Server owners, where do we go now?

The project "Pore" found here:
http://www.reddit.com/r/spongeproject/comments/2frfaa/pore_a_bridge_for_bukkit_and_sponge/

May be the most direct way to allow the bukkit-project plugins to work with Sponge.

 

Welp...

http://canarymod.net/news/canarymod-10-released was released today.

The Canary Team are basically quietly working on getting stuff done and less running around waving their hands in the air "The Falling sand's falling!..." hehe. But basically we've been really pushing ourselves day and night to get this release out the door to get the code base stable for at least 1.7.2 server.

Now yes its 1.7.2 - but - there's not much significant difference between 1.7.2 and 1.7.10 for client-side players (bug fixes and Minecraft Realms readiness). Now that being said we're also majority of the way right now for 1.7.10 release, as we wanted to get 1.7.2 out the door to let the code base soak first (why pick two fights at once). At the same time we're also working on a 1.8 update aswell, but there's some challenges there around attaching the hooks into the right place (as you can imagine the code moves around between releases, so finding the "ahah there's the familiar spot we need" is something of a annoyance / minor challeng!

None the less, the team are expanding and we aren't impacted by any of the DMCA drama. Instead we're more focused on the future!

Coming up next are more mod ports, and more tutorials. Over the course of the next month if you guys are interested, you're going to see some massive changes ahead for us as both a team and project.

What i'm trying to say is - There is a plan and there's a mature code base already in place. Yes Bukkit API is not compatible but we're thinking through some ideas on how to fix that so it can have less impact on Bukkit Mod developer(s).

 

mossyblog Isn't CanaryMod Helping Develop The API "Sponge"?

 

A little project I'm working on :
https://github.com/john01dav/ServerForge

 

Spigot is certainly not dead. Go check out their website to see what is going on. Spigot is fantastic, and is doing a great job at the moment.

 

Awesomedanguy As far as I know, no. We have a member of the CanaryMod Team listening and working with Sponge on an advisory standpoint but ultimately CanaryMod Team are 100% focused purely on CanaryMod. As we have a mature project that's VERY similar to CraftBukkit already and we think its more worthwhile continuing on this journey

 

Be careful with thread like these. I tried helping people with the same question in IRC and I got banned. Bukkit "ex-staff" still seems to be pretty hardcore about "advertising" or whatever.

 

Its not an ad, it's just a graphical helpful advice to folks who want an alternative while the debate between its future continues.

 

I agree with you, trust me. If I was Bukkit "ex-staff" I'd be encouraging this kind of conversation. I'm just saying they are still very hardcore against it and are throwing bans left and right.

 

While we're offering alternatives, I've added plugin support to my 1.8 server mod also
https://sites.google.com/site/superhappyminecraft/1-8-server-mod

Interestingly, I submitted a dev application for Sponge but never heard back. I put as reference I had one of the only 1.8 server mods, go figure!

Maybe they have enough developers than know what to do with currently.

 

Well my intent is definitely not to offend so it draws negative attention I'll definitely retract the graphic.

 

I like what you are doing is there a way to be able to set a hook for bungeecord so i can connect to my hub

 

I've never used BungeeCord so don't know what hooks it needs. If there was a specific list I could work through it, anything is possible.

My priorities right now are to just continue flushing out the basic API since it's still at infancy stage. My plan is an Agile-like approach where I add things along but keep a working build going along the way.

 

Im working on a spigot/bungeecord component and would like to talk to you.

 

There have probably been a lot of applications submitted, as with the uncertainty of Bukkit at this time, interest levels are quite high. Patience is a virtue.

 

With my server, i think i will stick to Bukkit v1.7.10 for some time and after that, i will switch over to MCServer.
MCServer is 100% opensource and they dont use any of the code from the official server software, so i think it would be the best option at the moment.

 

I agree it's Spigot's already in place and seems to keep helping their members, even with the restrictions(DCMA) in place.

