SMBX2 (2.0 Beta 3)

[The0x539]

I'm not sure if this is the right place to say this, but X2 is crashing my game. When I trigger an event, if I don't change it to at least 0.1 the game crashes, not that big of a deal, but mildly annoying. It also crashes when I exit full screen. Do you want me to send you the crash log or are these bugs already known?

What's the "it" that you have to change to 0.1?

[Artemis009]

I'm not sure if this is the right place to say this, but X2 is crashing my game. When I trigger an event, if I don't change it to at least 0.1 the game crashes, not that big of a deal, but mildly annoying. It also crashes when I exit full screen. Do you want me to send you the crash log or are these bugs already known?

What's the "it" that you have to change to 0.1?

"It" is the "trigger event" option you can use while making events. The time before it triggers the event is customizable, but leaving it at 0 crashes the game.

[ElectriKong]

I wouldn't really call writing a text file a proper enemy/npc editor and the enemy editor I'm familiar from certain other programs also allows you to create new enemies not just change properties of existing enemies.

Stuff like that is hard to do and isn’t priority right now. SMBX2 is still beta and a tool to create new npcs isn’t likely to happen anytime soon.

At least scripting is possible which one could say is more expansive than using an editor but is a lot harder to do and require knowledge of how to code.

Well if you know a way of creating AI without, you know, creating the AI, then I would like to know as well.

[Amyrakunejo]

I'm not sure if this is the right place to say this, but X2 is crashing my game. When I trigger an event, if I don't change it to at least 0.1 the game crashes, not that big of a deal, but mildly annoying. It also crashes when I exit full screen. Do you want me to send you the crash log or are these bugs already known?

What's the "it" that you have to change to 0.1?

"It" is the "trigger event" option you can use while making events. The time before it triggers the event is customizable, but leaving it at 0 crashes the game.

Likely some sort of index error.

Given I know that there are parts of any programming language that are finicky about zero values (can't tell you how many times I've run into the divide by zero error in Ruby), that could be it, or partly. But...not sure.

[PixelPest]

What's the "it" that you have to change to 0.1?

"It" is the "trigger event" option you can use while making events. The time before it triggers the event is customizable, but leaving it at 0 crashes the game.

Likely some sort of index error.

Given I know that there are parts of any programming language that are finicky about zero values (can't tell you how many times I've run into the divide by zero error in Ruby), that could be it, or partly. But...not sure.

I've honestly never had this happen with events in SMBX

[Emral]

The only case I can think of where this might happen is if you repeatedly trigger the same event, causing an infinite loop without any wait time.

Sizeable graphics work starting at a size of 2x2. This has always been the case. The "bug" is that PGE displays them wrong.

Well could you make SMBX2 display them like that too?

Can this be answered?

No we can't display them like PGE does. PGE renders these tiles differently. Just use non-sizeable semisolid blocks. You can still resize those to be taller than 1 tile if needed (by resizing the graphic file).

[Goomba98]

This just occured to me: Can you make it a vanilla thing that Bloopers can "swim" in midair, like in The Lost Levels?

[Emral]

It's easier to write a new blooper. We might do that at some point.

[Artemis009]

This just occured to me: Can you make it a vanilla thing that Bloopers can "swim" in midair, like in The Lost Levels?

You could always slow down a chasing para goomba to imitate it, obscuring the regular gonna of course.

[TheTrueMarioMaster]

Stuff like that is hard to do and isn’t priority right now. SMBX2 is still beta and a tool to create new npcs isn’t likely to happen anytime soon.

At least scripting is possible which one could say is more expansive than using an editor but is a lot harder to do and require knowledge of how to code.

Well if you know a way of creating AI without, you know, creating the AI, then I would like to know as well.

Scripting doesn't require creating an AI, Lunalua could be considered a form of scripting (though to be fair before SMBX 2.0 or even Lunalua was created I didn't think scripting or any kind of game changing modification could be added to Super Mario Bros X 1.3.0.1 because of everyone saying it would be impossible to decipher Redigit's code at the time). Though speaking of AI something that would be very challenging would be writing up a functional NPC that acts like a Player Character for a one player battle mode.

[HenryRichard]

It's easier to write a new blooper. We might do that at some point.

