I meant to reply to this earlier but it slipped my mind. Whoops.
Anyway, from what I understand Spigot is going to phase out all bukkit code and replace it with their own. Once they do that I can’t see any reason why they wouldn’t be able to publish the sources. Correct me if I’m wrong but, I’m also pretty sure they haven’t/won’t distribute complete game code anyway so nothing would be stopping them.
And as @modwizcode has already said, Forge or Sponge doesn’t distribute complete game code anyway so the idea that you’d have to consider “any server mod publishing sources to be very sketchy” isn’t true. This also true for granite since it states on their site that it “doesn’t contain or build against vanilla code.”
“Reflection is super horrendously slow” is a myth that hasn’t been completely valid for at least 3 years. It appears that Granite is caching classes, fields, and methods based on name, so the overhead is already greatly reduced. See the recent posts on this thread: http://stackoverflow.com/a/435571
The only real performance drawback I can think of is inability to optimize vanilla stuff like Spigot, which could be worked around with ASM or similar bytecode manipulation.
In addition, I don’t see why Granite would be “at least somewhat hard to update”. It seems like it would be slightly simpler than CB due to not actually modifying the vanilla code, and the developers are supposedly starting to automate their mapping methods.
Personally, I feel that it’s the most viable, legal option ATM. It goes without saying that this is blasphemy on the Sponge forums, but from writing a simple plugin and appraising functionality/ease of use/future update potential, Granite definitely takes the cake.
I don’t think it’s blasphemy at all. Granite appears to be our most serious competition if Husk doesn’t turn out to exist.
I think we should help them design a good API, I actually would rather team up with and have them work on a native implementation of the SpongeAPI that doesn’t rely on forge.
All I can say is for someone that doesn’t have a product out, you sure do put sponge on a pedestal, Knocking down everything else. Not that I’m against Sponge, but what have you guys got to show for it as of now? I’ve said it before and I’ll say it again Rainbow is the only working alternative at this time for a 1.8 server that i find has less bugs then the others mentioned in the OP.
I haven’t knocked anyone down and neither has @modwizcode. Yes some people are shitting on it for no reason other than blind hate but if you look at my [other][1] [posts][2] and [various][3] examples you’ll find that the only thing that the legitimate arguments are doing is stating what they feel are pros and cons about other products. Nothing more. Heck even @Inscrutable has [made some good points][4] in the previous thread.
Just because something is the only thing available doesn’t mean it’s great. And just because Sponge isn’t out yet doesn’t mean we can’t discuss the pros and cons of other servers out there. Heck I may even start using another server if one comes out with the plugins I need and looks like it has a decent future.
Also as for the “You have nothing out yet so you can’t discuss the cons of a server that is out” argument I have to say that’s simply not true. As I already stated there are pros and cons to both products and you could easily compare the two on certain levels like Dev reputation, community, api features both planned and already shown, etc. Plus, yes Sponge isnt available to download and use yet but any dev can look at what they have done and pushed to the github and see if they would want to work with it in the future. The core part of a server platforms life are the plugin devs and depending on how many are willing to make plugins for the product and which well known ones get official ports is very important.
It’s not a race to see who get’s theirs out first it’s what pros and cons each one has more than just having a working product. Again, just because a restaurant is the only place to get food in town doesn’t mean it’s particularly good for you.
I agree with you to an extent, but how is possible to say Sponge is better than (anything) because the people working on it? The same developers will be working on other platforms so this should not matter. What I’m stating is that Sponge is being being put on a pedestal, with no proof of what’s being done with it. SO yeah you have some good developers onboard, but nothing to show for it. Pros and Cons should be compared for what you have going, not what might be.
You seem to have a few misconceptions so let me clear them up:
I have Never said Sponge is better than anything and most of the valid people in this and the other thread haven’t either. If I have alluded to that then that wasn’t my intention. The point of the pros and cons is to talk about the pros and cons not compare them. And if you do compare, you can compare on the levels as I have already said. I may support sponge and hope for the best for it’s future but it isnt even out yet. Anything could happen. Heck Granite may be the best one or Rainbow a little bit down the line.
