate carry C', so V = 1 XOR 1 = 0. On the other hand, if there is no carry C, then there was no penultimate carry C'. so V = 0 XOR 0 = 0.

case 3) if A and B are negative, there will always be a carry C since the sign bits are both 1. If there is a penultimate carry C', that means the result is correctly negative with no overflow occurring, and thus V = 1 XOR 1 = 0. If there is not a penultimate carry C', the result is positive, and overflow occurred, and thus V = 1 XOR 0 = 1.

Take time to reason about this, write out a few examples, and you will understand.

For 65816 this covers the ADC and SBC (subtraction, which is really addition of a negative numbers) instructions. The other things that set V are straightforward to understand. BIT oddly enough copies the penultimate bit (not carry) of its operand into V. You can also clear V with CLV but there is no set instruction, yet inexplicably there is a pin on the processor that does just that. Lastly you can re/set V manually with REP and SEP or use the stack to pull a processor status register PLP. I think that's it.

>>4033130
>ASL <=> Multiplication by 2
>LSR <=> Division by 2
>a * 1.5 = a + a>>1

So wait, is this correct?

LDA (SPEED LOW) X
LSR
SEC
ADC (SPEED LOW) X
STA (SPEED LOW) X

With "D9" I get a result of "46" after the operation so this seems correct

So now all I'd need to do to take everything into account is :
>add +1 to the high speed value if the operation overflows
>check the high value, before the +1 I just mentionned, multiply it by 1,5 as well, and if it's odd, add +80 to the low value
>if after the +80 to the low value, it overflows, add +1 again to the high value

If this works, this will make things more simple, faster, and make me save about a hundred bytes because I won't need the table.

BTW is there any way to check if a number is odd or even? I mean, I don't exactly need it in this case because the high values concerned are never bigger than "4" so I can easily do CMP1, CMP3, but I'm just wondering.

>>4034776
>So wait, is this correct?
Not quite. There are actually 4 shift operations. ASL and LSR shift a zero in, while ROR and ROL shift the carry in. Both shift a bit out to the carry. You are dealing with a signed 16-bit number on an 8-bit machine. None of the shift operations automatically sign extend on the 65xx, so you have to do that manually. Also you will need to chain two bytes together. This is one way to do it for a variable called vx (for velocity x), although it's a bit optimized

LDA vx.hi ; Get sign into carry
ROL A ;
LDA vx.hi ; Roll hi byte right (sign bit shifts in, low bit of hi byte shifts out)
ROR A ;
PHA A ; Stash the shifted hi byte result for a moment
LDA vx.lo ; Roll lo byte right (low bit from high byte shifts in, truncated bit shifts out)
ROR A ;
CLC ; Add the shifted lo byte result to the lo byte and store
ADC vx.lo ;
STA vx.lo ;
PLA ; Recover the shifted high byte result
ADC vx.hi ; Add the shifted hi byte result to the hi byte and store
STA vx.hi ;

>BTW is there any way to check if a number is odd or even?
Think about the definition of even and odd numbers, and how binary is encoded for a few minutes.
even <=> x = 2 * n
odd <=> x = 2 * n + 1
the low bit in an unsigned binary number is adds either 0 or 1
if the low bit is 0 then even, if the low bit is 1 then odd
LDA number; AND #$01; BEQ even;

>>4034776
To add, you program would almost work if you were dealing with an 8-bit unsigned number. However you used SEC instead of CLC, so your result will be 1 off. (you CLC when you start additions, and SEC when you start subtractions)

>>4034818
Last thing to say, compare this with the 16-bit unsigned multiplication by 2

LDA vx.lo
ASL A
STA vx.lo
LDA vx.hi
ROL A
STA vx.hi

>>4034818
So if I'm getting this right, the reason you use ROR instead of LSR is because in the cases when the carry is 1, LSR would get that wrong?

Anyway thanks I'll give this a try

>>4034824
>you CLC when you start additions, and SEC when you start subtractions

Damn, this could explain why my shield hit detection doesn't work in some rare cases.

Okay, I think I got it to work using shift operations.

Thanks guys. Compared to what I had done, not only this is faster, more effecient, but since it doesn't require a table I've saved 100+bytes. Awesome.

>>4036920
nice,

the daxter one was great too

>>4002236
Panic in the Mushroom Kingdom 1 and 2
Return to Dinosaur Land

Bump

Hopefully you can tell who they are.

>>4021921
Neutopia used the password system. Passwords for 16 bit rpgs can fuck right off.

>>4037369
>Mike Jones
NICE

>>4037369
>no bananas in his ears
Shit tier.

>>4037535
Sorry to disappoint you anon. I never played StarTropics so I wouldn't know. Plus someone else made the sprite. If you want you can add bananas to the sprite if you so desire.

>>4037380
i just read the neutopia manual which suggests saving is possible with TurboGrafx-CD Or turbobooster plus???

>>4037597
Yeah, if you had those peripherals. From what I remember they weren't exactly cheap. Pretty sure either of those cost as much or almost as much as the TG-16 system. Also I honestly can't ever remember seeing a single turbografx-cd where I am. Not even sure if they were sold in the US, or at least in Georgia.

then it seems possible for >>4021742

We made some adjustments for the hack. You don't have to use them but I think they look better right?

do any of the everdrives incorporate fpgas to emulate coprocessors? i'm teaching myself how to use them, and the possibilities are overflowing now that I'm getting things working.

>>4037914
vidya anon, you should consider using these. I think you have a fantastic base, but you will probably want to do a round of touch up once once everything is there. This guy clearly put some love into Croft's tatas, and that's how it should be.

>>4033852
>Castle Smurfenstein
Fucking hell, that takes me back... Anyway, you should probably cut down on the Sonic stuff, because holy shit that is literally 80% Sonic there.

>>4037914
>General Custer is in the hack

>>4034316
it's almost fininshed!

I've been folowing your posts for a while. exciting

>>4039169
The front sprites, maybe, but the back sprites have a long way to go.

>>3998472
Played the original version and it was good. There were some extra heart containers though, so you had like 3/4s of an extra heart at the end of the game.


Never tried the MQ, which supposedly has a secret boss/dungeon, and I also didn't play the updated version. The new version supposedly fixed a lot of issues, bugs and design.

