Encourage Modding

[Sergeant Base]

Basically Minecraft blocks CANNOT MOVE...sure there are trains & pistons and some blocks are gravity-sensitive, but that is completele different from at least 10 10000+ block-models moving at 200 km/hthrough the same sector...

 

[Sergeant Base]

Noone\'s flaming...we\'re just discussing our arguments o.0

 

[Archer6621]

I know you Maple, I was actually defending you, none of my negativity was directed at you.

Again, the lock you mean is formally called \"obfuscation\", that\'s how it is named.

 

[Archer6621]

Still not enough explanation, how is it completely different?

 

[Sergeant Base]

Just imagine...if you build a house in minecraft & the same house in starmade as a ship (block-type doesn\'t matter for that). You can build the same house about 200 times in minecraft without real performance problems....all those houses are on a fixed position that\'s all the server needs to know. Now in starmade if just one of these 200 houses collides with another one it would initiate a giant chain-reaction and the server has to calculate every houses collision with others. Do you know how hard it is to calculate the collision-proxy of a house? Just imagine we build ships that are maybe 100 times bigger and every block every hull part needs to have collision. The physics behind all this is so much more difficult than the pretty simple basic physics minecraft has. A block in minecraft is A: static (means: not movable) or B: affected by gravity (means it\'s falling down...JUST down).

The whole system behind both games has to be completely different I don\'t think that even ONE LINE in SM\'s SourceCode can be found 1:1 in Minecraft. Sure they look very equal but that\'s probably just for the players. I don\'t know because I haven\'t seen the code yet, but I don\'t see anything in Starmade that acts like something in Minecraft.

 

[matheus.de.alcantara]

\"public final class AC\"

Okay, that\'s not what you mean, i know, I know.

@ONTOPIC : jjaquinta, if you need a help with SMCP, PM me, I can code some decompilers.

 

[kupu]

I for one am personally looking forward to seeing a Project subsection or player made content on the new site.

Coding issues aside, simply having a section availible could encourage a few more potential artists and modders.

 

[Archer6621]

What you are saying (the entire first paragraph) is mostly not related to modding, and the last bit of it only makes one little aspect of the whole game difficult to mod (the physics, but who\'s going to mod that anyway?).

Does this stop people from modding the rest of the game? No, look at jjaquinta\'s post above, the only difficulty is the obfuscation, nothing else is impossible, even if it may be hard, you need to understand that.




I don\'t think (etc etc)


Thinking is not enough, if thinking is your main argument then I call it invalid. We need to KNOW if that\'s the case, by first de-obfuscating the code and then trying to modify it.

Negative speculation and pessimistic views won\'t bring anybody anywhere...

 

[Comr4de]

go back to /b/ you moron

 

[ltmauve]

You know what would be nice? About six hundred little improvements to the client side of things. Ability to sort the navigation list. Far better weapon interface. But if Schema does decide to fix these little problems, it would probably delay the game progress by a few months. But if modders got involved, they could fix the little problems that bug us but aren\'t worth Schema\'s time.

 

[Archer6621]

Exactly, modders can try out stuff, and if it works well, schema could consider it for the main game.

 

[jjaquinta]

Technically, making improvements to the program itself is not modding, it\'s more like opensource. Modding is usually extending a program through well defined points. I wouldn\'t normally quibble over words, but there is a distinction in how to implement either.

For mods, Schema has to build in extension points into the code. (That\'s the right way. The wrong way is what Minecraft did and have someone hack them in.) For improvements, the code needs to be open.

Personally, I don\'t think making the code open source is the right direction for Schema. It would have a good chance of killing any chance of it going commercial. If it doesn\'t go commercial, it will always be a pet project with a small following.

There are other choices. The model I would recommend is the one Linden Labs used with their Second Life client. They made the source code open to read, and would let other people create their own derivitives of it, as long as they signed up for a license key. (Their servers wouldn\'t respond to unlicensed clients.) More pertinently to your suggestions, they also let people submit suggested changes to the code base. It\'s not true open source where anyone can post changes, but controlled.

This approach would give Schema a nice \"force multiplier\" to his development. This project has grown past what one developer can support. These \"600 little improvements\" are what the community needs, but he can\'t do that sort of thing on his own. But, if he spent one less hour a day developing and instead used that hour to approve/reject community submissions, the project would advance a lot faster.

 

[BraveSteelDragon]

Schema could do a pseudo-code documentation, and then open a \"code repository\" here on the site.

This way people could still suggest new features throught more practical means (trying to pseudo-code it would give people the right feeling of why certain features are possbiles and other are not), while still allowing community feedback on those given suggestions.

 

[Archer6621]

Couldn\'t he simply make parts of the code open-source, that people are likely to modify/improve? Such as the whole amc balancing thing or how those specific blocks are defined and assigned to certain IDs, that would be quite the improvement already, not sure if it still would kill the chance for it to go commercial though, I don\'t know much about licensing.

 

[BraveSteelDragon]

Making something like that would mean using a license which would be more awkward than useful....

 

[jjaquinta]

They key is to come up with something that gives Schema more benefit than it costs in time. Re-writing everything in pseudo-code would suck down time and give dubious results.

Making only part of the code available in open-source could be done one of two ways. The constraint is that if modders can\'t re-compile and run the code they cannot test their submissions. The quality would go way down and vetting them would take more of Schema\'s time then the benefit given. Two approaches:

1) Make only a whole unit available. For example, Linden Labs made the client code available, but not the server code. For StarMade, it might be more usefiul to do it the other way around. In a well designed system, most mods will be on the server.

2) Select portions could be made open-source, but the rest would have to be deliverable as a library/jar file. This is do-able, but would require some up-front work from Schema. The code would have to be separated and the obfustication approach changed. (Clear code can\'t really call into obfusticated code.) Given Schema\'s retticense in responding to anything, I don\'t think he\'ll pick something that requires any degree of work or change.

 

[Winterhome]

One word: Modloader.

The entire purpose of that mod was to force the engine to accept mod packages without the need for further modification of base classes.

The problem is the obfuscation. If Schema used a separate algorithm for content from the one used for the engine itself then we coild have a solution to the risk of code theft, at least for the engine itself. Not sure if that would be possible though...