Replies to this topic | |
|
|
rollin | Thu Aug-26-10 07:52 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#1. "RE: Source Code Work"
In response to Reply # 0
|
also
GAME
- able to create & use mech configuration files independently from singleplayer or multiplayer
|
|
|
|
|
|
rollin | Wed Sep-01-10 06:19 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#2. "RE: Source Code Work"
In response to Reply # 0
|
|
|
|   |
Magic | Fri Nov-26-10 11:06 AM |
Charter member
647 posts
|    |
|
#19. "RE: Source Code Work"
In response to Reply # 2
Fri Nov-26-10 11:15 AM by Magic
|
rollin, thank you your message box:
I entered the code to my 2.5 with some modifications. No need to pause the game and press enter. It will display the message for a 10 (or more) sec and close after that period of time.
Here is the code:
File controlGUI.cpp. ******************************************************************** // set the message text - mesage box by rollin modified by magic 26112010 /* by mgto */ void ControlGui::setInstantMessageText( const char* getString) {
instantStartTime = scenarioTime; //char tmp2<64>; //sprintf( tmp2, "startTime: %ld", instantStartTime ); //MessageBox(NULL,tmp2,"TEST",MB_OK);
instantEndTime = instantStartTime + 10;
renderInstantMessageText (getString, instantEndTime);
// set message state i_dispMsg = true;
}
******************************************************************** // render the message text - mesage box by rollin modified by magic 26112010 /* by mgto */ void ControlGui::renderInstantMessageText (const char* getString, int endTime) { // change text if getString is not empty if (getString != "" ) i_text = getString;
// if there is some text to display: pause game, draw message box and display text if (i_text != "") { //___________________________ // DRAW MESSAGE WINDOW
unsigned long itw_right, itw_left, itw_top, itw_width, itw_height; itw_right =10; itw_top = 10; itw_width = 400; itw_height = 200; itw_left = Environment.screenWidth - itw_right - itw_width;
unsigned long txt_width, txt_height; txt_height = helpFont.height(); txt_width = helpFont.width( i_text );
GUI_RECT rect = { itw_left, // left itw_top, // top itw_left + itw_width, //right itw_top + itw_height //bottom }; drawRect( rect, 0x44000000 );
//___________________________ // DRAW TEXT
drawShadowText( 0xff41a0f0, 0xff000000, missionResultsFont.getTempHandle(), itw_left+10, itw_top+5, true, "Message Window", false, missionResultsFont.getSize() );
//drawShadowText( 0xffffc600, 0xff000000, guiFont.getTempHandle(), itw_left+itw_width/3, itw_top+itw_height-15, true, "Press Enter to Continue", false, guiFont.getSize() );
// void drawShadowText( long colorTop, long colorShadow, HGOSFONT3D font, long left, long top, long right, long bottom, bool proportional, const char* text, bool bold, float scale, long xOffset, long yOffset, int align) drawShadowText( 0xffffffff, 0xff000000, helpFont.getTempHandle(), itw_left+10 , itw_top+40, itw_left+itw_width, itw_top+itw_height-20, true, i_text, true, helpFont.getSize(), 1, 1, 0 ); }
//___________________________ // CONTINUE
long expTime = scenarioTime;
if ((expTime > endTime) && i_dispMsg) { //change i_text to nothing i_text = ""; // set message state i_dispMsg = false; }
} ****************************************************************** void ControlGui::render( bool bPaused ) if (drawGUIOn) { /* by mgto */ if ( !bPaused && !MPlayer ) // show only in singleplayer and when pausemenu is not displayed
renderInstantMessageText("",instantEndTime); }
******************************************************************* File controlGUI.h
public
void setInstantMessageText( const char* getString);
private
void renderInstantMessageText (const char* getString, int endTime); long instantStartTime; long instantEndTime;
Can I use this modified code in my 2.5? Time can be entered to setInstantMessageText like this
setInstantMessageText( const char* getString, int timeToDisplay);?
|
|
|
|
|     |
|
rollin | Fri Nov-26-10 11:37 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#20. "RE: Source Code Work"
In response to Reply # 19
|
Starman01: don't see how this could be usefull ..
magic: thx for sharing the code!
