I have been seeing the progress of Sponge for well over an year now and whilst they have like the most strongest API out of any other server wrappers, it’s still not finished. So my question is, does anyone have any approximation on when it’s going to be stable and complete enough that it can be used in a major server? I am currently working on a server that has an average player base of 200 and I am really interested in switching to Sponge the same way I had been from day 1. Thanks!
1 Like
The biggest problem right now is that while the API is actually pretty stable now and has most of the features you would ever need, there are still a lot of gaping holes in the implementation. (Inventory i’m looking at you) Once those holes no longer exist, and most of the bugs are sorted out, it’s then just getting all the plugins you need. Some plugins have been or will be ported. Others might be reborn as completely new plugins that fulfil the same role. We can’t really tell you when it’s a good time to switch to Sponge; you just need to try it out yourself. Take SpongeForge or SpongeVanilla, dependending on what you need, and set up a test server. Throw plugins on it and see if you can’t set something up. Things will probably break along the way, but don’t be afraid to report bugs and stuff. You might eventually be able to get something to work.
7 Likes
Just to add onto this, the Sponge team need servers in order to find bugs. Having 1 person report a bug is good, but being able to get 2 people confirming the same issue is very good.
2 Likes
I guess I can soon start experimenting with Sponge again and see how much I can port things from the Spigot plugin. Thanks for your response!
Fair enough. We will need to soon start a server with like around 100-150 players in Sponge to see how it tackles the stress. If we find bugs, we will be sure to report it. I understand for each new project, stability is something that happens overtime.
1 Like
Minus some major features, and depending on your plugin selection some servers could make the switch now. Might be a while before some of the bigger name plugins are ported. As someone who just finished porting MoneyDrop, it’s a bit of a pain
If you don’t mind to share, can you tell me what part of porting was painful to you?
Sponge API and Bukkit/Spigot API are dramatically different, both in their construction, and in their philosophy. Often times, sections of plugins need to be completely rewritten to work on Sponge API. In some other cases, where plugins use NMS, the NMS code needs to be updated too, which is painful in and of itself.
Plugin development is hard. Plugin porting is almost just as hard, if not harder in some cases. It’s because a lot of plugins were developed specifically around the Bukkit/Spigot API, which isn’t bad, but some of them are so closely tied to the structure of that API that it makes it very hard to port it to the Sponge API, which has an entirely different “shape” if you will.
8 Likes
This will probably not even happen at all. For most servers, switching to Sponge isn’t a topic to discuss anymore. The SpigotMC community is not slowly dying, it’s growing faster than ever before.
SpigotMC has currently roughly twice as many members as in December 2015 (100,000, now > 190,000). It has never been growing as fast as it is now. The Sponge forums got about 500 new members since then.
These statistics are taken from MCStats.org:
Servers with the Bukkit API are altogether >90% without BungeeCord. To be honest, I have no idea what “BungeeCord” means - is it a Proxy instead of the actual servers it represents or just the Proxy while the servers are represented themselves? Anyway, SpongeForge and SpongeVanilla do not reach much more than 0.1% together. There is no way Sponge will be common enough to have enough plugins to compete to Bukkit in the forseeable future.
Also, over at SpigotMC, there are many members who consider Sponge inconvenient to work with it, mainly because of its complex design and some other issues like the “excessive” usage of Optionals. Besides that, there is not really more you can do with Sponge than with Bukkit. There are still the same limitations of server side modding (most servers don’t want client side modding and Forge) and some holes in the API can easily be filled with NMS (which is actually not considered as evil as stated).
1 Like
Yeah idc I’m done with bukkit. Also take them statistics with a grain of salt, and don’t see a complexity issue. Matter of fact I found Sponge way easier to work with. Sounds like a lot of people who didn’t know Java jumped into Bukkit and thought they’d learn that way to me. Last point, what’s so bad about Optionals? They eliminate null values. This is a good thing.
I’d also like to mention slightly off topic, that I got treated like crap on a regular basis in the Bukkit community and I know others suffered the same. Won’t find that here
8 Likes
Tbh I think the reason why so few people are using Sponge atm is taht Sponge doesn’t have a proper release yet. For 1.8.9 we only have Beta recommended builds and for 1.10.2 we even have only Beta latest builds.
I am currently setting up a Sponge server to upgrade my server to that when I’m done, but I’m only doing it because my server is in the buildup anyway and I don’t really have a community yet. If I already had a community on my server, I would also wait till we have at least a recommended beta build for 1.10.2.
Same problem applies to plugins as well. Most plugins are still in development and not all that stable.
So yea, I think when Sponge finally has some release builds and the plugins are farther in development as well more people will start using it.
1 Like
Of course, it’s all about personal preference and experiences. That’s the gist. That’s why you can’t just assume that Bukkit will die sooner or later just because Sponge has a better design. CanaryLib was also considered way better than Bukkit and didn’t become more popular…
Honestly it’s a combination of things, Sponge not being complete, developers and server owners who don’t like change, especially server owners, you know if it works why change it. And the fact that Sponge is years late to the game.
Probably a bit bias though because I never really cared for Bukkit. Everything always felt like a workaround when developing plugins and as I said earlier the community doesn’t create warm welcome to outsiders and really had no interest in improving the platform. Suggestions were treated with harsh scrutiny.
