@erlend_sh said:
As we discussed in the chat, your implementation is problematic. We should have started discussing the technical details a long time ago. You decided to do things your way, which is fine, but you can't go against best practices and at the same time expect to have your work included in the SDK just because people express an interest in it.
The best course of action right now would be to just release the code out in the open and see where that takes us.
I did it.
if you would like you can write some news about this event.
@mifth said:
Sorry, man. It would be a diificult task for me to dig into the scenecomposer.. As my skills in multythreading and other stuff are weak.
I wrote the editor during 3 weeks from skratch. I think you can adapt my code to SDK during 2-3 days really.
Cheap excuses. I explained to you that you just need to implement a widget interface in the SDK and not have to care about threading. I was even offering to help you with any code you come up with. So @all, please bug @mifth and not me, right now he's offloading the work to poor @nehon -.-
The content of this post is meant to be read as a straight information or question without an implicit dismissive stance or interest in having the other party feel offended unless there's emotes that hint otherwise or there's an increased use of exclamation marks and all-capital words.
And properties panel is missedā¦to adjust position or scale for example ā¦mouse operations are convenient but not really precise.
ā¦If it goes into the SDKā¦the problem will be solved by itself.
And properties panel is missedā¦to adjust position or scale for example ā¦mouse operations are convenient but not really precise.
ā¦If it goes into the SDKā¦the problem will be solved by itself.
Empty layer can be needed for parsingā¦
You can precise with 1/5/10 units (on the left panel). And you can easily move to 10 units or rotate to 90 degrees.
if you want to contribute a properties panel it will be welcome. As my editor is āSimpleā.
Not really understandā¦if i only fill the 2 first layersā¦in what way i need the others?..they are emptyā¦i only want my 2 filled layers in the j3oā¦others are useless in jme3 no?..empty node?
maybe iāve missed something ;).
Hum, yeah youāre right this can do the trick :).
ā¦all depend on the future of the softā¦if it will be merged in the SDKā¦no need to ;).
@haze said:
Not really understand...if i only fill the 2 first layers...in what way i need the others?...they are empty...i only want my 2 filled layers in the j3o...others are useless in jme3 no?..empty node?
maybe i've missed something ;).
Hum, yeah youāre right this can do the trick :).
ā¦all depend on the future of the softā¦if it will be merged in the SDKā¦no need to ;).
Well, 19 empty layers is not a big problem.
Iām interested in this too.
I hope @nehon will make magic.
@mifth said:
Well, 19 empty layers is not a big problem. :)
Iām interested in this too.
I hope @nehon will make magic.
Yeah, no drama, i simply asked if it had a reason to ^^. Why donāt you simply check if the layer rootNode has child or not before export it?
ā¦yeah @nehon is great but this is a not planned work for him, and he has some occupation for now with shaders/material editorā¦so donāt hope too for now.
@haze said:
Yeah, no drama, i simply asked if it had a reason to ^^. Why don't you simply check if the layer rootNode has child or not before export it?
ā¦yeah @nehon is great but this is a not planned work for him, and he has some occupation for now with shaders/material editorā¦so donāt hope too for now.
Indeed, I canāt promise anything, got already much to do, bug fixes, new features for the shader nodes and so onā¦ Also have a couple of personal projectsā¦and a family to take care of:p
But Iāll look into it eventually. The features Iām interested in are the nice transform widgets, and maybe the grid.
Maybe i misunderstood, but the layers should not be based on the scene graph. Usually you want an horizontal approach for layers rather than a hierarchical. Also since the layer concept does not exist in JME, I would save this information in a separate meta data file instead of in the j3o.
@nehon said:
Indeed, I can't promise anything, got already much to do, bug fixes, new features for the shader nodes and so on.... Also have a couple of personal projects....and a family to take care of:p
But I'll look into it eventually. The features I'm interested in are the nice transform widgets, and maybe the grid.
Maybe i misunderstood, but the layers should not be based on the scene graph. Usually you want an horizontal approach for layers rather than a hierarchical. Also since the layer concept does not exist in JME, I would save this information in a separate meta data file instead of in the j3o.
Ok, good. Donāt forget about good selection tools. Try them. Especially rectangle selection tool (radiobutton on the left panel).
Hi @Mimth i tried your editor on win 7 and found out that it freezes when i try to save a scene. Is this known issue or something with my machine only?
