@atomix said:
Can you provide more examples how move to Gradle can improve JME's game deployment or project structure .. :p Thanks
Maven has taught me to despise it over the years. In my day job, I went from ANT (worry about the build in hours per month) to maven (worry about the build in hours per day) to gradle (worry about the build in hours per month again). Not to mention how many “gotta get this out now” deadlines we ran into where maven was the thing holding us up with some issue… only to be met with maven dogmatic responses like “Well, you shouldn’t do that.” So, I will literally go miles out of my way to avoid maven… and I think it’s kind of a funny argument “my build tool is so complicated that I need IDE integration just to function and so now I need that from every build tool…”
On the whole, given all of the numerous maven-hell downsides, Maven has no strong improvement over the existing ANT build in the SDK… which would integrate nicely in gradle if we ever decide we want something more “module-friendly”.
Anyway, I only mentioned gradle because you are a groovy fan. I never use it from an IDE so that never occurred to me. I just setup the build file and then never think about it again (for my non-jme projects).
@zarch:
Indeed, that’s just a reference of how I have the idea. The post is all about Getdown’s magic and he just want to show that Getdown can be called with Maven. That’s nice but what about the opposite (the idea)!
Maven have all the download - dependency magic in one POM, and some deploy station (server provider) for open source projects all support Maven.
The blog was out 2 year ago, unti then I don’t even think about a game that can use Maven + Osgi at all. But then I think about it once again, I can deploy an Open source updatable application (don’t metion web-based one) with Maven and Osgi:
Free server for my files (mean libraries and assets).
Decrease the load in term of user’s software update for real server for handling multiplayer logic and data (database).
Really open-source spirit, free the dependency hell
So it work (consider oppositely) like this: Whenever I changed my POM and some other light-weight files by Getdown, the user then get his libraries updated by Maven via Maven’s central repository.
@pspeed: Well, at that time, I don’t know about Gradle and even Groovy yet! The simple and clean of Getdown script can compare to Gradle at some point, just about deploy and download aspect. I used Getdown anyway but the idea of Maven + Osgi based game is quite interesting as I said.
@pspeed: yeah, I experienced the same with maven. The funny thing is that maven praises itself to have a standardized build process. So in theory it would be easy to switch between projects and even companies without having problems understanding the build process used there.
Now in our company the maven build process is heavily bloated with plugins. So only some mighty dev’ god knows what’s going on in there.
But all in all it works well but as what I experienced: people should use maven only for larger projects with dozens and hundreds of dependencies. in other cases its just an administration overkill.
@ceiphren said:
@pspeed: yeah, I experienced the same with maven. The funny thing is that maven praises itself to have a standardized build process. So in theory it would be easy to switch between projects and even companies without having problems understanding the build process used there.
Now in our company the maven build process is heavily bloated with plugins. So only some mighty dev’ god knows what’s going on in there.
But all in all it works well but as what I experienced: people should use maven only for larger projects with dozens and hundreds of dependencies. in other cases its just an administration overkill.
From a purely systemic perspective, gradle has every advantage (minus the dogmatic limitations) of maven without any of the downsides. When IDE support catches up, hopefully we can see maven fade away as nothing but the standard format for public repositories (which gradle fully supports by the way).
In the phase Zero, I decided to use Arianne instead of Three rings. Why?
As I’m working in a game company, I’ve learnt by heart the importance of Game Design document and Workflow. And I also have fews failure attempts in the past as hobbist game developer with small team try to make BIG thing due to lack of these.
So Why Arianne?
Because it’s Open source, simple, have enough features, easy to deploy, test, and most important very well documented compare to Three rings. And i also have nice experience work with the team there.
Features
Stendhal - MORPG featuring hundreds of NPCs and quests
<li>Stendhal - Huge and beautiful world to explore</li>
<li>Stendhal - Statistics website and Hall of Fame</li>
<li>Marauroa - Game server handling client-server communication</li>
<li>Marauroa - Database persistence with asychronous access</li>
<li>Marauroa - Flexibility of game rules. Apply to drawing boards, card games, PacMan …</li>
<li>Keep It Simple approach</li>
<li>Release early, release often</li>
<li>Automatic client updates</li>
<li>Detailed tutorials to extend Stendhal or get started with Marauroa</li>
<li>Supportive development team in IRC</li>
It’s wiki with a lot of tutorials of making Quest and NPC, scripting: Arianne
o The idea
In Oriental culture, we all love the legend of Monkey king who traveled to the West and become a Buddha.
