MCForge is awesome...

[FlowerChild]

I sincerely hope Notch keeps adding lots and lots of aesthetic blocks. If he manages to use up enough, he'll effectively break modding and will have to come up with a block ID solution.

So keep bugging him about more aesthetic blocks :)

Rofl! That's a rather unique approach you're suggesting there :)

[Zhil]

Rofl! That's a rather unique approach you're suggesting there :)

But it's completely true. Notch using up another 50 IDs and he'll have the whole minecraft community up his ass about more IDs :)

As it currently stands, block IDs is the biggest single hurdle that can't be fixed by even a unified API like Forge.

I still think there might be a solution though, I have a couple of ideas floating around for making more block IDs available. Currently I'm experimenting with making a new block class that can contain all the blocks for a single mod in one block ID and I'm also playing around with a method that can salt the block ID with a mod ID, so each mod can use 256 separate block IDs. For the second idea I need a way to resolve save files, which I have ready too. Basically, there'd an extra set of saves for each mod. When the game loads a "mod block", it looks it up in the alternate mod save file. If I ever get an easily workable system, I'll release it as part of Forge or maybe as an extra API on top of Forge.

[DaveYanakov]

It's just crazy enough to be completely insane. Therefore it is an excellent plan! Quickly! We need a list of at least 85 blocks to add!

Rofl! That's a rather unique approach you're suggesting there :)

But it's completely true. Notch using up another 50 IDs and he'll have the whole minecraft community up his ass about more IDs :)

As it currently stands, block IDs is the biggest single hurdle that can't be fixed by even a unified API like Forge.

I still think there might be a solution though, I have a couple of ideas floating around for making more block IDs available. Currently I'm experimenting with making a new block class that can contain all the blocks for a single mod in one block ID and I'm also playing around with a method that can salt the block ID with a mod ID, so each mod can use 256 separate block IDs. For the second idea I need a way to resolve save files, which I have ready too. Basically, there'd an extra set of saves for each mod. When the game loads a "mod block", it looks it up in the alternate mod save file. If I ever get an easily workable system, I'll release it as part of Forge or maybe as an extra API on top of Forge.

Hang on, if it ends up being possible to salt mod IDs, does that mean a single mod would be able to use several mod IDs, thereby allowing an author to use several hundred block IDs?

[Fracture]

It's just crazy enough to be completely insane. Therefore it is an excellent plan! Quickly! We need a list of at least 85 blocks to add!

Rofl! That's a rather unique approach you're suggesting there :)

But it's completely true. Notch using up another 50 IDs and he'll have the whole minecraft community up his ass about more IDs :)

As it currently stands, block IDs is the biggest single hurdle that can't be fixed by even a unified API like Forge.

I still think there might be a solution though, I have a couple of ideas floating around for making more block IDs available. Currently I'm experimenting with making a new block class that can contain all the blocks for a single mod in one block ID and I'm also playing around with a method that can salt the block ID with a mod ID, so each mod can use 256 separate block IDs. For the second idea I need a way to resolve save files, which I have ready too. Basically, there'd an extra set of saves for each mod. When the game loads a "mod block", it looks it up in the alternate mod save file. If I ever get an easily workable system, I'll release it as part of Forge or maybe as an extra API on top of Forge.

Hang on, if it ends up being possible to salt mod IDs, does that mean a single mod would be able to use several mod IDs, thereby allowing an author to use several hundred block IDs?

LIMITLESS BLOCKS! AHAHAHA!

But in srsness, we could do so much with that.

[Zhil]

Hang on, if it ends up being possible to salt mod IDs, does that mean a single mod would be able to use several mod IDs, thereby allowing an author to use several hundred block IDs?

Yes, but depending on how this would work (still looking into it and it does SEEM possible), it might have a restriction of, say 16, meaning you'd only be able to run 16 mods in this way. If a single mod then uses multiple mod ID's, you'd still be at square one.

That being said, I'm leaning towards fully exploring the single block ID mod, since that would be easier to implement and more versatile to work with.

[walker_boh_65]

Hang on, if it ends up being possible to salt mod IDs, does that mean a single mod would be able to use several mod IDs, thereby allowing an author to use several hundred block IDs?

Yes, but depending on how this would work (still looking into it and it does SEEM possible), it might have a restriction of, say 16, meaning you'd only be able to run 16 mods in this way. If a single mod then uses multiple mod ID's, you'd still be at square one.

That being said, I'm leaning towards fully exploring the single block ID mod, since that would be easier to implement and more versatile to work with.

You might want to see how red power does it, the mod maker fits id say at least 70 blocks into one id with damage values up into the 1000s

[Zhil]

You might want to see how red power does it, the mod maker fits id say at least 70 blocks into one id with damage values up into the 1000s

I regularly chat with Eloraam, so yeah, I know about that trick. It solves some ID problems, but it's limited in use. BTW for example wouldn't benefit too much from it.

[Deepsniper]

I truely doubt that he will use up 256 blocks...