Does setting NPC+0x1C (Is Underwater) to 2 not work?

[PixelPest]

It's easier to write a new blooper. We might do that at some point.

Does setting NPC+0x1C (Is Underwater) to 2 not work?

No it doesn't. Fails to trigger underwater AI but has the other effects

[ElectriKong]

At least scripting is possible which one could say is more expansive than using an editor but is a lot harder to do and require knowledge of how to code.

Well if you know a way of creating AI without, you know, creating the AI, then I would like to know as well.

Scripting doesn't require creating an AI

Creating a new NPC does though

[Miguel CG]

Can you add Canadian French and European Portuguese, please?

For a game editor, having European and Canadian French separate is pretty trivial. The dialects are extremely similar

Well, I'm saying this because in Nintendo America, the languages are English, Canadian French, Spanish, Portuguese. And Nintendo Europe languages are British English, French (France), European Spanish (Spain: Castellano), European Portuguese, German, etc. Because I don't know how can I translate in French of France, because of reasons...

[TheTrueMarioMaster]

Well if you know a way of creating AI without, you know, creating the AI, then I would like to know as well.

Scripting doesn't require creating an AI

Creating a new NPC does though

I think we are thinking of two different AI's here because when I think of AI it's a complex program capable thinking near human level intelligence. That's something a Mario enemy doesn't need. Though I think your referring more to coding new enemy behaviour which is a lot more possible than you make it sound considering new enemies have already been added to the beta.

[Hoeloe]

Scripting doesn't require creating an AI

Creating a new NPC does though

I think we are thinking of two different AI's here because when I think of AI it's a complex program capable thinking near human level intelligence. That's something a Mario enemy doesn't need. Though I think your referring more to coding new enemy behaviour which is a lot more possible than you make it sound considering new enemies have already been added to the beta.

What you're talking about is more usually referred to as "machine learning", and isn't usually what people mean when they say "AI" in the context of games. Very few games use actual learning algorithms to control non-player characters. The vast majority, even those with enemies that appear intelligent, use basic techniques like search algorithms and simple behaviour trees.

In a gaming context, "AI" usually just refers to any algorithm controlling a character other than the player, regardless of what that algorithm actually is.

[TheTrueMarioMaster]

Creating a new NPC does though

I think we are thinking of two different AI's here because when I think of AI it's a complex program capable thinking near human level intelligence. That's something a Mario enemy doesn't need. Though I think your referring more to coding new enemy behaviour which is a lot more possible than you make it sound considering new enemies have already been added to the beta.

What you're talking about is more usually referred to as "machine learning", and isn't usually what people mean when they say "AI" in the context of games. Very few games use actual learning algorithms to control non-player characters. The vast majority, even those with enemies that appear intelligent, use basic techniques like search algorithms and simple behaviour trees.

In a gaming context, "AI" usually just refers to any algorithm controlling a character other than the player, regardless of what that algorithm actually is.

I see, in that case the complexity of the AI involves what the enemy is capable of like how a Chargin' Chuck AI would be more complex compared to that of a koopa troopa. Thinking about it out of all the classic Mario Platformers I would say Yoshi's Island enemies tend to be the most complex in behaviour, even some of the more simple ones like shy guys and goombas in that game. Speaking of which are any Yoshi's Island enemies going to be added in the future or is the main focus just the first three Super Mario Bros plus Super Mario World. Though to be honest I do expect to see a lot of Super Mario World stuff in this project.

[Artemis009]

I think we are thinking of two different AI's here because when I think of AI it's a complex program capable thinking near human level intelligence. That's something a Mario enemy doesn't need. Though I think your referring more to coding new enemy behaviour which is a lot more possible than you make it sound considering new enemies have already been added to the beta.

What you're talking about is more usually referred to as "machine learning", and isn't usually what people mean when they say "AI" in the context of games. Very few games use actual learning algorithms to control non-player characters. The vast majority, even those with enemies that appear intelligent, use basic techniques like search algorithms and simple behaviour trees.

In a gaming context, "AI" usually just refers to any algorithm controlling a character other than the player, regardless of what that algorithm actually is.