I compose the idea with wild jungle scenes in fictional oriental - western mixed scenery and theme, and adventures along the jouney.
o Plan
My initial plan for the phase Zero is to create a port to jme3 (means 3D) version of the existed game Stendhal with free 3d models and 3d gameplay. Then I let the team go in, try and discuss, learn, design as much as possible before we going futher.
In the first attempt, 3d enviroment and models will use open art website such as:
@atomix
I’d love to get involved however I’m still learning code and my issue is mainly time, struggling to find enough of it just to finish my TerrainTiler system.
At this stage I need to focus on that for my own learning but after that’s complete (and if you need/want to use it I’m all happy to share and work it with you) then I’ll be more open for working.
I do also have my own MMORPG game I’m wanting to develop and experience is all a help, (see http://thegame.radans.net for basic info) so I’m planning this path anyways, just need the time as I said, even considering taking the risk and just quitting my stressful day-job and going it full time, still might…
So not sure how I can help given the above but open to ideas. ( I probably get about 2-3 hours spare about 3 days a week at this point, some of with gets stolen by stress management (Neverwinter or Minecraft )
Never worked with a team before of software, always gone alone, but let me know if you think I could help, or ask me anything. (btw my timezone is Sydney Australia - I live 2hrs drive south of there)
Quote: “My initial plan for the phase Zero is to create a port to jme3 (means 3D) version of the existed game Stendhal with free 3d models and 3d gameplay. Then I let the team go in, try and discuss, learn, design as much as possible before we going further.”
I read that as taking a 2D game and porting it to jME as a 3d-Game (iso like diablo etc I assume) and building from there.
No?
If I’m right I’d say thats a good start, something with a known goal and style to feel the water and provide a base to build on and learn from.
@pspeed said:
From a purely systemic perspective, gradle has every advantage (minus the dogmatic limitations) of maven without any of the downsides. When IDE support catches up, hopefully we can see maven fade away as nothing but the standard format for public repositories (which gradle fully supports by the way).
Yep. The jME SDK will move to gradle at some point but the current ant build scripts can be integrated in any build environment already.
Its funny, in the times of jME2 people spent most of their time setting up their build environment, integrating libs etc… Now the SDK really makes that unnecessary and you can care for special requirements for your own projects later. Still many people seem to enjoy setting up theoretically nice build systems and then try to change the world around them to fit it (got POM? ;)) Mostly its because of really silly reasons like one function that seems slick (while many others don’t even exist) or the user being acquainted to some piece of technology (like Eclipse or maven).
It a) doesn’t help the community much as basically existing functionality is reimplemented and b) doesn’t help the user as he cannot apply functionality thats made available for the SDK and standard build process (desktop deployment, applets, multi-platform compilation, asset tools etc.).
@radanz said:
Quote: "My initial plan for the phase Zero is to create a port to jme3 (means 3D) version of the existed game Stendhal with free 3d models and 3d gameplay. Then I let the team go in, try and discuss, learn, design as much as possible before we going further."
I read that as taking a 2D game and porting it to jME as a 3d-Game (iso like diablo etc I assume) and building from there.
No?
If I’m right I’d say thats a good start, something with a known goal and style to feel the water and provide a base to build on and learn from.
My 2c anyways.
Cheers
Edit: PS The forum code tag bar is gone again!!!
mhh ok my bad, i mistook arianne framework with a 2D engine called Angle (I knew it was starting with an A )
@nehon:
Hehe, next time read more carefully
Regardings Arianne, Yes it’s 2D game engine indeed.
From Arianne to jME3
The reusable:
Network: Good for small scale game
DB: MySQL or integrated H2
Almost gameplay: Enities definition, Quest, Scripting
They also have a complete website for the game with tutorials, wiki
The different between a 2d and 3d, Arianne an jME3 engine:
Replace the Arianne game loop with jME3 states and update loop.
