Goodbye Craftbukkit, Hello Sponge?

Discussion in 'Bukkit Discussion' started by Gr0up115, Sep 8, 2014.

Thread Status:
Not open for further replies.
  1. Offline

    DoingItWell

    What's your ETA for a 1.8 build that supports the new blocks/entities?

    I'm also trying to "read the tea leaves" to figure out where things are headed and where to take my 1.8. server solution. I'll continue working on it as I enjoy it but would be nice to know timelines for things (like it's not clear to me when Spigot will support 1.8 similarly).

    I agree with you on tangible things over concepts. I offered in #spongedev to partner with someone to enable a Sponge-API plugin with my 1.8 server build (i.e. minimalistic at first but then expand the features with each pass). That way they'd have something you "can use now" as opposed to waiting for the full thing. The discussion didn't go anywhere though.

    So I'll just continue doing my own thing. At worst, the community just has another option to choose from even if on the fringe side of the house.
     
  2. Offline

    InsomniaCraft

    Want a timeline? 6 months to a year before Bukkit or Sponge is released for 1.8.
     
    lDucks likes this.
  3. Offline

    DoingItWell

    Well that is surely not true. It wouldn't take me that long working alone to add Bukkit support to a 1.8 customized server. So surely whatever group behind Spigot can get theirs updated. I was thinking at most a few weeks but maybe I'm being optimistic?

    I just wouldn't want to bake Bukkit support directly into the customized server .jar to avoid any copyright/legal entanglements (whether they have merit or not).

    I wonder if anyone has some code already available that loads Bukkit plugins that could be turned into a plugin for my server (i.e. a plugin that loads plugins). Maybe the Pore project that's doing something similar for Sponge. If so, then I'd like to partner with them to enable their plugin to load Bukkit plugins for my server mod. That would be really cool, someone of my simplistic Bukkit plugins (like IP Tracker) could be used today.

    Knowing timing of other projects and where things are headed would be great though for making wisest investment of time.
     
  4. Offline

    mossyblog

    The timelines heavily depend on projects like MCP. In the end deobsfucating the Minecraft server jar is not an easy to do from scratch but if you have projects like MCP you can determine where things are and how they line up in terms of changes. Now given 1.8 changed some of the code around in small parts whilst also adding new materials/block/entity types its a game of "pattern search" to figure out whats new, what's old and what's re-usable in terms of signatures from the previous version.

    MCP has got about 77% code coverage done for 1.8 even without thinking about what's new vs old given the previous version(s) signatures haven't changed..so thats a good sign its not to radical in terms of change(s). Yet all it takes is 1% to be the difference here and it can still hold back all releases in the modding layers.

    The problem here is when Minecraft gets a new release its the modding first response layer (deobsfuction teams) have to then pick over the bones to figure out where things that are known vs unknown exsit. When they get through that they push out a release and then the developer API layers get their respective updates which then moves onto the Plugin Devs which then gets the gamers back to hitting tree's and watching squids attack them :)

    I suspect this is probably what's contributed to the Bukkit Team's fatigue levels as if Mojang are constantly spitting out new server code and not giving free mappings or involving the team(s) more openly in that re-discovery process it just adds an unnecessary layer of friction to the issue. On the flip-side though i can understand why they'd probably be reluctant to do this given it would directly impact their IP ... but that's if you take the existing maturity levels as is from within Mojang. If you were to assume that these teams wanted to engage the modding community more direct then creating a specific SDK with EULA's attached would solve probably this problem, whereby you treat the vanilla server as a development platform (SDK) much like you do with Operating Systems. There's certain things you can do within the confines of the SDK but in the end its still their IP and they reduce the whole DMCA thing down to folks who break the EULA and go to far in terms of IP infringement(s).

    At the moment the whole Mojang vs Modding thing i put down to inexperience, in that these guys are most likely taking the talent pool they have and retasking them on growth of Mojang beyond PC and if they aren't they're probably spending the bulk of their time optimising the code base to get either more performance or get things ready for a 1.9+ release around a modding API.. but ultimately if they have a plan its not being shared or evangelised which in turn is where the inexperience plays out, as this tells me "Your community modding relations is not doing their job".

    Google, Microsoft, Adobe, Unity3D etc all have recognised the merit of putting "evangelist/advocates" on the payroll who's job is to broker the two worlds together to help keep a vibrate and uniform community in place. I'm not sure they have this and that's why we're probably in this negativity stall or miscommunication that's taking place in the Bukkit community today. It's not that people aren't fans of Minecraft its just people recognise they want MORE from the product, they want more creatures, more protection for their in-game assets, more mini-games and so on. Stating "We don't have enough bandwidth to handle the workload" is also probably a meh, not buying that response from me. If you can generate $100million+ per year in sales you can afford $700k a year to staff up in terms of "modding api" devs and more to the point if an existing cluster of community members are already skilled (proven to do so) and are doing this anyway, having them re-tasked to help you for the greater good - all win.