FUCK THE TELEPORT DUNGEON. FUCK I HATE THAT LEVEL.

>>4000135
Nice. Does this work with all 3 games?

>>4012151
The part where the gril goes missing is the worst section of the game by far. I wish you got Poo shorter after there so it would mix it up a bit and it wouldn't drag on.

>>4007815
I liked this. The boss AI is different in most places too. The sword has been nerfed, but it is still OP.

There were suppose to be more minor updates, but I think he stopped for now. He wanted to do a X4 style sword too, but that is a lot more involved.

feeling kinda lost still
Trying to figure out the Pocky & Rocky compression but I'm not that good with 65C816
In the attached image, the top bytes are the compressed data and the bytes below it are what's decompressed to RAM. They're almost the same, but there are apparent 'control words' here and there in the ROM which translate to some kind of copy commands leading me to believe it's a LZ variant. Wherever there's repetition in the RAM row, they're always bytes that have appeared beforehand.

Here's what I got so far
00 00 means 'print the next 16 bytes as they appear and read a new control character'
*0 00 is a LSR counter, take the 40 00 for example. 40 is copied to 00:001B, the
next byte (4B) is printed and a BCS is read. Branch doesn't happen, and the 40 is LSR'd. It's now 20. Print the next byte (7B) and shift the 20, et cetera, until there's been so much shifting that the byte becomes 00 and the Carry flag is finally set. Then you read the next control word.
The LSR counters are evident in the image because after the 40 00 there are 6 pure bytes, with 20 00 there's 5 and with 80 00 there's 7.

My method so far is setting a read breakpoint to 1AFCE0 and starting the game. The data I'm watching is the palette data for the intro animation but I'm confident the same compression is used for graphics too (1AF178 is intro sprite graphics)

how do i git gud? can someone help?

>>4039357
the image pff

>>4039357
Reverse engineering the Super Metroid compression routine was the first hacking I ever tried years and years ago.

http://www.romhacking.net/documents/243/

Honestly it was a mistake. Decompression routines are highly optimized and varied. They are difficult to fully understand, especially for somebody just starting out. From personal experience, trying to understand the routine from the input and output is far harder than just disassembling the routine, which itself is already a hard task.

That being said, reverse engineering decompression is an inevitable hurdle for any good hack, and that's not mentioning writing a compression routine to insert your own data. Of course you could bypass the routine by copying the expanded RAM back into ROM and then call a memcopy function in place of a decompression function, but you quickly exhaust ROM that way.

I might be willing to help, but you need to isolate and trace the assembly routine.

>>4039383
I've done that and I'm stepping through the commands with 'Step Into' but Geiger's SNES9X only lets me watch one place simultaneously so it's a bit cumbersome. There are important things written to 00:0010 through 00:001F during the decompression, at least.
I just wrote the input and output down so it would help me visualise it even a bit.
Thanks for the link though, I'll check it out when I wake up.

>>4039270
nope but if the dude releases his source it might be possible to extend its abilities to edit the three games
it's a very unorthodox editor; you load the ROM and it dissects it to different project files the editor can manipulate. You can't re-open an edited ROM again, only the project files.
I think it's a novel idea. Sometimes it's better to dissect the ROM to pieces you want to edit and only when the desired changes are made, smart reinsertion happens and pointers are recalculated. For smaller/simpler games it might be better to go the traditional way, edit the bytes where they are and hope you don't break anything.

Maybe this question belongs in /g/, but because I am wanting to learn this to take part in the /rhg/ community, and I was told by several folks here to learn C, then assembly, I will ask anyway.

So I am using The C Programming Language by Kernighan and Richie, and everything was going great until I got to the character input and output section. I have probably read it 5x now and still don't have a firm grasp of how it works. Is it ok to ask some more questions about it here or is it too general and would better belong on another board?

>>4041129
what, you mean like fprintf and fscanf and shit? C strings are just arrays of chars with a zero at the end.

>>4041241
>C strings are just arrays of chars
I suppose that makes things a little less murky; my only other experience with programming is Java, and iirc it treats strings as objects. If you want to manipulate a string you literally have to create a new string object; you can't actually edit it.

I guess my problem begins with their discussion of getchar() and putchar()

>The standard library provides several functions for reading or writing one character at a time, of which getchar
and putchar are the simplest. Each time it is called, getchar reads the next input character from a text stream
and returns that as its value. That is, after
c = getchar();
the variable c contains the next character of input. The characters normally come from the keyboard; input from
files is discussed in Chapter 7.

I think I understand putchar pretty well; it takes a char as its parameter and prints it on the screen when called. getchar is where I get confused. The statement

char c = getchar();

makes sense; we make a variable named c of the character type, and set it to equal what getchar() returns. Assuming the keyboard is the default for input for this function, when the statement is executed and getchar is called , the user is prompted to enter a string, which is then stored in memory as an array of chars. Is this correct so far?

A lot of my confusion came when I started playing around with the two functions. For example I would execute the getchar function (without assigning it to anything) then call it again in the very next statement but assign it to something, and it behaved not as I expected it to. I don't recall exactly what it did, or the other things I tried just playing around with the language, so I will stop here until I am and in front if my computer. Thanks btw.

>>4041298
Oops, this all is from the text I am using and was meant to be greentext

>The standard library provides several functions for reading or writing one character at a time, of which getchar and putchar are the simplest. Each time it is called, getchar reads the next input character from a text stream and returns that as its value. That is, after
c = getchar();
the variable c contains the next character of input. The characters normally come from the keyboard; input from
files is discussed in Chapter 7.

>>4041241
well calling it a zero is a misnomer

it's more like \0

>>3991907
IPS patcher

>>4041461
>it's more like \0
Which is octal for 0 inside a char/string literal. Calling it a null terminator is a little mysterious sounding, so I just called it what it is.

>>4041298
>my only other experience with programming is Java, and iirc it treats strings as objects.
Unlike Java's derived String class, C strings are primitive types. The null terminator is really the only convention, and that's because a string literal with double quotes " appends a null at the end of a string, making it 1 character longer. Single quotes ' omit the null. While sthing.h provides functions that assume null-terminated strings, you can easily write functions that do not exhibit this behavior.