<cite>@aniket0 said:</cite>
Hi @Mimth i tried your editor on win 7 and found out that it freezes when i try to save a scene. Is this known issue or something with my machine only?
Hello! Thank you for testing!
I tested it only on linux. Can you run the project from SDK and send me the log? Iāll try to fix it.
<cite>@mifth said:</cite>
Hello! Thank you for testing!
I tested it only on linux. Can you run the project from SDK and send me the log? I'll try to fix it.
Iām back and decided to continue developing the SWE. I know there is no big interest for the SWE but it can be useful for some people. So, I moved entire code to a separate project https://code.google.com/p/simple-world-editor/
Aims of SWE:
Create cool scenes for small game projects
Simple interface
Useful tools for level design
Possibility to create grid-based levels
Try to run SWE on Android (it can be useful for someone in future)
SWE binaries will be here:
SWE 0.13(Done):
Removed SWS file. Now only one SWE file is used.
fixes for loading/saving assets
Constarint value 0,5 is added
SceneList is now used only for custom selection
Various fixes and optimizations
SWE 0.2(Future 2013 RoadMap):
Redesign Interface (make it more cool)
Grid Menu
Lights - add/move/rotate.
Camera Views Buttons (top, front, left etcā¦)
Shortcuts description button
Classes refactoring
Remove System.out.println().
Object Parameters Menu (values of transform)
Make unlimited number of assets
Lock Layers feature
JSON formatting
Change Background Color
Change Camera settings
If model is not found - make red cubes
SWE 0.3 (Future 2014 RoadMap):
make multiple scenes in one file
merge 2 scenes into one scene
Groups feature
Group Objects feature
Add different objects with the same name
Shaders, Materials, Shaders
Adapt project for Android
load external model (not from assets)
Hide Objects feature
Collision Objects - creating/viewing/editing
Blend, Ogre, Obj models loading
Custom model format (SWM). To save external models.
this dark theme make me feel depressive and angry in the same time, as i really hate blender with all my soul. It seems that blender find always a new way to donāt do what i need, to hide precious informations and donāt react as expected. @mylife : When i found in the wiki sentences like āyou can do that with the EASY shortcut ctrl+alt+shift+fā (ā¦) + blender crashed when i went in user preference menu + blender collada import (from SU) is messed up + face normal is messed up (i can give you files, youāll see yourself) + the rotation pivot for the camera is not what we have under our mouse, so the camera just fly away all the time + ā¦ well, a lot of other things. I never had stomach ach before, i always have when i try to use blender.
i think that people use a fix computer (or a mobile one, but i mean : not a mobile phone) to develop their application, even if the target are mobile phones. So, why you use nifty ? I think that for editor, you should use swing (or something equivalent) that adapt its appearance to the look&feel of the os (not sure if i am clear). Nifty is ā¦ a bit cool, but when i look at some nifty code, i just feel like i am back in early 90ās with all the horrors of pre-css html and the mix of the presentation and the content.
I think that nifty is like a peg leg : we needed (and still need) some way to do gui, and nifty can do something ā¦ not too bad. But at the very end, itās just like reinventing the wheel (the wheel being html) and like every time someone reinvent the wheel, he gets something which is more square than round.
i didnāt see the āfallā option in your gui. In morrowind editor, you have an option like this, wich put the selected object āon the groundā (it performs a ray cast). You could have something like that, but with the physics engine. To be explicit, steps in my mind are:
-add a rigidbody to the object
-perform the simulation for this object (maybe with higher speed than normal)
-/!\ at every frame, ensure that the object is still on the vertical line from the original point : we want it at this exact xz location, only the y should change. This could be an activable/deactivable option, however.
-at the end of the simulation (when the object is still or then the user stop it manually), ask to the user if he wants to keep the result. This is not needed if you have a ācancelā option in your editor (and you should ^^).
-clean everything (remove the rigidbody etc. Well, you know what ācleanā means ^^).
Why i ask for this option ? Well, itās always hard to place object precisely on the ground, and only some game are in space. Things in space are a lot easier to do.
@bubuche said:
I will maybe appear to be a douch but :
.....
Blender is one of the best opensource projects. If you don't like it i can't help you.
Itās much faster to develop the interface with Nifty than with Swing. Also, Swing has some issues according to multythreadng. I think in 3-5 years you will use android on desktops easily.
Physics is in my todo list.
You can easily place objects with manipulator. Or you can use āClearLocationā button and then move X and Z axys.
I think iāll add āDropToGroundā button in future.