I and Wolfman are already using a furhter developed code (not mutch) where you are also able to set the style and size with fit files
Imo the pause is very important.
If you want to display story text you should let the player decide how long the needs to read the text
I'll need to update the code on my page
|
|
|
|
|         |
|
rollin | Fri Nov-26-10 12:28 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#22. "RE: Source Code Work"
In response to Reply # 21
|
Sorry Magic, I was in a hurry between work and stuff that I missed your question. Of course you may use the code in any edited form. That's why I released it in the first place!
|
|
|
|
|
|
rollin | Wed Oct-06-10 12:25 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#3. "RE: Source Code Work"
In response to Reply # 0
|
to fix the editor resize bug look into the editorinterface.cpp and change the code to this
search for the line "case WM_SIZE:" and change the case with the following code
case WM_SIZE: { float w = LOWORD(lParam); float h = HIWORD(lParam); /* disabled by mgto */ //windowSizeChanged = true; /* by mgto */ //MessageBox("wm_size"); // use this for debugging if (w != g_newWidth && h != g_newHeight){
if ( (16 < (w - g_newWidth)) || (16 < (h - g_newHeight)) ){ // make bigger windowSizeChanged = true; } if ( (-w + g_newWidth) > 20 || (-h + g_newHeight) > 20 ){ // make smaller windowSizeChanged = true; }
if (windowSizeChanged){ g_newWidth = w + 16; //scrollbar thickness g_newHeight = h + 16; //scrollbar thickness } }
/* by mgto */
/* disabled by mgto */ //g_newWidth = w + 16/*scrollbar thickness*/; /* disabled by mgto */ //g_newHeight = h + 16/*scrollbar thickness*/;
retVal = GameOSWinProc( m_hWnd,message,wParam,lParam ); if (editor && bThisIsInitialized) { editor->update(); } //EditorObjectMgr::instance()->update(); } break; }
|
|
|
|
|
|
rollin | Wed Oct-06-10 12:29 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#4. "RE: Source Code Work"
In response to Reply # 0
|
magic or anyone else who knows: how to play a sound from one of the mclib cpps ? (i guess it's mclib folder.. the folder where mech3d.cpp is in)
normally everything works from the code folder using the gamesound.cpp (which is also in the code folder but working as a subclass of soundsys)
but I'm not able to create an instance to the soundsys class (which is referenced in mclib) nor reference gemsound.h into the mclib stuff
I don't understand the c++ organisation well enough to get this going
|
|
|
|
|     |
|
rollin | Sun Oct-10-10 11:36 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#7. "RE: Source Code Work"
In response to Reply # 6
|
hi!
1. I can play sounds form files in the 'code' directory 2. There is no code I can find under mclib that is actually using a sound system. the bink code has something in it but I'm quite sure the bink code is not compiled any more
I can not use the normal way to play sounds 'cause I'm not able to use soundsys (can't get an Instance of the class to work) nor am I able to reference the gemesoundsystem which is under 'code' and not included in the mclib sub-vproject. If I inlclude it I get the error that aFont can't be initialized but I'm not able to fix that
|
|
|
|
|       |
|
TheOldMan | Mon Oct-11-10 10:03 AM |
Member since Jun 16th 2010
28 posts
|    |
|
#8. "RE: Source Code Work"
In response to Reply # 7
|
Correct me if I'm wrong (wouldn't be the first time). I'm new to C++ and don't know what's available through the compiler, and what has to be coded.
MCLib is part of the game Library? Primarily setting up access to existing C++ background functions like the sound system, so they can be called later. If that is the case, you wouldn't be able to access SoundSys, because it technically doesn't exist until it is compiled. Library functions have that tendency.