[FlowerChild]

You might want to see how red power does it, the mod maker fits id say at least 70 blocks into one id with damage values up into the 1000s

I regularly chat with Eloraam, so yeah, I know about that trick. It solves some ID problems, but it's limited in use. BTW for example wouldn't benefit too much from it.

It's not much of a trick. It's just assigning tile entities to pretty much every block in the mod, which is not something I'm willing to do.

[DaveYanakov]

Hang on, if it ends up being possible to salt mod IDs, does that mean a single mod would be able to use several mod IDs, thereby allowing an author to use several hundred block IDs?

Yes, but depending on how this would work (still looking into it and it does SEEM possible), it might have a restriction of, say 16, meaning you'd only be able to run 16 mods in this way. If a single mod then uses multiple mod ID's, you'd still be at square one.

That being said, I'm leaning towards fully exploring the single block ID mod, since that would be easier to implement and more versatile to work with.

If a mod is going to use up more than 512 block IDs, how many mods do you actually need?

[MagusUnion]

I sincerely hope Notch keeps adding lots and lots of aesthetic blocks. If he manages to use up enough, he'll effectively break modding and will have to come up with a block ID solution.

So keep bugging him about more aesthetic blocks :)

Or... he'll accomplish the goal of disallowing modding in the first place by blocking off all of the empty space away from modders... >.>

(seriously... do you honestly think he likes others tinkering with his creation?)

[Battosay]

(seriously... do you honestly think he likes others tinkering with his creation?)

Yes, he does. A lot actually.

Edit :
After checking what tinkering means
"To manipulate unskillfully or experimentally."
I'd say you're not using the correct word.

[Katalliaan]

Magus, in April there was a post on Notch's blog that stated that Mojang wanted to set up a sytem that allows mod developers to sign up for full access to the code. I highly doubt that there's any plans for that to change, especially when you consider that modding is what keeps players interested in older games. If such a system were to be implemented, I'm sure that someone would put out a mod to merge the IDs of blocks to free up IDs.

[MagusUnion]

Magus, in April there was a post on Notch's blog that stated that Mojang wanted to set up a sytem that allows mod developers to sign up for full access to the code. I highly doubt that there's any plans for that to change, especially when you consider that modding is what keeps players interested in older games. If such a system were to be implemented, I'm sure that someone would put out a mod to merge the IDs of blocks to free up IDs.

Yeah, so where is it? Did he change his mind? Decide it was too hard to implement? Or did he not like the idea of not profiting from mod development as he originally intended? (Since that was the initial idea that he had before everyone chewed him out over Twitter like they did...) I honestly do think that if he was going to bring about this system, he would have done it instead of buying off Pistons and placing it into vMC, (because this idea was in the works around 1.5~1.6 ish)... Plus, from what little he's actually said about the Mod system he wanted to create, it wouldn't have the same space saving functionality as ShockAPI or Forge, so why use it when you run into the situation of not loading each mod you would like to have (again)?

Edit :
After checking what tinkering means
"To manipulate unskillfully or experimentally."
I'd say you're not using the correct word.

Not every modder is as skillful as Flowerchild... There are quite a few badly created mods on the MC forums and planet-minecraft if you look... so yes, I did use it correctly...

[Tekei]

...

As Batto said: Notch does not have anything against modders since it is what's keeping minecraft alive for a lot of users. It wouldn't be good business to counteract them and Mojang knows this. Saying anything else is just... false.

[Katalliaan]

Yeah, so where is it? Did he change his mind? Decide it was too hard to implement?

I can't say, as there's been nothing from Mojang regarding it since June 2011 (when Jens made a comment on a Reddit thread about the official implementation of pistions saying that they plan to stop obfuscating the code when they release their mod system).

Or did he not like the idea of not profiting from mod development as he originally intended? (Since that was the initial idea that he had before everyone chewed him out over Twitter like they did...) I honestly do think that if he was going to bring about this system, he would have done it instead of buying off Pistons and placing it into vMC, (because this idea was in the works around 1.5~1.6 ish)...

They would still "profit from mod development" - if a mod or total conversion catches their eye, they would buy or license it and sell it themselves, or integrate it into vMC.

Plus, from what little he's actually said about the Mod system he wanted to create, it wouldn't have the same space saving functionality as ShockAPI or Forge, so why use it when you run into the situation of not loading each mod you would like to have (again)?

It's not an API, though - it's just access to the code, which we sort of already have (once someone deobfuscates the code each patch). I'd imagine that we won't be rid of things like ShockAPI or Forge, just have our mods updated for the new patches faster.

[FlowerChild]

It's not an API, though - it's just access to the code, which we sort of already have (once someone deobfuscates the code each patch). I'd imagine that we won't be rid of things like ShockAPI or Forge, just have our mods updated for the new patches faster.

Right. The "mod API" as described by Notch so far would basically only serve to take MCP out of the equation for mod-developers as we would no longer need it to deobfuscate the code. It *may* also provide us with more meaningful variable names and code comments, depending on how well structured the Mojang source is.

That aside, it will do very little to change the modding scene, and pretty much nothing from a user perspective, other than perhaps shortening the delay between a new version coming out and mods getting updated as we won't have to wait for MCP to update to move forward.