    So short answer .. 1.8 first needs the folks at the lowest end of the modding totem pole to exercise their specialised skills / talent before all others above them can prosper.
     
    DHLF likes this.
  5. Offline

    DoingItWell

    Interesting comments, thanks :)

    Btw -- I was experimenting just now and found I could dynamically load a Bukkit plugin pretty easily. This is awesome! I'll work on adding support to my mod to load Bukkit plugins. Of course initially it will only support a subset of the API though.

    I'll have to post back on the outcome, exciting stuff.

    I was surprised how easy this was, although completely "hack city".

    I was able to load/run one of my Bukkit plugins (built prior to 1.8)
    http://dev.bukkit.org/bukkit-plugins/joe-ip-tracker/

    inside my 1.8 server mod! Basically I created a "Bukkit Bridge" plugin using my custom API, that in turn loads plugins just like Craftbukkit would, and provides to delegate events.

    For this one, I added support to call into PlayerLoginEvent (along with onEnable/onDisable, regular plugin stuff). Since the plugin in turn requests the Bukkit PluginManager, scheduler, and console-sender, those have to be proxied out. But it all works! Fun stuff :D

    I'll add playerlogout and command-preprocess events then publish with the next version of my server mod. This was really eye-opening to me on the machinery involved.

    On the downside, when you proxy Bukkit you quickly see how many methods and classes there are to synchronize or stub out (a lot!!). However, I know from coding plugins most people use a select few calls and don't take full advantage of all these methods/operations so even a smart subset if implemented well would be welcome.

    EDIT by Moderator: merged posts, please use the edit button instead of double posting.
     
    Last edited by a moderator: Jun 14, 2016
  6. Offline

    xphoenixxx

  7. Offline

    lecraeman


    Are you implying that the guys writing sponge are somehow stupid, or don't know what they're doing?

    You do know it's by SK (Worldedit and Such), and several VERY well known developers? Sponge is not a joke...I think it could easily take over where bukkit left off.

    Plus, It's based on Forge API so it would provide a server plugins and client mods.
     
  8. Offline

    jwpwns

    There is no test versions for sponge downloadable yet right?
     
  9. Offline

    DoingItWell

    Sponge is not implemented yet.

    In other news though, I've completed adding initial Bukkit plugin support (!!!) to my 1.8 server mod.

    Although just a starting skeleton I was able to successfully load two of my Bukkit plugins.
    http://tinyurl.com/qbyfj2w

    Bukkit plugins are able to use Bukkit scheduler, listen to player login/logout/command/pre-process events.
    No support yet for other events but the machinery is there to add more events which is very cool!

    If anyone's interested I can share details, but just know this is "fringe" stuff only for the brave/knowledgeable. :)
     
  10. Offline

    jwpwns


    Nice but how do you turn off that annoying announcer? and the no cursing thing
     
  11. Offline

    DoingItWell

    Lol, well edit the announcements.txt file the restart server. You could set the delay to a ridiculously high number as one option or remove all its announcements.

    Eventually I plan to remove the 'baked in' features and just have plugins for it. Most of the support is there to enable it although I'm focused more on adding more event features than writing such plugins. I was kind of hoping some Java developers would contribute one or two using the API so far.
     
  12. Offline

    jwpwns


    It seems to crash a lot also any way to turn the no cursing thing off?
    her is my error
    17699ms behind, skipping 353 tick(s)
    [18:37:50] [SVR/WARN]: airgabe moved wrongly! @ world_nether(-156,60,128)
    [18:37:50] [SVR/WARN]: airgabe moved wrongly! @ world_nether(-156,60,128)
    [18:38:06] [SVR/WARN]: Can't keep up! Did the system time change, or is the server overloaded? Running 17039ms behind, skipping 340 tick(s)
    [INFO] File load: ./world/region/r.-2.1.mca
    [18:38:18] [SVR/WARN]: Can't keep up! Did the system time change, or is the server overloaded? Running 12007ms behind, skipping 240 tick(s)
    Killed
     
  13. Offline

    DoingItWell

    well for new maps you need to let new chunks get created a bit. If a particularly slow hard drive it could end up causing fail. I have yet to see it fail though.

    I'll put in an option to remove the curse filter next release. I'll put that in a plugin people can choose to add/remove now that I can separate it out.

    Let me know how it goes if you continue testing. I haven't seen it crash in my tests so far.
     