Where as Code accesses (includes) MCLib, and uses the defined functions. Therefore, you now have access to SoundSys.
The coding setup was most likely never intended to play any sounds from within MCLib, just make sound available.
Is there anyway you can accomplish what you are trying to do in the Code section? Which brings up the question, what exactly are you trying to do? I understand you're trying to play a sound file, but for what purpose?
A bit more information might help others point you in the right direction.
Which all brings me around to one last question...
Bink Code? You found Bink Code? Where is it? What filename?
Okay, that was actually 4 questions, but I'm interested in restoring Video capability to the game. It's not a high priority item with anyone else, but I think the game has lost alot of potential without it.
|
|
|
|
|         |
|
rollin | Tue Oct-12-10 06:31 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#10. "RE: Source Code Work"
In response to Reply # 8
|
yes, that it's part of a library could be true. Instancing a Class only after compiling could make sense but I'm not sure it's really true here
anyway.. the bink stuff (which is nothing more as a little bit code not doing the real video stuff) is called something like ...bink.cpp or ..bnk.cpp also within the mclib folder
I want to play a sound every time a mech foot touches the ground
the necessary information is within the mech3d.cpp but as it is part of mclib I need to get a soundsystem there. I see no easy way to transfer the mech-foot-touch-ground information into the 'code' section
and wolfman: you would also need someone who knows a bit about implementing a video codec not only handling c++ code. But as people are normally not able to create some quality videos I wouldn't call this a really big value to the community .. multiplayer would be better imo
|
|
|
|
|           | |
|             |
|
rollin | Tue Oct-12-10 10:37 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#12. "RE: Source Code Work"
In response to Reply # 11
|
the filename is : mclib/mc2bink.cpp
and the class I would love to call is 'SoundSystem' and there is already an instance of it called 'sndSystem' but it's used by /gui/ or /viewer/
as said mc2bink.cpp is using the soundSystem instance of the GgameSoundSystem class I guess. but mc2bink is not used anymore.. but it makes me wonder if it IS possible to get the sounds active from the mclib folder
and yes I'm aware how a bad foot step sound implementation would sound like .. and I also know there are some old files in the source showing that they where planing to implement these sounds but decided against it.. bc of lac of time or other reasons but I'm sure I'm able to create a good foot step sound implementation if I'm somehow able to get these sounds working
|
|
|
|
|               |
|
TheOldMan | Wed Oct-13-10 10:57 AM |
Member since Jun 16th 2010
28 posts
|    |
|
#14. "RE: Source Code Work"
In response to Reply # 12
|
No such file, I got one called MC2Movie in the Code. You may have one of the first downloads of the source, before all the Bink stuff was cleaned out. I got mine just a couple of months ago.
Anyway, back to your problem... Mech3d looks like it's setting up the animation frames. It may not be the best place for your Mod.
We can look at this logically.
Problem: You need to add sound based on a specific action/animation situation. Walk or Run.
So, other questions to figure out... What calls Mech3d? What other process has a sound associated with an animation action? Is there another direction to work from?
Possible solution: Find a similar process, and see how that works. Look at the process for Jump. It uses special animation, and has a specific sound associated with it. (Which only works after the first jump.} It also has several functions for it in Mech3d, but no calls to play an SFX in those functions.
Where is the sound process executed? SFX sound = JUMPJETS = 76
There has to be something that says process a jump action, and play this SFX with it. Check the GUI process for the Jump Action.
That's your way in. Start with the mouse click for jump. Follow it to when and where it plays a sound. Apply the same or similar coding to walk and run. Walk should be easy, since all mechs walk at the same speed. Run will be more difficult since each mech chassis has its own speed. You might actually find most of the code already there commented out, or a reference to what was planned.
Also check on how SFX Sound Channels are set up. There are a set number of channels and a priority system, so if there are too many things to be played at the same time, high priority gets played and low gets trashed. Set your priority appropriately.