It's really not an API at all as it has been described.

[MagusUnion]

...

As Batto said: Notch does not have anything against modders since it is what's keeping minecraft alive for a lot of users. It wouldn't be good business to counteract them and Mojang knows this. Saying anything else is just... false.

I'll believe it when I see it...

plus: "We retain the right to use your mod idea and implement it ourselves
in Minecraft." + "it’s highly likely we will want to license it." = a loss of ownership towards said mod due to the fact that, legally, you have to withdraw your ownership of said mod to Mojang for it to be ever implemented in the game...

That doesn't sit well with me, and hasn't since April when I first saw said post. These mods are, in essence, stand-alone games expansions based around the Minecraft engine, and personally I find it unfair to modders to lose their ownership over their hard work with this 'deal'. I get that this is what many major companies do this anyway with mod material, but I kinda would like to think Notch would think more about the work people have to go thru for modding, rather than worry about how said mods will 'break' his game in terms of both gameplay and liability...


But perhaps that's a misplaced standard on my end since I was one of the very first supporters back in alpha version. Absolute power corrupts absolutely, eh?

[Zhil]

It's not much of a trick. It's just assigning tile entities to pretty much every block in the mod, which is not something I'm willing to do.

Yeah, exactly.

I had a talk with Eloraam and Seronis last night and I think I may have worked out a solution. Doing the math, once this system is implemented, it allows for 110 mods, each having their own 128 block IDs and 128 item IDs.

The way it works, it could sit in the Forge API without altering save files, decreasing performance or anything. Will keep you guys posted.

[Damion Rayne]

I have installed ShaRose's IDResolver Mod that has both a Forge and Non-forge Compatible version. It requires Risu's ModLoader and GUIApi and works perfectly, it is even keeping track of non-forge mod ID's for Sprites, Blocks, and Items and if you want will automatically asign ID's to things so that you have zero conflicts. it's bloody brilliant, and it works with Minecraft Extended as well. Though MCEx doesn't work with Forge mods.

So this is a good mod to have until whatever it is that Gil and his friends are working on comes out.

Mod is located here: http://www.minecraftforum.net/topic/366 ... r-updated/

[Damion Rayne]

So I got a downgrade to 1.7.3 and installed EE, BC, IC2, and BTW along with Wireless Redstone and ShaRoses's ID Resolver.

Here's a Screenshot of ID Resolver listing all the mods I have running,



Here's a screenshot of the Beta 1.7.3 start page, adn you'll notice in the top right a listing of numbers showing me the following,

64 Block ID's Left
31680 Item ID's Left
124 Sprites Left



Now ShaRoses ID resolver worked in 1.8.1 as well, when installed properly it will auto-detect Item and Block ID conflicts and allow you to set a mod to a certain priority, and automatically assign ID's (first avail from the top or bottom of the list) to each mod in turn.

Proof that with a little bit of work, Forge is effing awesome and allows large multi-mod worlds.

-DR

[Zhil]

Just a heads up that we got the block ID solution mostly working, so you will probably get unlimited block IDs with Forge soon. Thanks go out to Eloraam and especially Chickenbones, who took my idea, ran with it and implemented it after we talked about it :)

He added a few ingenious things. For example, if you remove a mod, the modded blocks should turn into air.

[Damion Rayne]

Just a heads up that we got the block ID solution mostly working, so you will probably get unlimited block IDs with Forge soon. Thanks go out to Eloraam and especially Chickenbones, who took my idea, ran with it and implemented it after we talked about it :)

He added a few ingenious things. For example, if you remove a mod, the modded blocks should turn into air.

Care to explain how this is even possible?

[Zhil]

Care to explain how this is even possible?

It's a bit complicated, but basically MC stores block IDs in a 32k array, it's just a question of save files storing them as bytes. I proposed a system where all mod blocks were a single ID and bait/switching them for the correct one on chunkload. Eloraam and I fleshed it out further, but she had some reservations. Today, ChickenBones came in and made a few tweaks to the idea and implemented it. It's all working now. So, you have now got 32k block IDs.

The specifics are a bit more complex, but that's more or less the gist of it. Vanilla can still read the save files, the memory overhead is minimal and mod makers will probably need to use the correct interfaces, so it's not going to work for non-forge mods like MCE.

[MagusUnion]

Care to explain how this is even possible?

It's a bit complicated, but basically MC stores block IDs in a 32k array, it's just a question of save files storing them as bytes. I proposed a system where all mod blocks were a single ID and bait/switching them for the correct one on chunkload. Eloraam and I fleshed it out further, but she had some reservations. Today, ChickenBones came in and made a few tweaks to the idea and implemented it. It's all working now. So, you have now got 32k block IDs.

The specifics are a bit more complex, but that's more or less the gist of it. Vanilla can still read the save files, the memory overhead is minimal and mod makers will probably need to use the correct interfaces, so it's not going to work for non-forge mods like MCE.

Wait, so that means that I can't install planes + forge under this new system?