That being said the fact can’t be ignored that there’s a possibility that bukkit will always be the go to for most. Howecer there’s a very large group of mod pack server owners that are already relying on the success of this project. The statistics may show low percentage of cauldron users but considering the real size of the Minecraft community ignoring them is ignoring potentially millions and that’s something to be considered.
2 Likes
Considered by who? I’ve looked at Canary and didn’t think much of it. I’d like to read about what was different / better with Canary then Bukkit if you have sources.
No, it’s not about personal preferences.
The fact of the matter is this.
Bukkit/Spigot is already very well established. Sponge is not.
There are a few causes for this:
- Sponge is not actually done yet. There isn’t a release version of Sponge. It it still in Beta, and has been since the beginning of this year.
- There is virtually no plugin support for Sponge at the current moment. The reason is list entry 1 above. Also there hasn’t been enough time.
- Because of list entry 2, there is not any benefit to using Sponge over Spigot right now, other than native Forge Support, which also still has its hiccups.
At some point in the near future, Sponge will actually start looking complete, which means there will be much more plugin support, which means people will actually have a REASON to switch to Sponge. At that point we’ll start seeing a lot more sponge.
8 Likes
One of the principal motivations for the creation, and continuing evolution and development of SpongeAPI and it’s implementations was to create an effective replacement for Cauldron, more so than Craftbukkit. The fact that we have the former lead dev of Cauldron as one of the lead developers of Sponge speaks volumes, and at least gives us the confidence that we’re headed in a sensible direction. That we happen to compete with the surviving Bukkit implementations via SpongeVanilla is a side effect of the process, in some ways.
Sponge could probably do more to boost it’s profile, but that would come at a cost. We really need to have a decent stable, and complete implementation available for the latest MC in order to compete properly, and that is still some way off. Everyone using Sponge implementations at present should be well aware that the whole project is still in-development, and relies on the feedback of users and devs to get where it wants to go. That includes considering the perceptions and quibbles from across the full spectrum of Minecraft communities that could stand to benefit from the project. So it’s good to see a fair comparison being made between these different server platforms, and where they might learn from one another.
10 Likes
Sorry to drag up an old topic but the fact that this topic is old is an example of the problems being faced by many server owners trying to implement Sponge.
I am one of those server owners and whilst I have no problem with change one of the major problems I am experiencing is a severe lack of documentation, not about Sponge specifically but plugin documentation.
There are currently around 100 or so plugins for Sponge, and whilst the main ones are there, they are very lacking in basic features compared to the older bukkit variants.
In addition to this is seems alot of those plugins are discontinued or have been abandoned.
I must confess, I am concerned Sponge may never really get off the ground and that another 1.10+ version of cauldron will be required for us to even get close to what we had available in 1.7.10.
It feels like a serious step backwards in terms of forge servers.
This is not a criticism of Sponge or the Devs, it’s obvious to anyone how much work you guys have put in, I’m just not sure it’s enough when we are used to all the candy that bukkit has given us over the years, bukkit allowed us to customise our servers and make us unique compared to other servers out there, with Sponge and because of the lack of plugins, most 1.10 forge based servers using Sponge are almost identical and offer nothing special compared to the next server out there. This is one of the largest issues regarding Sponge right now I think.
I understand it’s likely impossible to create a bridge between Sponge and bukkit but I am having problems seeing any other solution.
1 Like
Not Quite Impossible. Not perfect, by any stretch of the imagination, but some of it is doable.
Feel free to prod developers about absent documentation, that’s always been a common shortcoming with plugins in general. We do our best to document our own stuff, and even that has a number of significant gaps. Assistance is always appreciated.
Resurrecting Cauldron for the purposes of continuing to mash together the Bukkit and Forge APIs is a process fraught with frustration, and that is one of the reasons Sponge even exists. They just plain don’t seem to like each other. Whilst it may be possible to make a bukkit plugin hobble along on a Sponge server using some implementation of Pore, it’s still desirable to port it to Sponge so that it can run more efficiently in the native implementation and also make use of the interoperability that SpongeForge provides to mods in general. Obviously YMMV.
It seems that a major inhibition of porting to Sponge for larger servers is the need to rewrite a large body of custom code, which is how they stand out from the pack. It will take time for a larger body of public plugins to develop, once a few parts of the implementation have stabilised. Hopefully the imminent launch of Ore will help kickstart that process. I guess it very much depends on your needs and wants in a server.
2 Likes
Its not impossible at all. Devco has their servers connected rather well. We have 9 Sponge servers running and still 21 (?) Spigot servers going. The chat is global (This is a private plugin so don’t ask), our store is setup in a way that it works for both Spigot and Sponge… there really isn’t much that isn’t integrated actually.
You basically just need a good development team, or hit the plugin request forums. There is actually a lot of resources available, some people just don’t want to take the time to investigate what is available.
4 Likes
But new servers use Spigot by far more often than Sponge at the moment, too. It’s not simply too much work to switch.
For most people except those who need to run mods, Bukkit suits their needs and they see no reason to try Sponge. It’s not like you can’t do things with SpongeVanilla that you can’t do with Spigot.