Jump to content

paul

Members
  • Posts

    2
  • Joined

  • Last visited

Converted

  • Gender
    Male
  • Location
    MD
  • Personal Text
    I know a thing or two about Java...

paul's Achievements

Tree Puncher

Tree Puncher (2/8)

0

Reputation

  1. Don't get me wrong, from what I've seen elsewhere I think Minecraft Forge has the best API and build process going. If having a boatload of tasks works for the developers and modders, great. My opinion is that the number of tasks could be greatly reduced by eliminating IDE-specific tasks through an IDE-agnostic build process. It's also possible the codebase may lend itself to being organized into subprojects but I haven't had a chance to play the with code yet. I just today finished another side project (building and securing a Multicraft server) so with any luck I'll finally get to tinker with some code! Paul
  2. It's not "supposedly" - Gradle is better than both Maven and Ant for many reasons. First, Maven: Maven is happiest if you do things Maven's way. Customizing your pom is a chore and a half, and despite Maven's dependency management you'll still run into problems with conflicting plugins in a moderately complex project. There's nothing wrong with Maven's build process - it works great out of the box...if you want to play strictly by Maven's rules. Ant's strength is ease of customization (relative to Maven). While it's true you get nothing free with Ant (i.e. an out of the box build process) the joke is that there's only ever been one build.xml ever written and it's been passed around and modified ever since. Practically speaking anyone developing with Ant will grab an existing build script and quickly tweak it for new development work. Ant does not come with any dependency management - that has to be added on via Ivy. Ivy itself is a pain to configure so it's not like Ant+Ivy is a dream solution. A major problem with both Ant and Ivy is their respective build scripts (build.xml and pom.xml) are both written in XML. XML is not a scripting language - it's a markup language. Any solution based on scripting with XML is going to be problematic. This is not a knock on Ant (which I love) and Maven (which I don't) - when both projects were started XML was arguably the best choice for the next generation makefile. Enter Gradle (of which I'm a huge fan)...Gradle combines the best of Ant and Maven while improving on both. Gradle's "makefile", build.gradle, is written in a domain-specific language based on Groovy. Groovy itself is pretty, uh, groovy, but that's a lecture for another day Gradle comes with out of the box capabilities and dependency management a la Maven and is even more flexible than Ant. Gradle uses Ivy under the hood and the integration is done so well that dependency management is easier to configure than it is in Maven. Whoever had the foresight to go with Gradle should get a raise That being said, the Forge build process is kind of a mess (the absurd number of tasks in build.gradle are evidence of that) and I think that's a big source of confusion for a lot of people. No matter the build tool of choice (Ant, Maven, or Gradle), the use of an IDE should not be a factor if the build is constructed properly. It should be possible to pull down the code out of [Git, SVN, whatever] and build a working distro straight from the command line (yes, even with all the hoops that must be jumped through for Minecraft modding). When it's possible to build a freshly exported codebase from the command line then it's trivial for a developer to use any IDE or text editor. The more time spent mucking with the build means less time doing the fun stuff...writing code! Oh, and hi, I'm Paul. I'm pretty handy with Java and such but completely new to Minecraft modding and Forge so I'm sure I'll have a lot of questions. I'm happy to answer what questions I can and hope to be able to contribute to the project at some point. Paul
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.