*nix philosophy treats everything as a file, which is just a series of bytes (there are exceptions, but this thinking gets you really far). stdio.h functions are wrapper functions for low-level OS file manipulation commands like read() and write(). The purpose of stdio.h is two fold; portability and buffering. Stdio.h hides OS specifics to make your code portable, while buffering is incredibly important for performance because the system commands are expensive to call (your program often has to be suspended). By reading in large chunks of files at a time into a buffer internal to your program, you only infrequently have to make system calls, and instead mostly rely on local calls.

So how does this relate to you? getchar() and putchar() work on the file streams stdin and stdout, respectively. What exactly are stdin and stdout? Well, every process on your computer actually gets set up with these standard file streams when the OS creates them. For instance, if on the command line you run a function without any input, the keyboard becomes stdin. In contrast you could pipe a file in, in which case that file becomes stdin. Regardless, both streams look the same to the program, a series of bytes. The bytes get read into a buffer, and are consumed when you call getchar().

>tfw you want to play FF6 T-Edition but don't speak Egypt

So what's the next best FF6 romhack?

>>4041504
Wow, thank you for the detailed response. Unfortunately, I don't know all that much about operating systems (I've only ever used Windows, never Linux or anything else) and have never really used command line / prompt. I am familiar with what these things are, but my knowledge is only surface level at best. I don't really know how a computer actually works under the hood.

But I am determined to learn and after reading your post a few times and looking up some of terms that were foreign to me, I think I understand (for the most part) what you are saying.

In a few minutes I will open the book again and see if I can make more progress. Thanks again!

>>4041919
Just know this, calls within your program are cheap, while system calls are expensive.

The I/O buffers I am talking about are circular buffers that have two indexes associated with them; one index points to the next byte to write to, while the other points to the next byte to read from. If you are, say reading a file, you first get a file descriptor using something like fopen(), and then you start reading the stream with stuff like fget(). In this case, the initial fget() just wants a single character, but accessing the hard drive for a single character each fget() is prohibitively slow. Instead when you first call fget() the stdio.h library actually asks for a big chunk of data from the file, writes it to the buffer and advances the write index, and then returns the first byte pointed to by the read index. Each time you call fget(), the read index advances until it catches up to the write index. If they ever equal each other, more data is loaded into the buffer from the file, or no more data exists and an End-of-File is returned.

To add, something like unget() is usually implemented as a 1-character buffer, so you can only unget once reliably. If you want to unget more than 1 character, then you have to write your own code.

Of course old systems don't really have fprint and fscan, let alone system calls.

>>4041970
So now you can understand what and how stdin and stdout are implemented. When a *nix OS executes a program, it automatically sets up that program with essentially 3 open files: stdin, stdout, and stderr. What these files are associated with depends on how a function was called.

If you call a program from a CLI with no arguments, usually the keyboard is tied to stdin. When you type, the data is stared in an I/O buffer and the write index advances. When your program uses getchar(), it reads from the I/O buffer and advances the read index. If you call getchar() while the keyboard is associated with stdin and no data is available (the read and write index equal each other), then the programs waits for input.

Bump

Trigger warning

https://m.youtube.com/watch?feature=youtu.be&v=lBHnoBOEEqU

>>4043792
lost it at the end

>>4043792
lel very nice

>>4043792
hooly shit that's loud

>>4043792

>>4039361
>>4039357
I disassembled the Pocky and Rocky decompression routine that you were looking at. I'll upload the code later when I clean it up a little.

Apparently the data is de/compressed with an LZ-like strategy. The code is pretty sloppy, but quite simple.

You call $00:8CB8 with arguments passed in by Y; Y.hi holds the bank with the compressed data, and Y.lo holds an index into a pointer table located at the start of the bank.

Decompressed data is always sent to WRAM address $7F:8000.

The data stream has the following format. The first byte flags if the stream is uncompressed (false) or compressed (true). The next word (2 bytes) is the size of the decompressed data. After that depends on the mode. If the mode is uncompressed, then the routine directly copies ROM into RAM. If the mode is compressed, the stream of data is partitioned into 'lines'. Each line starts with a header word. The header contains bit-packed flags on the next 16 entries, specifically if an entry is an uncompressed byte (false), or a decompression command word (true). If it's uncompressed, the data is simply copied from ROM to RAM. If it's a decompression command word, the lower 11-bits are an offset subtracted from the current output address to grab previously decompressed data, while the top 5-bits are the amount of data to copy (2 bytes are always copied, so the range is [2,34] bytes). For every byte written to output, the number of bytes in the stream is decremented, and the algorithm exits when that number hits zero.

That's it. It's way more simple than SM's routine, probably less efficient, and definitely written like crap. However knowing that there are pointer tables at the start of banks with compressed data makes finding said data trivial.

>>4043792
*glurkle*

>>4045603
Ex: the first time $00:8CB8 gets called after reset, Y = 0x1414 (eh, not a great example), which means go to bank 0x14 and decompress the 0x14-th address found in the table at $14:8000. Looking up the address, data is located at $B1A0, or $14:B1A0 when you take the data bank into account. Going to $14:B1A0, the first byte is 0x01, meaning the data is compressed, and the following word is 0x0800, meaning the data is 0x0800 bytes long when decompressed.

If you are using Gieger's Snes9x debugger, you can set a breakpoint to watch when $00:8CB8 gets executed, and then use 'Skip OP' to prevent the call from happening. In this way, you can selectively disable decompression, and by seeing what happened, have a good idea what the data that just got decompressed is.

In the case of Y=0x1414, it seems the data is the license screen background table (not the character nor the pallet), because when you disable this particular call the license screen is garbled. The fact the data is 0x0800 bytes supports this, because this is precisely the size of a background table (0x20*0x20 tiles * 0x02 bytes/tile = 0x0800).

In the case of the picture you posted here >>4039361, when $00:8CB8 is called Y = 0x1A??. Looking at bank 0x1A, entry 0x0D contains address 0xFCE0, so Y must have been 0x1A0D. Setting a breakpoint for $00:8CB8, and looking for Y = 0x1A0D, I deduce that the data you are examining is the palette for the title screen.

Aside: Address $00:0004 hold the current game program mode. Mode 0x20 is the main game. Try doing shit like going to the title screen and then setting this byte to 0x20. You'll find yourself playing the title screen like the game.

>>4045635
Proof of title screen palette fuckery.