Hope this helps.
|
|
|
|
|           |
|
wolfman | Tue Oct-12-10 12:04 PM |
Member since Apr 15th 2003
351 posts
|    |
|
#13. "RE: Source Code Work"
In response to Reply # 10
Tue Oct-12-10 12:04 PM by wolfman
|
yes, that it's part of a library could be true. Instancing a Class only after compiling could make sense but I'm not sure it's really true here
anyway.. the bink stuff (which is nothing more as a little bit code not doing the real video stuff) is called something like ...bink.cpp or ..bnk.cpp also within the mclib folder
I want to play a sound every time a mech foot touches the ground
the necessary information is within the mech3d.cpp but as it is part of mclib I need to get a soundsystem there. I see no easy way to transfer the mech-foot-touch-ground information into the 'code' section
and wolfman: you would also need someone who knows a bit about implementing a video codec not only handling c++ code. But as people are normally not able to create some quality videos I wouldn't call this a really big value to the community .. multiplayer would be better imo
You know that I believe video and multiplayer are the two big things missing and a lot af great campaigns had video in it. I totally disagree with your judgement on the skill people have to do this. The German campaign "Operation Fimbulwinter" is almost video only with a great story and actors. I can live without foodstep sound but understand the challenge to implement it and the good feeling once it work check out wolfman's missions and wolfman's forum at http://s2.invisionfree.com/wolfman_x_forum/index.php
|
|
|
|
|   |
|
rollin | Thu Nov-25-10 09:10 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#16. "RE: Source Code Work"
In response to Reply # 15
|
that was a good idea!
unfortunately the bounds are only creating a 2d rectangle. Not sure for what that was or is used. Changing the values doesn't seem to do anything 0.o
|
|
|
|
|       |
|
wolfman | Thu Nov-25-10 04:08 PM |
Member since Apr 15th 2003
351 posts
|    |
|
#18. "RE: Source Code Work"
In response to Reply # 17
|
Hm, maybe it's disabled in the code ? That would explain why they didn't bother to change the values. Karl found this in the code, maybe it's related and you can add the z-axis : // convert to block and vertex
long tileCol;
long tileRow;
land->worldToTile( data->vector, tileRow, tileCol );
int blockI = tileCol/20;
int blockJ = tileRow/20;
data->blockNumber = blockJ * land->blocksMapSide + blockI;
tileCol -= blockI * 20;
tileRow -= blockJ * 20;
data->vertexNumber = tileRow * 20 + tileRow;
countObject(data);
data++;
}
I just don't know where he exactly found it. @Karl : In which file was this located ?
It can be found in objmgr.cppcheck out wolfman's missions and wolfman's forum at http://s2.invisionfree.com/wolfman_x_forum/index.php
|
|
|
|
|           |
|
rollin | Mon Nov-29-10 06:08 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#24. "RE: Source Code Work"
In response to Reply # 23
|
depends.. In which file did you find it?
Looks interesting indeed!
|
|
|
|
|             | |
|               |
|
rollin | Mon Nov-29-10 06:28 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#28. "RE: Source Code Work"
In response to Reply # 25
|
Magic, I've updated my message window code here
http://blog.maginot.eu/index.php?p=till.5__mc2_code
The biggest change is the implementation of the buttonLayout(...).fit Files to set the messagebox size and colors for every resolution in a proper way
It should be easy to implement the changes into your version of the code.
