Discussion in 'Bukkit Discussion' started by logoster, Nov 4, 2012.

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


    Hi. I took the effort for 1.3.2. At final reckoning before giving up because 1.4.2 had happened, I had put about 250 to 300 hours into it. As well as the previous efforts to build the 1.2.x bukkit builds for forge, using an older technique that doesn't allow for mod<->bukkit interactions in any feasible way.

    I estimated another 300 hours or more of work to get to what I would consider a good state. Most of this effort would have to be redone for every single MC update. I'm afraid to say, that's not "reasonable effort". So I abandoned it.

    I don't expect you to help. You seem unwilling to yield the smallest inch. We've had this discussion before, and you've always either ignored me or told me to go away.

    By the way, the reason we can't use the MCP mapping generator tool is because we can *only* download craftbukkit jars. To run and generate the diff for MCP, we need the mc-dev jar that is decompiled to generate mc-dev.

    Anyway, clearly, you believe modding to be trivial topic unworthy of your time. I don't think many others agree with you but that is your decision. I see no reason to continue.
  2. Offline

    Wolvereness Bukkit Team Member

    Ya know... That IS available... /commence infinite facepalm? This is why I'm saying it's TRIVIAL. It doesn't take hundreds of hours, it takes 2-3 hours to write a tiny script that outputs the mappings in the format you need them. We will NOT write this script for you. I cannot link to the mc-dev jar (and any direct links will likely be deleted by mods), but try compiling craftbukkit yourself and you'll notice something special as far as what jars you need & get.

    Asking any more of us is either a complete lack of effort on Forge's part, or is asking us to change how we handle updating and coding for craftbukkit.
  3. Offline


    So we should say the same thing about porting CraftBukkit? Lets give up the hours and hours of repetitive tasks because its hard work. You don't need to keep doing that, however, we will. Our community will continue because we keep up that hard work and are dedicated to the project.

    You started ForgeBukkit only to drop supporting it because its hard. Its your Forge community that suffers. The Bukkit community will not because we will put in the effort to keep it running.

    We put in "dozens of hours of work every update". Why are you above that very same thing?

    This is getting to be a circular argument. I understand you put in a lot of work. Understand that we do the same thing and what you are asking from us is impractical.
  4. Offline


    We don't use a format anything like what you want so it would take just as much work for us to prepare what you want as it would take you to do the same. Since you're the one that wants it...
  5. Offline


    Actually, you'd be surprised the formats we can consume Amaranth. If you have proguard files, we can read those too. Likely we can build a tool to read any format you'd care to pass us.

    Supporting an ancillary platform like CraftBukkit, however, especially one that seems to be completely unreceptive to anything we have to say? That seems to be masochism. Especially when every update would take us about 100x the vanilla update for the same thing.

    By the way: the server in the maven repository only has an 80% hit rate vs the vanilla server using our mapping tools (Lex just tested it). That means we have to identify 20% of everything by hand. That's still a *lot* of work and will produce a very buggy forgebukkit.

    TnT Are you seriously arguing that we don't put in the effort for every update? Currently fml/forge seems to be averaging about 12-18 hours lag time from when a pre-release is announced to it's partner forge being release. Most of that time is spent, well, updating. Also, do you know how hard it was for me to abandon forgebukkit? You know that I have a whole pile of unreleased code for bukkit/mod integration, that I CANNOT use because I can't get the platform working? Secured ironchests? not going to happen. IC2 personal safes with bukkit group permissions? not going to happen. Abandoning all that work was one of the hardest decisions I had to make. I would pick it up in a heartbeat, if I didn't see 600 hours of work sitting in front of me. 600 hours, multiplied by every release between now and doomsday. Do you have any idea how dispiriting that is? When I was working on it, I played maybe 10 hours of minecraft in 6 weeks. Every waking moment was spent hacking on craftbukkit. I slept maybe two to three hours a night.
  6. Offline

    Wolvereness Bukkit Team Member

    First hand experience, writing something to consume the mc-dev jar would take a fraction of the time to work with our insane mapping format... Seriously, just grab
    You are doing something wrong... Like, horribly wrong... It requires a little coding by hand, but the script can easily be completely automated, and as I said, this automation would be easier than trying to convert what we use for 'mappings' if you even want to call it that.
  7. Offline


    No, I can fully imagine the amount of effort you need to put into ForgeBukkit. I understand because the team puts in at least that much effort, every update, with the pressure of the majority of servers out there waiting for the update.

    It is a lot of work to mod Minecraft. There is no group that understands that more than Bukkit.
  8. Offline


    Err, have you looked at FML? Check out one of the libs we download. Also check out
    We know a little bit about ASM. That is using ASM to merge the client and the server into a single jar. And yet, mc-dev is still a complete pig. We don't know what you've done to it. But our tooling cannot generate a good "hit" against it at all. Also, don't forget, this is phase 1. Phase 2, is to decompile craftbukkit jar using those mappings into MCP format, so we can generate diffs of craftbukkit against vanilla MCP. Then we have phase 3: merge those diffs into the FML and Forge diffs in a meaningful way. At the end, we would have something resembling craftbukkit for MCP. Then I need to wrap that in a coremod that injects this craftbukkit into the system at certain key points allowing for both client integrated server and dedi server.

    Yeah. Trivial.
  9. Offline


    You say that writing a small script to automatically identify these mappings would be easy, trivial, or any other word. But from actually doing it, it's not. If it were, then we wouldn't be talking here. We wouldn't care cuz it'd be done.
    However, I propose a solution, why don't you just spend 10 seconds to send us your mappings file as is, no alterations and see what we do with it. Let us decide if it is easier to write this 'trivial' script you speak of.

    I can garente you it won't be.
    Nolzi likes this.
  10. Offline


    The party with the lower hand, does not have the ability to propose solutions. The work you say is "a simple copy/paste" is no such thing. You are requesting something which takes us hours on end, pouring our sweat and blood into in order to provide the community with a server that is both stable, and expandable. We, the developers, spend hours, days, weeks, every update, using our well constructed tools (re-constructed over time) to provide everyone with the product we deliver.

    Your demands will never be answered, in fact, you have been told time and time again, that we will not hand you our work. The manner in which you are constructing yourselves is something that is intolerable of any organization.

    The very fact that you have to bring this to the forums is intolerable. This thread should have been locked on day 1. I will make an attempt to sympathize with you, however, in regards to this:
    Welcome to the development of CraftBukkit. I myself haven't played minecraft in over a year.

    This topic has been locked.
    blha303, mbaxter and afistofirony like this.
Thread Status:
Not open for further replies.

Share This Page