Brainstorming: how to get capcom to add functional netcode to future fighters

You know that article was talking about the game mechanics, rather than the online. No one on SRK thinks the netcode is good.

that would be cool but youre ignoring the reality of every single review being bought

Well no shit. That doesn’t stop the article from being a senseless piece put on the front page in shallow hopes of shaming people into liking MvC3.

Oh. Oh yeah.

yea i mean im bein kinda wacky there but really when does any review manufacturer criticize something

maybe imma make a vid of me yelling at my tv while i mess w/ input delay and the videos in slow motion so people are like “nothin happens for eons after he presses the button”

Jojo’s will have the same problems that 3s has(maybe even more) if same amount of people were in the room. Also Petshop isn’t banned.

It’s a Capcom USA vs. Capcom Japan thing. Capcom USA obviously wants to use the code, hence we see them first use similar implementations for HDR and MvC3 before actually using it in FF:DI. Meanwhile Capcom Japan seems as if they have no idea how to make it work.

Frankly, I still think it’s just that Capcom Japan doesn’t see any fiscal gain in trying to make good netcode work, when they can just jury rig the crap they’ve been shoving down their customer’s throats so far.

They’ve clearly done things to help compensate for input lag. Extremely easy basic chain combos in MvC3 ( one complaint from SSF4 was how difficult tight links were, especially online ), lag training mode. And as I said, if you’re an unexperienced player, you probably won’t even notice input lag if it’s in the 1-5 frame region. Anything beyond that they can excuse away with “well it was just a bad connection”.

also this.

There are plenty of people out there fully aware of the lag who just don’t give a fuck. Look how many damn lag switchers are online in both SSFIV and MvC3 just for starters, and then all the Blanka/teleport spam/etc. lag abusers beyond that.

Such is the incredible suck of the internet.

there needs to be a skullgirls online multiplayer demo so ppl can get on the sweet netcode crack

If the net code lives up to the expectations I’ll hype this game to no end.

Then something needs to be done to get them to change their mind. Most obviously, support those games that do go out of their way to use GGPO or similar functional netcode. Heck, here’s hoping Namco (who actually seems to have a working implementation of GGPO for a 3D game) uses it for TTT2 or TxSF just so Harada can rub it in Ono’s face.

That’s a start, but I don’t see it being a good solution as long as the large base of players who don’t recognize modest amounts of input delay is as … well… large… and vocal as it is. Basically, you have to win the hearts and minds of groups like Capcom-Unity.

Honestly I think that Namco suggestion could even backfire on us. Again, with input tricks and just in general poor eyesight, people will brush off 1-5 frames of input delay as “lagless” when they might notice quirks induced by rollback code at that range, in which case you’d get a very real life “outplayed by stupidity” moment.

I harken everyone back to my big idea. I never received any feedback, so I’m under the assumption no one read it.

I harken everyone back to my big idea. I never received any feedback, so I’m under the assumption no one read it.

yeah me too. also skullgirls is on a 3d engine yeah?? like the “sprites” are just rendered onto 3d models or something right??? so theres another game for the 3d-ggpo list i think

i thought someone said it was being used for ttt2? anyway good net code would be a sweet thing for them to advertise, especially if it was a point used directly against capcom fighters

This is a point that’s going to confuse a lot of people, so it needs to be clarified.

Skull Girls is a 2D game. It uses “sprites”. That’s because every 2d game now uses “sprites” in quotes because of how modern hardware works. All 2D elements in modern games are, at some level, flat bitmaps being moved around by geometry with no z-axis - no depth. Blaz Blue, etc etc.

The difference is that their animation simply consists of a series of individual bitmaps that can be called up instantly (from ram) to adjust the position or animation timing of the characters if rollback happens.

The argument for problems with 3D fighting games and rollback code, comes from characters who are constructed of three dimensional polygons and animate by following vertex data. In other words, if Ryu in Street Fighter IV throws a fireball, none of his animation pre-exists. He follows a script (essentially) that moves the limbs of his polygon puppet around to make him throw a fireball.

If a character in Blaz Blue throws a fireball, the game is merely flipping between a number of pre-rendered frames that make it look as if the character is throwing a fireball - traditional animation.

The whole point is that it takes a lot more processing time and resources to move a polygon puppet through a set of animations, than it does to simply flip back and forth between a few pre-drawn “sprites”. The theory goes that the lag created by the game having to literally re-render all the animations of polygon characters and any other special effects going on, on top of that, may nullify the appearance of saved latency on a local client. Flipping through a set of pre-drawn sprites = fast. Re-rendering 3D characters and their special effects too, quick enough to seem to have zero input latency during rollback = not so fast.

If Tekken Tag 2 does in fact use some degree of rollback code, it may be possible due to the fact that Tekken doesn’t have to animate as many synchronized 3D objects as some games, like oh, Marvel vs Capcom 3, which is a clusterfuck of geometry on the screen.

[media=youtube]w_2u7-K40rQ&t=2m[/media]

It will be interesting to see how 3rd Strike Online plays considering that they have actually used the word “online” in the title.

This game should represent their best attempt.