Note: only controlGui.h and .cpp have changed and you need to add the new block to each buttonLayout(...).fit file
|
|
|
|
|                 | |
|                   |
Magic | Tue Nov-30-10 08:26 AM |
Charter member
647 posts
|    |
|
#34. "RE: Source Code Work"
In response to Reply # 29
|
oh and something else could you implement the following bug fix?! it's only a very tiny visual bug of the mech torso rotation. I'm quite sure nobody else notice it but me it drives crazy  mech.cpp old line if ((torsoRelFacing < -fireArc) || (torsoRelFacing > fireArc) ) new line if ( (torsoRelFacing < -fireArc) || (torsoRelFacing > fireArc) || ( (torsoRotation <= fireArc) && (torsoRotation >= -fireArc) ) )
OK will do, and will post with my latest release, possibly tomorrow.
|
|
|
|
|                 |
Magic | Tue Nov-30-10 08:24 AM |
Charter member
647 posts
|    |
|
#33. "RE: Source Code Work"
In response to Reply # 28
|
Magic, I've updated my message window code here http://blog.maginot.eu/index.php?p=till.5__mc2_codeThe biggest change is the implementation of the buttonLayout(...).fit Files to set the messagebox size and colors for every resolution in a proper way It should be easy to implement the changes into your version of the code. Note: only controlGui.h and .cpp have changed and you need to add the new block to each buttonLayout(...).fit file
OK but why doing so, it can be done in the code by checking resolution and switch.
|
|
|
|
|                   |
|
rollin | Tue Nov-30-10 09:53 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#36. "RE: Source Code Work"
In response to Reply # 33
|
it's simple for others to change the look without working with the source code. You are the code right now.. but you also have to think about those who want to mod your mod (mission builders like me for example)
Magic, I've updated my message window code here http://blog.maginot.eu/index.php?p=till.5__mc2_codeThe biggest change is the implementation of the buttonLayout(...).fit Files to set the messagebox size and colors for every resolution in a proper way It should be easy to implement the changes into your version of the code. Note: only controlGui.h and .cpp have changed and you need to add the new block to each buttonLayout(...).fit file OK but why doing so, it can be done in the code by checking resolution and switch.
|
|
|
|
|                     | |
|                   |
|
rollin | Tue Nov-30-10 09:53 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#37. "RE: Source Code Work"
In response to Reply # 33
Tue Nov-30-10 09:55 AM by rollin
|
THAT is great 
Already did that except I left color at the same level and used its 30000 for polys (70000). Tested on a large map with big forests - no object disappearing, but again my maps are not that complex.
Wolfman, you have bigger and more complex maps, have you tried this? Plus directly above there is something like this 40 * 1024 * 1024 (I upgraded my comp yesterday and Im still in data transfer, my studio 2008 is not yet up, I cant look in the code), make it 80 * 1024 * 1024.
Its in my latest update (upload - tomorrow).
|
|
|
|
|                 |
Magic | Thu Dec-16-10 08:23 AM |
Charter member
647 posts
|    |
|
#65. "RE: Source Code Work"
In response to Reply # 28
|
Magic, I've updated my message window code here http://blog.maginot.eu/index.php?p=till.5__mc2_codeThe biggest change is the implementation of the buttonLayout(...).fit Files to set the messagebox size and colors for every resolution in a proper way It should be easy to implement the changes into your version of the code. Note: only controlGui.h and .cpp have changed and you need to add the new block to each buttonLayout(...).fit file
rollin, since you work with wolfman and I dont plan to work on the editor can we agree about using the same code for your message box. Will you agree to use hybrid version because if we use different versions missions created in wolfmans editor may not be compatible (portable to 2.5)? I freed mechs and vehicles and I will free everything else of object2, so in theory missions created in wolfmans editor should be playable in mc2.5
My old missions are playable even without any change in mission fit files (with new obj number for mechs and vehicles).
I plan to change my corebrain.abx to be compatible with the original and wolfmans corebrain.
|
|
|
|
|                   |
|
rollin | Fri Dec-17-10 06:14 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#66. "RE: Source Code Work"
In response to Reply # 65
|
I have nothing to say in this case. It's wolfmans thing and not mine.
I'm just helping here and there with code stuff
What you or wolfman do with the code, if you use the same or not, is something both of you have to communicate.