Black Pocky and Missgno-esque Rocky

>>4043792

>>4045654
>>4045638
>>4045635
>>4045603
Holy fucking shit dude
Thanks so, so much!! You saved my wedding (I'll elaborate on that later)
Is there a way I can repay you? You got PayPal or a BTC address or something?

>>4045710
dude, i just enjoy hacking old video games. glad it was helpful

>>4045739
Well huge thanks anyway man
/vr/ truly the best board

>>4043792

Laughed at "& Knuckles".

>>4043792
wait a minute this is all 4chan themed

I don't want 4chan in my poke

>>4046759
Where have you been? The protagonist is /vr/-tan, the rival is /vp/-tan, Professor Oak is Moot, etc.
It's always been 4chan themed.

>>4045603
>I'll upload the code later when I clean it up a little
did you mean what you said in >>4045635 or you got some pseudocode/c ?

>>4046809
it... grew.

https://github.com/gewballs/pocky

you are interested in lines [512,750], the Decode routine.

When tracing the decompression routine, I traced the NMI incidentally. I was studying the license screen as a working example, and was so close to Reset, I just ended up tracing that too. Only about half is disassembled/decompiled, but there are interesting things in there. The psuedo-C decompile might look a little weird, but that's because I lined it up with the assembly as best as I could.

Grab an SNES dev manual reference and the NMI will read like an open book where various graphics registers data are stored in WRAM.

>>4046959
thanks man
probably decompressing everything later today and see if there's some unused cool shit somewhere

Just got redirected here from /v/
Is there any way to rip the art assets from an ISO of Blood Factory/Loaded? (PS1, Dreamcast)
Specifically after the splash arts, I've never done anything like this before so I wouldn't know if it's possible. Apologies in advance if this isn't following this threads guidelines.

>>4047115
pop the disc in, mount it, check the file structure
if there's something obvious like a folder called "GRAPHICS" or shit, check out the files
easy way out is to try and use TIMView or PSicture to see but you might have luck with Tile Molester too

can you mount the ISO so we can check out the files?

>>4047115
Try PSX Media Converter

>>4047115
If you have to ask, you wouldn't know.

Go back to /v/.

>>4047129
fug off

>>4046759
The creator literally said there'll be a memeless version a couple of posts above. You would know if you read the thread.

Bump

>>4046759
YOU'RE FAR TOO LATE FOR THAT.
EVEN THOUGH IT'S NOT RETRO

>>4003932
http://ngplus.net/index.php?/files/file/11-super-mario-rpg-armageddon/

>>4041906
http://ngplus.net/index.php?/files/file/12-final-fantasy-vi-brave-new-world/

>>3991949
There's this really awesome hack of OoT to recreate Link's Awakening in 3d. I fucking love LA so I'm excited for it, I just hope it gets finished. It looks like it's got great progress, and from what I can tell, the author is being smart about how he reveals progress i.e. doing some interviews, some screenshots and scant footage, and to avoid Nintendo's DMCAs he's hidden the project and ceased all new information until it's finished, which is smart. Anyway it looks like if there's going to be a proper OoT hack, Link's Awakening 64 is it. But there's no way to actually check on progress since half a year ago, so... yeah. It was in development for a year before hiding, so I wouldn't take this as a bad omen.

>>4050326
I've heard of this project.
The author is adding a bunch of dumb edgy fanfiction shit to the point that it may as well not even be Link's Awakening anymore.

>>4050446
Source? If so, that's highly dissapointing

>>4050641
Just look up videos. It's all there.

>>4046959
disassembled most of the nmi and pushed to repo

So what's the latest cool super metroid romhack?

Last one I played was Hyper.

Ok, so I'm making a simple Castlevania: Adventure (GB) bootleg to make it look like a game of another franchise. I've got tileset editors and a friend who is able to assemble stuff and create a legit cartridge out of it, but I'm stuck at one particular point; I cannot find an utility for creating music compatible with GB. Any recommendations? I'm pretty desperate.

>>4051214
Y-Faster is pretty rad, though it's a short gimmick hack.

If I don't understand what the fuck you guys are talking about, but I do know Calc 2 and an entry level CS course, then what do I need to do before I can understand how you guys are actually tracing code registers and shit to figure out how to change them? Like, is everyone here a CS grad or is it not that hard?

>>4050446
Why is he doing that...

>>4045603
>I'll upload the code later when I clean it up a little.
Last online: 4 years ago

>>4051758
what are you talking about, i uploaded it to github a few days ago and have already added more >>4046959

Oh how I wish for an actual romhack of oracle of seasons / ages. Those games have so much potential to be romhacked and make a new entry altogether.

>>4051974
Do it

>>4050446
I looked it up, and the "fanfiction shit" is all straight out of the official Link's Awakening manga. So... actually I dig what he's doing.

bumping with more shit.

>>4014797
Sonic 2 long version does this
It's a platformer but I recommend for you to check it out

>>4053764
Sonic 2 Long Version's new zones are shit though

>open up the secret files in BNW
>in the technical FAQ the author literally calls people "someponies"
>dedicates the game to "Princess Luna"

>Sans got that DIS FUCKING GUY expression on all the time

Any cool new hacks come out in like, the past 2-3 years? Haven't really posted here since late 2014.

Preferably Mega Man and SMW hacks

>>4054273
VLDC 9

>>3991980
That's one of the reasons I loved Quake 1 so much. Models are limited to a very small amount of polys, most textures are 32x32 and 128x128 at most with a limit of 256 colors, and a big portion of game coding is kept separate from engine coding so you can more easily add stuff in by learning simple QuakeC.

Not romhacking though, so I digress.

>>4054194
I'm so proud of our species to see that that kind of autism is able to travel through time and space and make me giggle

>>4049673
this.

I have played SMRPG into the ground. This actually makes it pretty damn challenging. OHKO's means you actually have to pay attention. Nice to be doing 1k+ damage by croco too.

>>4054273
Pokémon Prism finally got almost-finished (it got C&D'd right before release and was leaked with some of the content not done, but there are unofficial patches to fix most of the bugs from the leak).

tfw hack willow to modify 1 piece of the map

Psst! What's the password for the Brave New World extra zip?

Don't die

yoo so apparently the newest POC||GTFO has a chapter called "Reversing Pier Solar"
get hype

>>4054416
Is it any good?