As far as I know the plugin devs may or may not make plugins because the amount of time to port to each platform isnt worth it. Even if many do, not every one will take the time to port and what matters is which ones and how many are developed for what platform first
There is plenty of proof of what’s going on with it. From the Dev irc channel to the announcements to the Github. It’s not like a secret project only the head devs are able to access. Heck they even opened up applications for other devs to work on the api and you can ask any of them.
As I said above, there is work to be shown for it.
If pros and cons are to be compared they should be compared to everything. Yes, you shouldn’t say something like “Sponge will have better performance than Rainbow” because that would be a guess. But you could and should compare things like the future plans for each project or the way each of the api’s are shaping up. But again, the point isnt to compare to sponge necessarily but to give pros and cons on the platforms stand alone. Of course some people will make the comparison but the valid ones will compare to things that are actually, well, valid at the current time. Like some of the api problems some people have stated earlier.
I never said you , i was referring to the original poster.
In essence the brief description of the OP is a personal opinion regarding the server platforms.
I’ve tried Granite and several other platforms, none of them except for Spigot , can offer what i currently need as Rainbow does. Yes it’s not perfect, but i can guarantee no platform will be,
including Sponge.
Anyways my intent is not to get into a tit for tat with anyone, were all in this for the same thing ( the love of the game) I wish Sponge the best of luck., but in the meantime I’m happy with my Rainbow.
Ah, since you replied to me I assumed when you said
that you were referring to me. I’m sorry for the confusion but you should make it a little clearer who you’re talking about next time.
And alright. That’s your valid opinion and the op has theirs. Yes some of the comments may have been unnecessary but overall imo it was a reasonable pros and cons list.
Also what other platforms have you tried? I’d like to know what other ones are out there besides granite and rainbow.
Why we’re so sure that Sponge will succeed?
Well, we are actively developing Sponge, not only code-wise; thing is, the other projects only have small communities.
A big project with no community = can’t survive
A small project with a big community = can survive
Thank you ModWizCode for your concerns about Granite. We are aware that reflection will slow the server down slightly, but we have just come across something that could change the amount of reflection that we have done so far. Also I appreciate that people think that updating our project will take longer. In my opinion I don’t think that we will. I spent 2 nights (around 12 hours) updating to 1.8.1 (purely me deobfuscateing, comparing methods, classes and fields to 1.8.1) which will be release when Minecraft 1.8.1 is out. The only thing we have to do I update our mappings file and that’s it, Granite does the rest. As someone said ‘Finally, a use for Granite’ (This has become somewhat of our motto).
To Admin/Moderators:
Please do not take this as advertisement, I am purely expressing views on the projects current state.
Thanks for your response! I enjoy seeing participation from the projections I mentioned. Just a helpful hint modwizcode is the correct capitalization from my username. Not really a big deal at all.
I was wondering how hard it would be for updating Granite in an update like the one from 1.7 to 1.8 was, where large changes happened in the way it was written. The struggle I see, is that the reflection is a little hard to read as a developer, and functionality might be hard to understand because of the way it’s written.
Um I’m a bit unsure really, most things now im looking at the NBT route of things which will cut a load of reflection. As for the implementation of reflection, we have a mappings file like MCP has to work out what method/class/field does. We then load that up in to a Map. if we want to reflect something all we do is go to the Mappings class and do Mappings.invoke(“readableName”, Args); this will then pass through to the mappings and will be worked out but the Mappings class what needs to be invoked. It will then pass the Args through to the method that has been selected.
As for updating from 1.7 to 1.8, most methods were kept so there wouldn’t have been a lot to do. If a new method was added then we you just reflect it and add it to the API (Easy Peasy). And as for it nobe being clear and readable I am currently going through and adding javadoc to everything to make it clear to everyone.