 

Going to be flat out blunt here - there are several projects here, upstream at the top of the hill is Mojang's server code. (and with the new mojang managers stupid EULA the crux of our current issues..) Next down is the bukkit core built on it - Down stream from bukkit core is several projects managed by several groups and subsets, many of which kept starting "sh*t fights" with each other. Most you will not have heard of many unless like me you poked around the old jenkins and the like. The two most well known downstream projects are DING - "CRAFT BUKKIT" and wait for it... "SPIGOT" There is also FTB and tekkit the client side projects. For the most part the sh*t fights were not obvious, as any reference to spigot even in passing was usually locked or deleted with no explanation beyond a "protocol hack"

Both were reliant on upstream changes to bukkit core. Both could run the same plugins - although spigot had some very specific capabilities such as retaining legacy calls to allow abandoned plugins to keep working longer, and the ability to run as a distributed server platform.

Trouble is those working on bukkit core also worked on craft bukkit, but very few (publicly at least) admitted to working on spigot - even tho it was a downstream too - thus any work there was outside of their control. The main issues here was spigot who's aim was supporting heavier duty servers - would rapid prototype experimental builds to keep server uptime high, while bukkit focused purely on high quality code - instead seeking to minimize errors.

Spigot's focus was on error handling, craftbukkit's focus was on stopping the errors to begin with. Both approaches are perfectly acceptable from a development sense - and for the most part transparent to the end user.

Subsequently Spigot often /appeared/ to got a release out faster than craftbukkit and as a result the bukkit/craft bukkit team often got a lot of angry feedback, by players and operators alike, who did not understand that all the work in bukkit core (by the craftbukkit team) automatically filtered to spigot helping them get out builds faster anyway.

Put simply, bukkit core and craft bukkit are two projects managed by the same team, in any update bukkit core needs to be sorted out, BEFORE craftbukkit could be updated. On the other hand spigot only needs concern itself with spigot alone; so the second the craft bukkit team got the core sorted, spigot could immediately throw out a release. A programmer cannot do two things at once - but two different ones can.

It is not a case of craftbukkit being slow, it was a case of they had to update two separate projects to get a release, while spigot only had to update one. If anything the bukkit side were fast as hell, often getting /both/ core and craft builds ready within moments of the spigot version.

Spigot was also taking a lot of market share, taking the spotlight away from the hard working craftbukkit team who were in effect contributing to both projects willingly or otherwise. Programmers are only human, and with nothing in return for their work but fan adulation that is going to hurt.

I can't speak for the spigot side of the "flinging match" but at no point did spigot remove the "bukkit" reference in their version, nor pretend otherwise. While all I ever saw from the craftbukkit side was accusations against them.
Admittedly this seemed principly from the moderators, not the coders; but in keeping quiet the programmers are as much to blame.

I could be totally wrong, it is all in the way they acted and marketed themself. I was not privy to the internal political arguments between the two teams.

To be honest I have been alternating back and forth between spigot and craftbukkit, and to my thinking, had the two teams worked together; the results would have been very impressive. Both teams had merits.

The question is now, with a mutual threat can the two teams (what is left of them) put aside their distrust and work together to save the mine craft community? Or do we just let 5 years of our lives just fade from memory?

Rumor has it many contribs are moving to sk's sponge project.. I can only hope it works out, but ive no info on it.

To be frank as much as i hate DMCA, it was probably justified. It is one thing to see how a protocol works, it is another entirely to hack someone elses code. Had bukkit been built on original code instead of built on mojang code it would have been a non issue.

There are also other factors in play, it is a protest against a million dollar company secretly profiting from volunteers work, many of which have as much right to those million dollar mojang bonuses as they did. Also the new minecraft manager's EULA which has killed community servers - and the appearance this same manager may have just betrayed notch by selling his company out from under him. Now that is not to say this is what actually occurred - but it looks that way which is just as bad, but MR EULA seems to have overlooked the fact that at least 700million of that 800million on mojangs balance sheet came directly from players buying the game to play on those community servers, servers i might add that are so damn awesome BECAUSE of those unpaid coders; many of which covered their own costs running those servers. Ye sure call it pay to win, and sook about not getting a cut; everyone else called it pay for the damn community server.

The fact he implemented "realms" then did something that in effect killed off all the public servers at the same time suggests this guy is a riaa/sony/bmg grade anti-trust a$$hole. This is mojang not Microsoft... oh wait..

Nevermind uhuhuh...

EDIT by Moderator: merged posts, please use the edit button instead of double posting.