I see, in that case the complexity of the AI involves what the enemy is capable of like how a Chargin' Chuck AI would be more complex compared to that of a koopa troopa. Thinking about it out of all the classic Mario Platformers I would say Yoshi's Island enemies tend to be the most complex in behaviour, even some of the more simple ones like shy guys and goombas in that game. Speaking of which are any Yoshi's Island enemies going to be added in the future or is the main focus just the first three Super Mario Bros plus Super Mario World. Though to be honest I do expect to see a lot of Super Mario World stuff in this project.

Mario World 2's tile sets are extremely expansive and hard to use properly in the wrong engine. Source, I attempted to use YI2 tiles in Mario World. There are just so many variations and a lot of them have to be used in conjunction with other specific tiles, otherwise they don't look right.

[Westretroman]

What you're talking about is more usually referred to as "machine learning", and isn't usually what people mean when they say "AI" in the context of games. Very few games use actual learning algorithms to control non-player characters. The vast majority, even those with enemies that appear intelligent, use basic techniques like search algorithms and simple behaviour trees.

In a gaming context, "AI" usually just refers to any algorithm controlling a character other than the player, regardless of what that algorithm actually is.

I see, in that case the complexity of the AI involves what the enemy is capable of like how a Chargin' Chuck AI would be more complex compared to that of a koopa troopa. Thinking about it out of all the classic Mario Platformers I would say Yoshi's Island enemies tend to be the most complex in behaviour, even some of the more simple ones like shy guys and goombas in that game. Speaking of which are any Yoshi's Island enemies going to be added in the future or is the main focus just the first three Super Mario Bros plus Super Mario World. Though to be honest I do expect to see a lot of Super Mario World stuff in this project.

Mario World 2's tile sets are extremely expansive and hard to use properly in the wrong engine. Source, I attempted to use YI2 tiles in Mario World. There are just so many variations and a lot of them have to be used in conjunction with other specific tiles, otherwise they don't look right.

But he is mainly, mentioning the enemies from YI, some which appeared in Super Mario Advance 4's World-E. It would be nice to have something like a Bumpty, like 1.4 has.

[Teemster2]

I think we are thinking of two different AI's here because when I think of AI it's a complex program capable thinking near human level intelligence. That's something a Mario enemy doesn't need. Though I think your referring more to coding new enemy behaviour which is a lot more possible than you make it sound considering new enemies have already been added to the beta.

What you're talking about is more usually referred to as "machine learning", and isn't usually what people mean when they say "AI" in the context of games. Very few games use actual learning algorithms to control non-player characters. The vast majority, even those with enemies that appear intelligent, use basic techniques like search algorithms and simple behaviour trees.

In a gaming context, "AI" usually just refers to any algorithm controlling a character other than the player, regardless of what that algorithm actually is.

I see, in that case the complexity of the AI involves what the enemy is capable of like how a Chargin' Chuck AI would be more complex compared to that of a koopa troopa. Thinking about it out of all the classic Mario Platformers I would say Yoshi's Island enemies tend to be the most complex in behaviour, even some of the more simple ones like shy guys and goombas in that game. Speaking of which are any Yoshi's Island enemies going to be added in the future or is the main focus just the first three Super Mario Bros plus Super Mario World. Though to be honest I do expect to see a lot of Super Mario World stuff in this project.

Having NPC's from YI would be cool. I asked the question about using NPC's from the DKC universe ( I really wanted Zinger, Croctopus and blast barrels) and was told the main focus was on the first 3 Mario games and SMW at the moment but that was months ago that I received that answer. It takes awhile from my understanding of the dev team to create NPC's and there is still some pretty well known and common NPC's from the main games not even claimed yet. (Urchin, Porcupuffer, AngrySun, AmazingFlyingHammerBro, IceBro and BoomerangBro to name a few) I would love to see hundreds of new NPC's added but the fact is only a handfull of people can make new ones so we just have to be happy with whatever we get but for all of the NPC's we can get. We should see a lot more before the final release of the game but I don't see why we need to have a final release as there is always stuff that could be added to SMBX and it could be a never ending thing . How big can the game get before it is too big or gets to be to slow? I say make it as big as you can before you finally decide to make the final version or just keep adding stuff.

Thanks to Enjl for the response to my other question and to Yoshi for adding some Super Mario Land NPC's that I actually wanted. (Bombshell Koopa and Bunbun)