[Approved] New branch in code and management

Site & Project
1

Important Update 11th June 2009

mojomonk wrote:
It looks like you've done due diligence about taking over the site, no one seems vehemently against it. I certainly have no problem with it, I've only continued keeping the site paid for and up because no one else would. If you are ready to move to the next level, let me know how you want to proceed.
(...)

Also note: Thread starter 'erlend_sh' was formerly known as 'Sadr'



I have two rather simple but highly potent proposals that I would like to offer the jMonkeyEngine community:

Firstly:

This week I sent an e-mail to Mark "mojomonk" Powell, asking him whether he'd cooperate with me to do a complete transfer of ownership for jmonkeyengine.com, provided I managed to get the community's approval. He said, if I got my approval from the community, he'd work with me to complete the transfer. The only thing this transfer would mean initially, is that the jmonkeyengine.com domain and all of its content would be moved from mojomonk's webhosting at Lunarpages over to my personal Dreamhost webhosting, or to the private server (hosted with hosteurope.de) of a trusted friend if need of more space. Oh and, it was kinda this 'new generation?' topic that gave me the idea.

In conjunction with the transfer, mojomonk would be relieved of all his administrational duties (and burdens) and I would step in as a temporary administrator of the entire website, and I say temporary because I intend to find people more competent than me to deal with the more technical parts of maintaining such a website. Any one interested in joining me from the start in the role of administrator would be more than welcomed to do so, provided you've shown the necessary enthusiasm and competence to get the 'confidence vote' of the community, same as I'm aiming for now. Actually, it might be that the change in administration will happen before the actual switch of servers, but either will definitely happen shortly after the other.

If you're curious about my general intentions as a co-administrator of jME, you might as well just look at the old 'jMonkeyEngine 3.0' debate. My views expressed there haven't changed all that much with the time that has passed. I wouldn't necessarily do everything according to that plan. Rather, I would make perpetual invites to open debate in the community, then do something when talk has gone on for long enough. My ability is restricted to the administrative, but I do that one thing fastidiously so.


Secondly:

I want to see jMonkeyEngine 3.0 take shape. Most definitely jME 2.0 must be maintained and will probably be the engine of choice for many projects still to come, but in order to remain a viable Java game-engine (maybe that's not how it's described, but that is how it is predominantly used), the 2.0 branch must come to an end and we must see a new beginning take shape. It's been said already. jME 2.0 has done well enough in the hands of the community this past year, so we might as well keep it that way.

Now I've been away for the past 4 months, just to come back here and have MomokoFan (friend and partner in past projects) tell me "check out my 'jME 3 test' thread"... Lo and behold, he actually pushed out that 'from-scratch','next-gen', 'fresh approach' to the 3.x-line of jME that I always wanted, catchphrases and all! So here is MomokoFan, able and ready to spearhead the development of a next-gen branch of jMonkeyEngine as head developer, merely asking for his work to be accepted as the official next version and future of jMonkeyEngine, and not just a 'test'. Yes, why not?

The way I see it, bold moves like these will always be a 'test' in some sense until it really kicks off big and faces widespread adoption, or rejection. But the test won't thrive without proper support. The active part of the contemporary jME community now largely consists of people who adopted jME 2.0 for their project some time ago. That is fine really, and it keeps those minor patches coming daily. Making MomokoFan's 3.0 branch official would not entail anything more than a couple extra boards added and a new section in the wiki getting started. The code is already there in the repository, working, growing, developing. By embracing this new branch, we'd be looking at a potential influx of new developers and members of the community. I would be more than happy to help manage this influx, and make sure to keep every urgent task under the spotlight.


So what say you? Would you be willing to give me and MomokoFan your vote of confidence, trusting us to step up to the part of 'administrators', to be in charge of the development of jME's next generation, the 3.0 branch?

Mind you, all current developers, moderators and even admins could maintain their position if so desired, as we would merely direct near all of our attention to the development of 3.0, and leave 2.0 in the hands of the community and whatever management is still around to support it.
2
Alric said:

I would be concerned that a branch with only a single developer could quite easily end up running out of steam.

It is highly unlikely that I would adopt such a branch in the forseeable future.

You're right, this is why me and Sadr have decided to move this forward. We believe that the only way we could pull this through is by involving the community. The fact that already some developers have reviewed the code shows that this is not a single developer effort.
If I were using the jME2 branch, I would be more worred though, as it currently has no developers at all to support it.