 

There are possible downloads for the latest versions of bukkit and spigot. You just have to google a bit. But don't use them, as they are illeagal because of the dcma, you know!!!???

Spigot created binary patches to update the latest version of spigot (#1649), they provide md5 hashes to verify the files. Spigot allowes 1.8 clients to join the 1.7.10 server. The patching progress is easy and detailed explained.
http://www.spigotmc.org/threads/spigot-update-20140911a.29318/
http://www.spigotmc.org/wiki/spigot-patcher/
http://www.spigotmc.org/spigot-updates/

 

I am just not interested in letting 1.8 clients connect if 1.8 features are not available. I would rather not add to the confusion as to exactly what works and won't especially since I have a creative server.

 

I think Sponge will go well, & CanaryMod. From what I've heard I think Sponge will go better. They have the Spigot Devs ( like md_5 ), forge developers, and a lot of other people, including 1 staff from CanaryMod. That is my opinion. Anyone else agree?

 

For 1.8 long term, Sponge looks good.

For 1.8 clients connecting today but no 1.8 features (yet), Spigot looks good.

If you want 1.8 features today, you can use my server mod.

It would be interesting if someone from spigot had an ETA for when they thought 1.8 support would be available.

 

Canary struck me as basically patch hack on vanilla, as such its legal standing always struck me as ambiguous, admittedly it appeared to have some epic plugins, none of which ever worked for me, so i quickly switched to bukkit when i first started my server.

How hard would it be to build a "dumb" middle man client interface? Ie a "dumb" api that can communicate in client protocol, and can accept connections, but has no mob/ai/terrain/weather logic. This would contain no mojang code at all, although it would need to "talk" the same protocol to the client.

Then code the /other/ side of this "dumb" api to allow incoming (or outgoing) connections using some sort of standardized command set. This API could then be used to "talk" to the clients - and on the server side, the actual system running the world/mobs/terrain could pretty much be coded in anything from C, perl, python, basic, morse code whatever - so long as it speaks "dumb api" commands.

This would make client updates almost a complete non issue, since all that needs to be done is add a few more functions to the dumb api to support new features, change the reported server version, and add a new command for it. Designed well enough this "dumb api" could require little more than dropping in an updated conf/ini file - and when server operators "feel like it" they can add things to support newly added commands - or ignore them entirely. Totally client independent.

IE.

CLIENT1 Client2 Client3 etc
\ \ \\ \\
DUMB API SERVER
Command /translation interface
/ /
Actual server running, c, perl, pascal, basic, java, python, sign language,
lua, morse code, smoke signals whatever

This would also free up a lot of wasted talent/resources wasted bug hunting, to create an epic server, or an epic plugin for a server. Plugins would never go "out of date" ever.

This also adds something new, automatic fail over. Since the client interface is separate to the server - a server crash could be recovered by switching to a backup server automatically. Maintaining the "DUMB API" would also require a lot less work debugging, the only real work would be maintaining a document listing the current supported command feature sets.

Suggested water theme project name: wishingwell ?

... Assuming the playstation/xbox/tablet versions of the game have multiplayer support, that also allows the possibility of all platforms talking to the same world server.

EDIT by Moderator: merged posts, please use the edit button instead of double posting.

 

To avoid feeling upset I made a funny picture about the situation... I hope you find is a comic relief as well.

 

I think that piece of lithography is pretty accurate.

 

I think the Microsoft Herobrine needs to be bigger Your depiction is totally accurate, give it some time to see it really happen.

 

Are you referring to Sponge to "SpongeBob?" It's the buckteeth that gets me.

 

I say false. I updated my plugin two days ago. The file was approved the day after uploading it. BukkitDev is still looking after us! Thanks guys, for those who stayed.

 

Just saying, Bukkit WILL be updated for 1.8. Nathan Adams (Dinnerbone) himself said he will personally see it through that Bukkit will have a 1.8 update. Check out this link: https://storify.com/lukegb/the-tale-of-bukkit-for-minecraft-1-8

 

That's a nice story and all, but it drops off right before Microsoft buys Mojang and the uncertainty that ensues. Yeah, before that happened, Dinnerbone may have planned on updating 1.8, oh and lets not forget the DMCA action shutting it down as well? Your Storify needs updating.