FPGA is easier than I thought. Making a VGA image was a snap. It would be totally possible to make a custom tile-based render circuit, like the SNES PPU or the MD VDP. There are so many possibilities I almost don't know what to do.

So do we all agree that n°2 is more faithful and just better?

Initially I didn't want to use that colour because it made Protoman look like he had jaunedice... but now I'm thinking of changing back, especially because of how bad the charge look when I play on my CRT TV

Does anyone know where to get or have any information on this Willow nes hack ?

http://www.romhacking.net/forum/index.php?topic=5832.0

supposedly there is a partial willow disassembly too

>>4058652
how is 1 or 2 more faithful? they both have completely wrong background color

>>4058652
damn i wish i could see that

>>4058959
it's a png, if it's blurry it comes from your computer/browser.
In this case, save it, and open it with paint.

For now I've settled for an "in between" solution, I kept the face colour beige, like in 2, but I set the body colour during charge as that gold yellow, like in 3.

The hack for Phantasy Star II is old in the mediafire folder. Latest version is 3.1 and mediafire has 3.0.

>>4057162
I personally think it is. It's not perfect, and the difficulty curve is a bit rough occasionally (you get to an early cave and it has a huge encounter rate and really tough trainers relative to what you've been used to, though each trainer is arguably a miniboss given the game's plot), and though they were supposedly going to expand the postgame with post-release patches if not for the C&D, as is the Rijon region (the region from Pokémon Brown, which is the first hack Koolboyman made) has a lot of the same issues as Kanto from Gold and Silver, where there's not a lot in the region besides the Gym Leaders to fight. It's still pretty nostalgic to go through Rijon again if you played Brown before Prism, but it overall feels a bit empty. It really encapsulates early Pokémon game design though while also having the improvements from later generations like abilities and the physical/special split, and for someone like me who's been fed up with Game Freak casualizing the already-easy games more and more with shit like XY and SM, it was the best Pokémon game I've played in a while.

>>4058652
>especially because of how bad the charge look when I play on my CRT TV
Nice that you're testing on a CRT. I know a lot of romhacks and homebrew where stuff aren't visible on CRT due to the mask, it breaks my heart that they haven't tested it properly, almost as much as those romhacks for mainly N64 that only work on emulators.

>>4051362
Check out here:
https://github.com/avivace/awesome-gbdev

>>4051746
There's some emulators with debugging/romhacking features. FCEUX for NES has all kinds of tools for picking a rom apart while you play.

>>4059553
Don't forget that roughly 85% of the Super Mario World hacks do not work on real hardware, and at least 5-10% of the remaining hacks will freeze or glitch out. Of all the ones I've tested, so far Super Mario 3X is the only non-Kaizo SMW romhack that actually works flawlessly.

>>4059610
SMWCentral don't allow hacks that don't work on real hardware anymore though. It was caused by the method used to add custom music.

>>4059553
It's not so much playing on a CRT that I care about but playing on real hardware.
It feels great playing the hack on a Famicom plus if it doesn't work on real hardware, I might as well make a PC fangame and not a romhack.

Anyway I've tried my best to make the pal swap of the charge shot charging accurate with MM9/10, but when played on TV it looks weird, as if the different pals mixed together and the face looked purple, even though the face isn't affected by pal swaps.

It doesn't appear to be a problem on emulators

https://www.youtube.com/watch?v=QjRaRJtZIq4

I don't think there is much I can do though because when I compare with 9/10 side to side, this looks very close. Perhaps mine is a BIT faster to perform pal swaps but not from much, and considering how the pal swap is coded the best I could do is double the time for each palette which wouldn't be right.

Bump

I'm working on a dash attack inspired by Protoman's dash move from MM7 and the Power Battle series, except it'll dmg enemies.

It'll replace Beat, so only accessible once you collect all the letters.

In retrospect it's very similar to Charge Kick, which is why unlike c.kick I'll make it work in mid air (once per "mid air time"), do more dmg, work on shielded enemies, but also eat up more energy.
You can't shoot or slide while the weapon is equipped, only dash move, but the dash move itself will also deflect bullets.

So it should be a bonus, optional, very powerful attack that you have to use sparingly. If I make it work in mid air you could also use it to help reach places or skip spikes.

https://www.youtube.com/watch?v=nf0jk5pE82E

it's all WIP, thoughts?

>>4062498
it needs to look stronger!!!!!

>>4062575
I know you're right, I'm not very good at the art thing.

The main issue is making this look good, and powerful, while restricting myself to a tolerable amount of 8x8 sprites.

These frames have between 12 and 15 sprites which is already a lot. MM have between 8 and 10 tiles and that already flickered a lot in the OG.

I could save some if I sacrificed the colours, that is to say, if the 'effect' used the same palette as the body, I could draw both in a single sprite and not have 2 sprites on top of each others. This is what charge kick does. However, I could never get that to look decent for the life of me.

For instance MM Unlimited has a great looking "comet dash" attack but he didn't have to worry about tile count and flickering...

http://www.megaphilx.com/Game/Sprites/Heroes/MMUSprites-Megaman.png

>>4060378
Sure, they SAY that, but I highly doubt they are actually testing that. Case in point, The Second Reality Project Reloaded. Patched a proper rom, everything is golden. Yet everytime I beat the first castle on my SNES, the game hangs. They have it listed as working on real hardware.

I think they just play a level or two at best.

>>4062498
What about having his palette flash red, yellow, and white, and maybe adding a speedline or two behind him?

>>4062741
I'll try something else from scratch tomorrow. Something more whole like Comet Dash in MM Unlimited, and see what's the tile count then. Maybe it wouldn't be so bad.

>>4062601
Maybe the boys at /mmg/ can help you out in the art department
Just a thought.

How long until the next big SM romhack

>>4062965
Dunno. X-Fusion seems to putting along at a "when it's done" pace.

>>4041504
>Single quotes ' omit the null.
wtf. single quotes enclose character constants, not strings. of course they don't have a null.

Bump

>>4065045
Seconded

What improvement hacks would you recommend for a first run of Super Metroid, if any?

I heard Project Base mentioned earlier.

>>4067505
First run? Nothing.

>>4067505
>>4067538

This. Go with the original game on the very first playthrough.