Alric said:

I would also be concerned that Sadr you have not been a consistent member of the jME community. I have the impression that things build on jME are your primary interest, rather than the core engine.

Sadr is not a game programmer, but more so a game designer, with experience in project management. I have worked with him quite a lot and I think he could be a great member for the jME community.

Alric said:

I have no wish to hamper your efforts, but there I have spoken my mind as requested.

Thanks. Constructive feedback is always appreciated.  :)

mcbeth said:

if we are to considering moving forward to a jme3 that is decidedly "next gen" then consideration must be given to the fact that alot of hardware will be be left behind mine included

Mcbeth, I have already answered that portion of your post, so I am just going to quote it here:

I take back what I said that such an engine will require powerful hardware. Now that I have completed the renderer portion I see that the main requirement is OpenGL2, with support of advanced DX10/GL3 features on compatible hardware. That means you should be able to run jME3 any cards supporting Pixel Shader 2.0, that includes GeForce FX 5200, Radeon 9550/9600/X300 and later cards. Additionally, the engine supports a pluggable renderer architecture (similar to jME1/2), which would allow writing a renderer that doesn't use shaders and thus supports virtually all video cards.

I am also going to add, that the JOGL renderer in jME3, is planned to support all video cards with an OpenGL1.1 minimum requirement. Some changes in the latest local copy adds support for RenderState specification in the material files, which would give further support of fixed-functionality in jME3.
3

I agree with all that was said and I support Sadr all the way  :slight_smile:

If we want jME to continue to be the best Java 3D engine, the only way to do it is by starting from scratch. The existing jME has served many people for quite a while, but with time and lack of major support, it lost its shiny and polishing it won't help. That's what jME3 is for, to fix the mistakes of our past and create an even better product, supported fully by the community.

4

I think this could be a move in the right direction. Momoko, you have surely shown an interest in moving forward and you have something to show for it. I have no objection, you have my vote.

5

I agree too

6

There is quite a bit gong on behind the scenes of jme2.

With blaine working on blender support and other things, wonderland working on their own branch, various fixes and contributions coming through the contribution depot.

That really could use some management/coordination. What jME 2 is lacking most currently are proper releases/quality control etc.



If your new hosting server can provide better support for nightly builds / set up hudson etc, that would be nice to have.



What i miss about momoko's jme3 idea is some kind of roadmap and more input from other experienced people, it seems to be a bit of a one man show, and i just think its too much for one person to create a whole new engine in a resonable timeframe.


7
Core-Dump said:

There is quite a bit gong on behind the scenes of jme2.
With blaine working on blender support and other things, wonderland working on their own branch, various fixes and contributions coming through the contribution depot.
That really could use some management/coordination. What jME 2 is lacking most currently are proper releases/quality control etc.
Yes with the right management I'm sure jME 2 could be in a completely different place right now, and its widespread use and adoption cannot be ignored. And you might find that management if you go look for it, but I'm not it. I have my faith in jME 3.x and MomokoFan's competence to back it up, which is why this will be my center of attention. Of course, if I'm granted permission to make the migration to a new and better server, 2.0 will certainly have all the same technical benefits as 3.0.

See, part of the big picture here is that a 'fresh start' is exactly what is needed in order for people like me to be able to actually wrap my head around things and get properly involved, instead of dipping my head into that which is the jME 2 monster and have it bitten plain off. jME as both a team-effort and community has accomplished great things, quite astonishing even, yet its scope and direction no longer anywhere to be found. But enough of this; I am not really interested in debating far and wide regarding my issues with the jME project as-is, I'm just trying to offer a solution to some problems, hoping to see something new and exciting take shape in the process. What you should be considering are the pros and cons with me taking over the hosting from mojomonk.

If your new hosting server can provide better support for nightly builds / set up hudson etc, that would be nice to have.
If we go with the private hosting, that is definitely something we can accomodate. I'm sure there's a lot of disagreement about ideal solutions for this though, so whereas it might be easier for us to make quick efficient decisions as the small compact team we will be behind 3.0, the 2.0 line with its myriad of involved and influential contributors is a whole different story. That is my fear anyhow. Yet again, this is another reason why I've chosen to chiefly back up 3.0 alone.