  14. Offline

    jwpwns

    It takes a good 5 players and 10 mins for it to crash its not hosted on a peace of crap ether its on a 8gb vps
     
  15. Offline

    DoingItWell

    Interesting, which version are you using I'll check to make sure nothing unusual started going on.

    The memory usage wouldn't be a factor with a small number of people and time up. The biggest factor is hard-drive speed with new chunks (my best guess). Especially if there was someone on with speed hacks.

    I've run with 30-40 players on and haven't had issue but that was on a much earlier build. You can type "/diw mem" in game to see memory usage in case something unusual reported there (like too little memory seen). But other than a bunch of new chunks loaded quickly, again not sure without digging further.
     
  16. Offline

    jwpwns

    Its a generated map and i am using the latest version
     
  17. Offline

    mossyblog


    Nope on all accounts and (yes we're all aware of who's who in the little geek celeb pool). I'm merely poking fun at how the overall situation is a bit chaotic and that anyone and everyone is attempting to solve this problem with "File-New-Project" (irrespective of who and how skilled the are).

    Like I said, going after something brand new and ignoring the existing is a bit of a "eh?" moment..

    P.S
    I converted that drawing into 3D :)
     
  18. Offline

    lecraeman


    I really don't like the fact that you're using this whole bukkit thing to promote Canary.

    I would be a lot more apt to use Canary if it was using up to date versions (but from what I understand, latest version is on 1.7.2).

    I am not saying it's easy to update, but I need to use an API that is on the latest version. Which is why I am a bit more excited about some other projects...some which are already on 1.8 :L

    Plus, Reading the sponge documentation I thought they were going to use Forge as a base, and then canary/spout for yet another layer base, and then develop from there.
     
  19. Offline

    mossyblog


    It's a community message about an existing open source project that can help folks who want to start moving away from bukkit given well... bukkit is in a vacuum of leadership / contributors right now. It's a way of just moving forward.

    Everyone needs to use the latest version, and have you noticed there's not a lot of communication around when that will be? given well, it has a strong dependency on the MCP team(s) unpicking the obfuscated code and then moving on from there. As for "Forge" API what does that even mean? are you wanting a client-side API to work the same way a server-side? also the "Hook" concept in forge is similar to Canary? and lastly Forge also relies on MCP so everyone is in the same starting line per say around "1.8" deliverable until the code has been deobfuscated & translated to human readible syntax.

    Oh and Canary is now finishing up the last pieces to 1.7.10 ..but really the difference between 1.7.2 and 1.7.10 is not really noticeable server-side as those releases were more for Mojang "Realms" support more so than anything else.

    As for Sponge.. Cool i hope they get something going but right now there's nothing to hang ones hat on for today.. tomorrow sure its possible, but nobody is saying much about when they will deliver and what support they will/won't have exactly. It's just a lot of open faith to go by right now in a environment where people just want to get on with playing the game?
     
  20. Offline

    jwpwns

    DoingItWell found the problem it was when people tried to go to the nether it crashes.
     
  21. Offline

    Gr0up115

    How about you guys get off my thread? It's totally off-topic. Make your own.
     
  22. Offline

    lecraeman


    Wow now buddy. You're the one breaking the rules posting external links. TSK TSK.


    All I am saying is...it's ironic you're here dissing other server APIs, and then blatantly self promoting your own.

    EDIT by Moderator: merged posts, please use the edit button instead of double posting.
     
    Last edited by a moderator: Jun 14, 2016
  23. Offline

    mossyblog

    I reject your defined use of irony and the words of "dissing" are your choice not mine. I'm simply stating facts, the fact is there are existing code-bases today, the fact is the current formation of "sponge" is either in ideation form or is in kickstart form, it's not at the level of completion other "existing" server projects are presently at.

    "Promoting my own" is to assert that this is a zero sum game, that if you adopt one you can't have the other. I'm merely advising of "an" option (via signature) but you're the one taking a defensive posture against the existence of it because of some loyalty or <insert emotional trigger here> lineage with the Sponge group?
     
  24. Offline

    smcallah


    Really? You think a company that paid $2.5 billion for something is going to stop development of a part that makes them millions upon millions already? They know there are a lot more PC Minecraft players than Xbone. And they know there are millions of Android players as well, and more buying the Android version everyday. $2.5 billion isn't chump change. They aren't looking to lose money when they spend $2.5 billion.
     
  25. Offline

    Alshain01

    Keep in mind, that was posted before the dollar amount was made public. And also, I'm not saying they will do it intentionally, just that they have destroyed PC gaming companies in the past and they will do it again. They have a track record.
     
Thread Status:
Not open for further replies.

Share This Page