>>4067505
http://www.romhacking.net/hacks/3211/

>>4069124
No.
This is for after you've played the original game, if you want to have a different experience.

Bump

Alright, I have a challenge for you all.
Make the biggest impact on a game by changing only one bit.
Not a byte, not +1 or -1, but flipping only one bit.

>>4070589
As an example, here's one for Megaman for the NES:
Rom address 1513D: 1C -> 0C
Starts you off with 40% life.

>>3989831
I once saw a screencap of CV1 with 16bit graphics (wasnt chronicles) and its meant to be a texture pack that can be loaded on an emulator or something. Any idea how I can find/make these? Thanks.

>>4070589
that's an interesting challenge and reminds me of something I did years ago – I was dicking around with a hex editor, trying to find the level format of the Wonderswan game 'Rockman & Forte: Mirai Kara no Chousensha' and changed one byte somewhere only to find out it was some kind of a global game speed byte. The game as it is is painfully slow and the bullets you shoot practically crawl through the screen, but that single byte made everything so fast that it felt like a much more polished game.

so this just happened: https://github.com/pret/pokepinball/commit/e235064ebc8bc17cb41eeed9e456f5b301433729

now you can rearrange things in the pinball disassembly however you much you want without breaking pointers. it's basically "complete"

the list of complete disassemblies is now pokered, pokeyellow, pokecrystal, pokepinball (pokeruby sort of counts, it's completely disassembled but not yet entirely decompiled to c)

wonder when they'll get poketcg and pokepuzzle and pokegold done

>>4072008
excellent disassembling friendo

>>4072008
Is there any reason to use pokegold over pokecrystal besides the cooler intro?

>>4072748
pokegold isn't even close to being finished

it'll be way easier to just port the gold intro to pokecrystal than to deal with an unfinished disassembly

not that pokegold is bad. it's just that removing that last incbin (which pokecrystal did years ago) makes a huge difference in a disassembly's usability

Huge megaman fan here, just found out all these mods that are out for it that I'm seeing in the pastebin. What should I play first to get a feel for what these can do?

>>3989831
My project showoff MUSHA hack :D

>>4073559
>pastebin
>MM romhacks

Shit, I hadn't checked it in a long time. Rockman 6 innocent is in there, I had no idea this was even released! I saw a video a long time ago and it looked awesome, but thought it was cancelled.

Also thanks for putting my hack in there

Hey does anyone have a recommendation chart for sonic rom hacks

>>4031201
Thank god your still around!

Is there a patch for the FLUDD assets?

https://github.com/Fiachra1993/SuperMarioSunshine64DS

>>4075178

I don't have a chart, but I can recommend checking out all of these in no particular order for one reason or another:

Sonic Classic Heroes
Sonic 3 Complete
Sonic 3 - The Challenges
Sonic 3 - D.A. Garden Edition
Robotnik's Revenge
Sonic 2 Advanced Edit
Sonic 2 Retro Remix
Sonic 2 Recreation
Sonic VR
Sonic Megamix
Sonic Regen
Sonic 1: The Next Level
Pana der Hejhog
Jester's Challenge
Big's Fishing Derby
Knuckles Emerald Hunt
Metal Sonic Hyperdrive
jubbalub's Untitled Sonic 1 Hack
Sonic Boom
Sonic Zoom
Mobius Evolution

>SM Redesign: Axeil Edition

is it easier or harder than Redesign proper?

>>4075992
Easier. The author took all the criticism of the original hack to heart, went through it with a fine tooth comb, and fixed most, if not all the problems people had with it. And on top of that he added a bunch of QoL features, such as altering the map tiles so that it's easier to read, or adding an auto-morph for those morph ball tunnels you have to jump and mid-air morph to get into.
He also completely redesigned Tourian.

You all made me do this...
http://www.mediafire.com/file/5gelklth837pox5/SouthPaw%27s_first_Sonic_Hack.BIN

>Playing Super Mario Star Road for the first time
>Get to Hidden Palace Finale
I think I'm at my breaking point now. I managed to get through everything else without savestates using sheer stubbornness, but I think the game has truly worn me down at this point.
Anyone else had trouble with this rom hack?

>>4077231
Star Road is a broken hack. It's full of bugs and shit that just doesn't work.
I don't understand why everyone sucks it's dick.

>>4077289
It's cause it's really the only"full game" sm64 hack

>>4077372
What about Last Impact?

>>4077376
Haven't heard of that one. Regardless, there are very few.

>>4077378
Oh man, you're in for a treat then
https://www.youtube.com/watch?v=JG4PCVfu4qo

>>4075750
>Robotnik's Revenge

The second one is miles better.

>>4077289
I did end up getting all 130 stars. I did enjoy it for the most part, but man did it get pretty frustrating at points.
He's making a remake of the hack for the DS remake of Mario 64. It seems he's dialing back on the difficulty. Combine this with hopefully less bugs, and I think I might end up enjoying that a lot more. Even has a program that will allow you to use an Analog stick with full-analog movement while emulating it, taking the DS remake's fatal flaw.

>>4077289
>>4077372
>>4077376
>>4077378
>>4077381
Last Impact is garbage. Star Road is amazing.

>>4080682
Yes, it is amazingly broken.
I'm glad we agree.

>>4080730
Just like the original game.

>>4080731
The original game only breaks when you're actively trying to. Star Road is broken in regular play.

>>4080738
lol no, I accidentally found lots of bugs when I was a kid just by playing it normally for the first time

are there any hacks of SOTN that add the underground garden + other content from the saturn release?

>>4080760
I'm sure you did, Jimmy. Does you dad work at Nintendo, also?

>>4080775
what

I've done a little bit of looking around and not really found much, so I thought I'd ask here - is there a hack of Gunstar Heroes where you play through the game with just your fists (no guns), like in the curry + rice fight?

What's up with Rockman 6 Innocent ? Is this a glitch or is the hack just a demo?

I've tried several roms and emus

>>4077376
>>4077381
Is Last Impact harder than Star Road?

>>4081560
It's harder because you'll have to fight against boredom as you play. That's how boring it is.

>>4081607
It looked like it had a bunch of new stuff, is Last Impact just not that interesting?
And genuinely, is it harder? If Star Road was much longer after the Hidden Palace, I don't think I'd have gotten every star.

>>4081607
That's exactly how I felt while playing Star Road!