What i miss about momoko's jme3 idea is some kind of roadmap and more input from other experienced people, it seems to be a bit of a one man show, and i just think its too much for one person to create a whole new engine in a resonable timeframe.
Well that's exactly what we're trying to address here! :D The roadmap will show up as soon as we start seeing interest in 3.0 on a larger scale. Probably the prime reason why me and MomokoFan chose to go with this line of action is because he quickly realized that, like you say, it would be impossible for him to uphold the standards for a professional grade engine all by himself, which is the end goal we're looking at here.
8

i'm massively in favour of the momoko's jme 3 branch and have started looking at porting the scene worker project to it with specific regards to scene workers shader editor…

but if we had a road map, feature list and a tentative time frame it would enable people like me who are not core developers to figure out when we can best allocate our time to moving our existing code bases over to the jme3 branch…

which is something i am very very eager to do but i don't want to drop working with jme 2 until i know when jme3 will be up and running, even from an alpha perspective as i'd like to have my editor pegged to the same release schedule…

also the name gorilla 3d is great  :slight_smile:

9
ncomp said:

also the name gorilla 3d is great  :)


thats the only thing i have to disagree with. While the hint for a "bigger and stronger monkey" may be funny, a name change will just confuse people.
10
ncomp said:

i'm massively in favour of the momoko's jme 3 branch and have started looking at porting the scene worker project to it with specific regards to scene workers shader editor....
but if we had a road map, feature list and a tentative time frame it would enable people like me who are not core developers to figure out when we can best allocate our time to moving our existing code bases over to the jme3 branch...
which is something i am very very eager to do but i don't want to drop working with jme 2 until i know when jme3 will be up and running, even from an alpha perspective as i'd like to have my editor pegged to the same release schedule...
also the name gorilla 3d is great  :)
This type of feedback means a lot ncomp. We suspected a lot of people were feeling this way, that eagerness to move on, hence this attempt to gain a bit more privileges in the project so that we get the chance to properly show that jME 3.0 is charging on strong and welcomes anyone eager to jump on the wagon.

And hey, Gorilla3D is all momoko! :P I believe he came up with it while we were still pondering the idea of starting a new community from scratch (oh boy) in case we failed to get the support of the jME community. I suggested the ever so awesume name Silverback3D. eh? EH!?

Anyhow, @dhdd, to address your concern: Though a name change is not completely out of the question for the faaar off future, changing jME's name right now would, like you said, only confuse people, and it would be bad for another multitude of reasons such as broken bookmarks and mixed results in search engines.
11

Hi!


Sadr said:

So what say you? Would you be willing to give me and MomokoFan your vote of confidence, trusting us to step up to the part of 'administrators', to be in charge of the development of jME's next generation, the 3.0 branch?

I don't trust people who say that JMonkeyEngine 2.0 is an engine of choice whereas its JOGL/AWT side is not yet usable in production phase. I would like to get much help in the future. I have been working on some new fixes with a freelancer (he made the biggest part of the job as I have to work on the cells-and-portals algorithm), I will deliver them as soon as possible but I would have preferred to find volunteer people to do it with me. If you prioritize the development of JMonkeyEngine 3 with LWJGL in mind, I won't trust you. The OpenGL layer must be abstracted correctly and adapting the JOGL/AWT side whereas the design of the engine is centered on LWJGL is not the right way to do except if you like spending months and money to fix the resulting bugs of such a naive approach (lol maybe you don't feel concerned). Sorry to be harsh. The NASA uses JOGL, some other corporations too. I would be sad to use Ardor3D, Aviatrix3D or Xith3D at work because of a lack of investment in JOGL for JMonkeyEngine.
12

http://jerome.jouvie.free.fr/OpenGl/Projects/GraphicEngine.php

13
darkfrog said:

http://jerome.jouvie.free.fr/OpenGl/Projects/GraphicEngine.php

Thanks, I'm looking at the source code.
14
ncomp said:

i'm massively in favour of the momoko's jme 3 branch and have started looking at porting the scene worker project to it with specific regards to scene workers shader editor....

Glad to hear that, but do note that core components may change without notice, breaking your code. I would suggest waiting until the engine becomes more concrete.

ncomp said:

but if we had a road map, feature list and a tentative time frame it would enable people like me who are not core developers to figure out when we can best allocate our time to moving our existing code bases over to the jme3 branch...