Delete the Render task of the game view JPanel
Terrain : I have my own tiled Terrain implement in jME3 for almost top-down game.
Characters: Use 3D Models of Open Art resource
GUI : Use pure swing gui (port to Nifty later)
2d to 3d Gameplay: Map entities, trigger
2d to 3d Picking : I use a simple translation
We use git. check the Google code link and PM your gmail.
We also can use skype, facebook, gmail for contact.
The new google account of the project is : ancientmonkeyworld@gmail.com . No supprise!
I also made a simple website, dropbox, google+ …etc with that new google account for this project Cheers!
this post is 2 years old, but the idea is fantastic. Is there any news here? Why the project is stopped? Are there any showstoppers?
It will be great, if we reanimate it?
“Just stop” become a common MMO game problem even for a big team because the amount of works and financial strength, ideas and manpowers needed to make it actually happen.
Base on my personal experience at the time, when I was working fulltime at a big company who procedure big titles. There is NO FREE TIME. So the lesson for me is: even I can spend months continue to roll the code, my time budget exhausted quick. I also want to mention the trend of MMO change very quick. If I was fast enough I may keep up with the trend. But because I was extremely slow in developing this project. Well, it should end as the interesting lost.
My summary about making such project, if you are a very skilled developer and familiar with Multiplayer game coding, including servers and networking, you can make a small MORPG prototype, is quite doable with 1-3 devs.
The result of my project is also a prototype: a 3d game with a generated ( so can become massive) world, players can play together in a big scenes and interact like fight or chat. Beside of that, I’ve created a lot of tools to create various types of characters and composable mini gameplay, which is very satisfied result.
I stopped at the point I think I can not go futher because the lack of knowledge and skills in creating beautiful world and characters.
I don’t think I have time for MMO rightnow because I moved toward mobile games. If you want to learn and try, you can send me an email. Restart this kind of project is useless. You should create a new one with your idea, I hope I can give you some code to help you start up quickly.
Hello, it might sound a bit weird, but I just saw your post, saw the
answer you just send and I must say, I might be interested in some of
the stuff you have if you are willing to share. I started building an
MMORPG like 6 months ago and so far the project is going pretty well,
even if I am the only programmer active at the moment. I am using, of
course JME3, but decide to go on object DB for the database. It also
uses ToneGod for the 2D part and both TCP and UDP so I can use the
best of both. We are for the moment a little team, build up of me and
2 guys who study in game design.
I got a little formation in project management and I am seriously eager to keep the
project going, even if your story is one of the many out there who
show how difficult this type of project can be.
So far all the showcases of the project are in french so… it might need to make some traduction
before I can really put anything on the forum, but I aim to make it
bilingual in the end, so I might as well start it already,
Anyway, I just started to look at everything you did, but if you got anything, like advice, or piece of
software, or people you think might be interested, just send it to
me, i’d be more then happy to get any help i can!
I’m very into open-sourced but I don’t really want to manage one. Managing Opensourced project usually return in headaches at very first glance. Anyway i you still insist i will push some of my code into git:
A tool to create/ compose characters from patches like clothes, hair style, ammor. It’s a Netbean plugin which integrated with JMP at that time, don’t know if it will run in current SDK because it’s Nb 7.3 now.
A 2d/3d tiled map editor, world editor which will export xml in tiled format or save to an un optimized j3o format. Decal placing is possible but I also not optimized it. The terrain tile paged algorimth is base from Forester, note: also abandoned.
An “interesting” lightweight actor framework to enable multi-agents interactivities. Base in http://jumi.fi/actors.html
… something else
As you are developing an MMO with your team. I might say it’s going to be a big challange, not only technical aspect but also the “output” of the project. What it will become if you can not find a sponsor to put it online. Maybe gambling with something like steam greenlight or kickstarter, maybe sell your house for a big amount of $ to support the team and server.
As for research purpose, it’s likely we should invest into RPG part of the game more than massive aspect. Getting result quicker, getting notices quicker, not as much of time, efforts and money.
I think you should put me an email at atomixnmc@gmail.com .About the repo I will put it online soon but don’t know when
)
)