>>4081607
>>4081629
Wow! Differing opinions! Cool.

>>4081626
Well dude the patch and rom is free, you'll just have to play it and see.

>>4081630
Why do you sound incredibly mad?

>>4081637
Do I?

>>4081640
You totally do.

>>4081629
Your fault for going in the only 2 boring levels in the entire game as the first thing you did

I can't make this fucking dash work correctly in mid air.
Even when I look at the code of the "air slide" hack for the same game, which has a very similar result, I'm lost between all the movements routines that get in the way.

I'm sick of this so atm I'll stick to making the dash work on the ground only. Let's just pretend it's because it's closer to how Protoman's dash works in MM7 and Power Battle.

>>4081525
what is going on in that picture??

>>4082021
Just an empty unfinished room and you can't go on. Apparently it's just a demo, there is only about 4 or 5 stages done.

What is done is impressive, with some great ideas, but the author isn't good at difficulty balancing, which is a shame because with his skills this could be great.

>>4082067
did u google anything

buh-bumppp

>>4053717
any idea how this is coming along?

>>4076040
Axeil is GOOD

How does S&K's Lock ON shit work? How do two games become one

For example can you make a lock-on enabled ROM hack?

>>4084229
Yes it is, Anon. Yes it is.
It's just too bad people who were turned off by the original Redesign won't give Axeil a chance.

Are there rom hacks of the sonic games that moves the camera ahead of the character instead of keeping them in the center?

I've played the games countless times but I'd like to see how a camera shift can effect them.

>>4084884
What's wrong with the original Redesign? I liked it

>>4084824
it's simple.

The main program is on S&K.
It checks if the connectors are connected on the lock on cart.
If it is, it addresses the top rom, reads the name of it, and runs the boot code for STH2&K, or STH3&K.

Then, it runs a program off the Lock On Cart (S&K) that references data in the top cart's addresses and itself (stage and music data).

Then it patches the code to include expanded areas in the S&K cart (hidden areas only Knuckles can reach, etc)

All you need is a conditional program that checks the contents of the extra cart and patches the rom as it runs.

https://retrocomputing.stackexchange.com/questions/1514/how-exactly-does-sonic-knuckles-lock-on-technology-work

>>4085640
I didn't think it was too bad either.
A lot of people didn't like the original's physics, especially underwater. Especially if you ended up in quicksand.
A lot of people despised the Hell Run to the Ice Beam.
A lot of people found the map to be pretty confusing to read.
I'm sure there are other issues I'm forgetting, but those are some of the ones I see complained about the most.

Goodbye Beat, you were a good friend.

>>4077381
>[DMCA'd]
So I guess Nintendo's been taking out even more ROMhacks since Prism?

>>4086918
I bet they (which could be the owners of the site he uploaded it to) just removed random ROMs and Kaze took the opportunity to make it more popular.

>>4037914
They all look great except Bocke, the old one was better.

Who is making this? Saw it in another thread.

>>4087950

I think that's just a mockup.

>>4085676
Wow that's an awesome website... Really detailed answers

>>4037914
Tried to make Agumon look more like he does in the v-pet and games

What kind of new SMW romhack have come out in the last year?

>>4041970
>>4041986
I realize it has been a month and I apologize for my hiatus (had personal matters come up that has kept me from the hobby and from here), but I would like to sincerely thank you for your help, and let you know that I am now about to resume my journey in C programming and, eventually, rom hacking.

>>4089337
How dare you?! If you ever ask another question here again, I'll drop a whole load of crap for you to think about. That will teach ya!

>>4088527
The last "big" SMW hack I can think of is The 9th Vanilla Level Design Collab, but it was released in 2016.

>>4091046
That fits the descriptor of "in the last year"
It would count technically to 06/30/2017

>>4088047
Looking good, dude!

>>4084156
Yeah, it's going fine. I'm just taking a break and doing other stuff.

>>4053717
how many of these are evolutions? does simon belmont evolve into dracula?

>>4053717
shouldn't optional/alternate be changed to trainers?

>>4053717
I remember that space in which Zero Suit Samus stands being labeled "Varia Samus" at one point. What inspired you to change your mind?

>>4092306
Does it matter? Not all of those are trainers.

>>4092420
Because I want to get this project over with already. If someone makes Varia Samus then i'll gladly change it. But I have Zero Suit right now so that's what I'll put in the meantime.

>>4092303
They can if you want

>>4091046
I was thinking of trying that out. Is there a recommended emulator for that? Or is Higan 1.03 just fine?

>>4092504
No I prefer Zero Suit, I was only wondering.

Gotta work out the bugs but this goddamn dash is finally shaping up

https://www.youtube.com/watch?v=hOtACC4b0gI

A friend of mine was a big help in making the animation. I ditched the old one.

>>4093110
nice dash, rude language

https://www.youtube.com/watch?v=KWIV0Oz-3Bs
w h o a

I wonder what the other codes do now.

>>4092529
It should work just fine on all major SNES emulators as well as the real hardware.

>want to make gameboy game in C
>too retarded to get a background to display
>extremely poorly documented
>gone through a few tutorials but nothing is working
>error message unhelpful

For the love of god someone help me

I can make a sprite move around the screen but any kind of background or map data is making it shit itself or display a black screen

>>4095115
Okay, got it to display a garbage version of the background.

I'm learning

>>4095271
>>4095115
That's how we all picked this shit up.

My advice is bite the bullet and learn Z80 assembly. It's not that hard, really, and if you know C, understanding how C gets mapped to assembly will be extrememly illuminating.

With this skill you can use a debugger and make sure your program is doing exactly what you think it is doing. If something unexpected happens, you'll either realize you misunderstood what the CPU does, realize you have written a bug, or realize that you misunderstood the hardware.

If there is a dev manual anything like the SNES, I would go through the section describing how to use the background, and pay VERY close attention to how registers are initialized and then set. It's REALLY easy to forget to set a bit here or there and get garbage out.

Lastly, it's impossible to help with the specifics unless you post your code somewhere so others can review it. You should start a github repository and upload what you have there. If you do that, I'll take a look at your program and see if I can offer advice.

Every time you make a teensy bit more progress, you'll cement hard-earned knowledge into your head, and feel exhilarated seeing your creation come to life.