I already have a roadmap which I use for development, but since many people requested it I will try to make it available on the wiki.

gouessej said:
The OpenGL layer must be abstracted correctly and adapting the JOGL/AWT side whereas the design of the engine is centered on LWJGL is not the right way to do except if you like spending months and money to fix the resulting bugs of such a naive approach (lol maybe you don't feel concerned).

That might not be necessary, I have made some changes that gave the context and renderer control over the render thread and environment. This makes it possible to implement LWJGL's AWTGLCanvas as well as the JOGL standard of handling the render loop.
In fact the display layer of JOGL is already complete:

15

Maybe admin/global admin could send out an announcement about this thread for increased exposure? I imagine most people go to the development boards before checking out 'Site'. I'd say we need feedback from at least 10-15 more people before any final calls can be made.

16
Momoko_Fan said:

That might not be necessary, I have made some changes that gave the context and renderer control over the render thread and environment. This makes it possible to implement LWJGL's AWTGLCanvas as well as the JOGL standard of handling the render loop.
In fact the display layer of JOGL is already complete:


Is it a joke?
17
Momoko_Fan said:

That might not be necessary, I have made some changes that gave the context and renderer control over the render thread and environment. This makes it possible to implement LWJGL's AWTGLCanvas as well as the JOGL standard of handling the render loop.

Shouldn't we be using the faster Display.setParent() method for LWJGL canva's instead of the old AWTGLCanvas?
See here: http://www.jmonkeyengine.com/jmeforum/index.php?topic=10569.0
18
gouessej said:

Momoko_Fan said:

That might not be necessary, I have made some changes that gave the context and renderer control over the render thread and environment. This makes it possible to implement LWJGL's AWTGLCanvas as well as the JOGL standard of handling the render loop.
In fact the display layer of JOGL is already complete:


Is it a joke?

No its not. Do note that only the display layer is implemented, that includes creation and destruction of the GL display. The renderer and input portions are nearly non-existent, hence why the screenshot shows a black screen. Once the JOGL renderer has been implemented fully, then you should see the jMonkeyEngine logo shown instead of a black screen in that test program.

If you want to see it for yourself, I checked in the changes just now:
http://code.google.com/p/jmonkeyengine/source/browse/#svn/branches/mf_jme3test/src/com/g3d/system/jogl

SomethingNew said:

Shouldn't we be using the faster Display.setParent() method for LWJGL canva's instead of the old AWTGLCanvas?
See here: http://www.jmonkeyengine.com/jmeforum/index.php?topic=10569.0

Yes I know, AWTGLCanvas can still be supported though, if it is needed.
19

this is pretty much my view on the issue won't add more to it… yet

http://www.jmonkeyengine.com/jmeforum/index.php?topic=11290.msg85041#msg85041

20
Alric said:

I have no wish to hamper your efforts, but there I have spoken my mind as requested.
And it is appreciated ;)

I haven't looked at the code, but clearly momoko_fan is one of the community members capable of moving the engine forward.

I would be concerned that a branch with only a single developer could quite easily end up running out of steam.
Similar concerns have been addressed here already.

I would also be concerned that Sadr you have not been a consistent member of the jME community. I have the impression that things build on jME are your primary interest, rather than the core engine.
Of course things that's built with jME is my primary interest; at the core I'm a game developer after all :) That doesn't mean I would be any less suitable as an assisting administrator of this project, because there are more sides to the matter than just developing this engine. jME lives on to this day because of essential components such as its community, larger projects that depend on it, smaller projects that ensure a steady flow of promotion (backlinks & mentions etc.) and a surge of new members who keep challenging the general level of "accessibility" of the whole project. There's enough love and care going around for the core engine so that some of us can dedicate ourselves to other essentials of the project.

I wouldn't dream of being the "main guy" in management for jME really, I wouldn't have a broad enough spectrum of interest and abilities for that. But right now I'm offering to be there for jME when a new branch is established, because building teams I would say is my area of expertise.


mcbeth said:

this is pretty much my view on the issue won't add more to it..... yet
http://www.jmonkeyengine.com/jmeforum/index.php?topic=11290.msg85041#msg85041
I can't read that, so I'm assuming that's in the moderator's board. Maybe if the thread doesn't actually touch sensitive topics (more so than this one) then it could be moved into the public area?