I don't have time to do or look into anything atm anyway.
|
|
|
|
|                     | |
|             | |
|               |
|
rollin | Mon Nov-29-10 06:12 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#27. "RE: Source Code Work"
In response to Reply # 26
Mon Nov-29-10 06:12 PM by rollin
|
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Karl you are close to beeing the hero of the day..
Magic and Wolfman (when you're back)
open mission.cpp (not mission2 or anything else yet) and double all Pool Values
(stuff is around line 3075)
It does work here but I've already changed some values in the mid level renderer (MLR) so please do only this and nothing else and test it with a very crowded map (big forest or something like that)
I'm not saying we have it.. but we are certainly very very close!
Great find Karl!
|
|
|
|
|                 | |
|                   | |
|                 | |
|                 | |
|                   |
|
wolfman | Tue Nov-30-10 10:52 AM |
Member since Apr 15th 2003
351 posts
|    |
|
#39. "RE: Source Code Work"
In response to Reply # 35
|
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Karl you are close to beeing the hero of the day..
Magic and Wolfman (when you're back)
open mission.cpp (not mission2 or anything else yet) and double all Pool Values
(stuff is around line 3075)
It does work here but I've already changed some values in the mid level renderer (MLR) so please do only this and nothing else and test it with a very crowded map (big forest or something like that)
I'm not saying we have it.. but we are certainly very very close!
Great find Karl!
Already did that except I left color at the same level and used its 30000 for polys (70000). Tested on a large map with big forests - no object disappearing, but again my maps are not that complex. Wolfman, you have bigger and more complex maps, have you tried this? Plus directly above there is something like this 40 * 1024 * 1024 (I upgraded my comp yesterday and Im still in data transfer, my studio 2008 is not yet up, I cant look in the code), make it 80 * 1024 * 1024. Its in my latest update (upload - tomorrow).
I double all the pool values in mission and editor.cpp and can convince myself it is better or it has not chnged on dense forest maps. Need to make amore objective testing. This seems to be more psychology and the wish driving perseption check out wolfman's missions and wolfman's forum at http://s2.invisionfree.com/wolfman_x_forum/index.php
|
|
|
|
|                     | |
|                     |
|
rollin | Tue Nov-30-10 05:26 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#41. "RE: Source Code Work"
In response to Reply # 39
|
There are some settings with the mid level renderer that might have an influence on what karl posted
the first place can be found in mlr.hpp around line 117
something like Max_Number_Vertices_Per_Frame = ... and so on
this is "just" a place that initializes basic values that can be changed on function call anyway but as this function seems to be not always called with all arguments this is not that unimportant.
so for the first test it would be interesting what happens if you double the values here (make sure they can be divided by 2 like 1024, 2048 ...)
the 2nd place is in mechcmd2.cpp seach for "MidLevelRenderer::InitializeClasses(
2nd test would be to change values here but not in mlr.hpp
3rd test would be to change values in both files
here is my current state where I see now a lot more objects on screen
in mlr.hpp i use still some old values for testing which are way higher the the defaults: vertices : 16384 primitives : 16384 the others values are the default ones
now in mechcmd2 i use also some old test values but this time lower ones! (512, 512, 0,0,true)
that's all I can tell for the moment
|
|
|
|
|                       | |
|                         |
|
wolfman | Wed Dec-01-10 11:46 AM |
Member since Apr 15th 2003
351 posts
|    |
|
#44. "RE: Source Code Work"
In response to Reply # 42
|
|
|
|                       | |
|                         |
|
rollin | Wed Dec-01-10 04:30 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#45. "RE: Source Code Work"
In response to Reply # 43
|
is it possible to read the settings for the pool stuff out of the settings or options cfg? OK, of course it's possible but how much effort?
Would be cool as soon as we know for sure what's necessary to increase the models per screen.
Wolfman says he's not sure if doubling the values alone did something significant
|
|
|
|
|                           | |
|                             |
|
rollin | Thu Dec-02-10 10:56 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#47. "RE: Source Code Work"
In response to Reply # 46
|
ok, so far we are not sure what the MLR is doing (bc I also didn't saw any real change)
but changing the values in mission.cpp does increase the amount of objects one can see (I tested it and magic tested it)
Magic, when are you releasing your latest build?