>>4095115
>gameboy game
>in C
Aren't you supposed to use assembly for old specific machines like this?

>>4095476
Not him, but I know that the dev manual exists. It's really complicated though, isn't it?
Also, looking for it on Google I found this huge resource of information I had no idea about. I would have read at least some of it by now if I did.
https://github.com/avivace/awesome-gbdev

>>4096226
look up SDCC

>>4095476
Yeah I'm looking into assembly now, never learned it before so seeing if I can find a good guide, it'll be much handier to do this kind of thing once I can understand it. I also want to start doing some disassemblies, so got to get some ASM under my belt. Any tutorials for assembly in general you'd recommend?

Good advice on the repo, I've been meaning to make a 4chan-specific one (for various /g/ projects too), so I went ahead and made one, I'll throw up the tiny bit I have written some time today. I'm trying to make a short RPG (or Yume Nikki-style exploration) to get more familiar with it.

>Every time you make a teensy bit more progress, you'll cement hard-earned knowledge into your head, and feel exhilarated seeing your creation come to life.

Looking forward to it, it's been a lot of fun so far!

>>4096226
For production-quality yeah, but for gameboy there's the GBDK, which compiles C for it.

iirc there were actually a few games back in the day (for Genesis and NES anyway, can't remember where I read it though) that were written in C, but they tended to be laggy as fuck

>>4096561
>GBDK
>2001
you might as well build your own toolchain

bump

>Introducing the FCEUX Debugger

See animated image for a rundown of the Debugger window.

The Debugger allows you to examine the state of the NES at any given time and view the code as it is executing. 6502 Assembly and the NES architecture in general is beyond the scope of this post, but resources are available online that describe these topics in detail.

When a game is running, if you just pause the emulation as it speeds through frame after frame, there is no telling where it will stop, so we use breakpoints to stop the program where we want it to. The breakpoints are a trap you set on a certain address (or range of addresses) that will pause the program when it tries to do something to that address, such as read from, write to, or execute and instruction on that address (or any combination thereof.) Conditions can also be specified ie: only break when you write 0x05 or higher to this address.

Once the breakpoint has triggered, the code the will display in the left window. To the right you will find all the different sections of the processor, such as what the current stack looks like, the contents of the registers and so on. By clicking the buttons near the top of that section, you can step through the code and watch it execute. "Run" simply continues running the program as normal. "Step Into" executes only one line of code and will follow into any subroutines that may get called. "Step Over" executes the next line of code but will do a subroutine before breaking again on the line after it returns. "Step Out" runs until the next RTS command, useful for getting out of a function and getting to the code that called it. "Run Line" will execute as much code as would be done in the time it takes to complete one scanline. "128 Lines" does the same but for about half a screen worth of scanlines.

Bookmarks can help you identify locations in the code.

Hint: Double clicking a breakpoint will toggle it on and off, and hex editor will highlight recently changed values.

Any good dbz rom hacks worth playing?

>>4090123
Sorry. I had a death in the family and have been extremely busy. I am back at it now though.

>>4098759
Thank you for this!

>>4098847
My condolences. Hope you are feeling better.

Fuck year, figured out backgrounds

Sprite moves slow as fuck now because I'm probably implementing it in a shitty way, but definitely progress.

Interesting to see how the GB renders things, definitely a better way to do it but I'm having fun learning about how the machine works.

Need to learn more about ASM so I can read that 8-Bit Wonderland doc though

>>4098849
I got one for using the cheat search, but need to redo the image.

Which hack for Pokemon FireRed would you suggest: FR251 or Throwback?

anyone have any idea how to edit the echo buffer in SMW tracks? trying to get the second reality project reloaded to work on higan

rec me some sanic hacks /rhg/.

>>4099649
scroll up, fella
>>4075750

>Finding Variables With Cheat Search for FCEUX: A Primer

See animated image for a rundown of the Cheat Search window.

One way to find where data is stored in RAM is with the Cheat Search. The Cheat search works by making a list of every ram location and the value in that address at that point in time, then paring it down based on how that value changes.

So to start play the game until the variable you want to track down is sure to be loaded and hit the "Reset" button (on the Cheat Search Window!) That will populate the list on the right with every address in ram. Now start playing a bit of the game, then narrowing down the results.
If you think you know the exact value, you can put it in the box next to the "Known Value" button and hit the button.
If you know the value hasn't changed, use the "Equal" button.
"Not Equal", "Greater Than", and "Less Than" are as they say, but if you know exactly how much the value has changed by, you can put that in the box next to the button, and check the check box between them, to specify the parameters for that search.

This way you can easily narrow down the possible addresses to a few locations. If you mess up, or don't find it, just hit "Reset" and start over.

Some things to note:
1) This can only do 8bit values, anything that goes higher than 255 will be stored among two or more bytes and this cheat search will be of less help.
2) The value may not be kept as straightforward as you think. It may be stored in an unpacked Binary Coded Decimal format, or based on the display tile, or maybe the value is calculated as needed whenever called for.
3) The search only acts on the values in the results of the previous search. "Reset" will start the search over with the current ram. Keep this in mind if the game happens to briefly swap chunks of memory around at regular intervals. Notice the option at the bottom of the window.

>messing around with GBDK
>using spritesheet related, from a disassembly someone did of Pokemon Red
>implement directional turning for the sprite before I start working on a walking animation
>huh, that's weird, Right isn't working properly
>realise there's no right-facing sprite
>those hacks at Game Freak just flipped the sprite every time you pressed the Right button
>now have the choice between rewriting the code around an entirely different spritesheet, or writing a load of spaghetti to flip it each time

FFFFFUUUUUUUUUU

>>4101114
I say use a code to mirror it. That's what MegaMan do as well. I assume many games do the same to same graphics space, imagine if you needed to draw every sprite that needs it the other way, even if it looks the same just mirrored...

In MM there is a "display flag" ram address, one of its bits determines if it's mirrored or not.

>>4101131
Implemented something in move_player_sprite() to handle it:

https://github.com/Siro451/gameboy_rpg/blob/master/game.c

Probably a very shitty way of doing it, but all part of learning.

Implementing walking next, then that's the basic sprite movement done I think. Just need to abstract it a bit more so I can use it for any sprites.

>>4101332
hahah excellent
i assume set_sprite_prop() is setting that bit that NESfig refers to