And I hope we can get this into MC2X too
|
|
|
|
|                               |
Magic | Thu Dec-02-10 11:17 AM |
Charter member
647 posts
|    |
|
#48. "RE: Source Code Work"
In response to Reply # 47
|
|
|
|                               |
|
wolfman | Fri Dec-03-10 12:42 AM |
Member since Apr 15th 2003
351 posts
|    |
|
#49. "RE: Source Code Work"
In response to Reply # 47
Fri Dec-03-10 12:43 AM by wolfman
|
ok, so far we are not sure what the MLR is doing (bc I also didn't saw any real change)
but changing the values in mission.cpp does increase the amount of objects one can see (I tested it and magic tested it)
Magic, when are you releasing your latest build?
And I hope we can get this into MC2X too
I tripled the pool values (kept color) in mission and editor cpp and left everything else in any other file as is.
Great result about 50 mechs in superdense woods and no missing textures.
Definitely much better than before. Still you can reach limitations but at other level compared to before.
I tried to change one entry from 40*1024*1024 to 120*3072*3072.
This variant crashed the gaming exe with the error message: Heap TinyGeom Error bad 0009. I took this back to the starting values.
Will use the settings in next MC2X revisioncheck out wolfman's missions and wolfman's forum at http://s2.invisionfree.com/wolfman_x_forum/index.php
|
|
|
|
|                                 | |
|                                   |
|
rollin | Fri Dec-03-10 08:45 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#51. "RE: Source Code Work"
In response to Reply # 50
|
I also hope we can close the technical gap between wolfmans and magics mod. For me it would be great to port my maps easily from one mod to another and just deal with the balance changes based on the different gameplay
I know wolfman has it's way of thinking about mc2 as magic has too. But getting conform on a basic technical level would help both!
That's also my motivation why I release all my changes and modifications to everyone!
|
|
|
|
|                                     |
|
Starman01 | Fri Dec-03-10 08:53 AM |
Member since Nov 10th 2010
110 posts
|    |
|
#52. "RE: Source Code Work"
In response to Reply # 51
|
|
|
|                                   | |
|                                     | |
|                                       | |
|                                       | |
|                                       |
|
rollin | Sun Dec-05-10 06:40 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#57. "RE: Source Code Work"
In response to Reply # 56
|
I say 512 is too much.
Of course it's never bad to loosen restrictions but I doubt there are many people around capable of filling all these pixel with a good texture.
Also this would lead to redoing all default textures to get an consistent look doubling or tribbling all objects polycounts to really make the texel density worth the effort
I say, 256 would be nice and still difficult to fill
A Texture size of 512 or 1024 would mean to use normalmapping and creating a mechcommander 3
for everything else, less is more then enough!
- my opinion
|
|
|
|
|                                       | |
|                                       | |
|                                     | |
|                                       | |
|
Magic | Tue Dec-07-10 09:50 AM |
Charter member
647 posts
|    |
|
#62. "RE: Source Code Work"
In response to Reply # 0
|
Rollin,
have you seen this?
struct Limits { static unsigned Max_Number_Vertices_Per_Frame, Max_Number_Primitives_Per_Frame, Max_Number_ScreenQuads_Per_Frame, Max_Size_Of_LightMap_MemoryStream;
enum { Max_Number_Vertices_Per_Mesh = 1024, Max_Number_Vertices_Per_Polygon = 32, Max_Number_Of_Texture_Bits = 14, Max_Number_Textures = 1 << Max_Number_Of_Texture_Bits, Max_Number_Of_NGon_Vertices = 9, Max_Number_Of_Multitextures = 8, Max_Number_Of_Lights_Per_Primitive = 16, Max_Number_Of_FogStates = 4 }; };
in mlr.hpp.
You are much more versed in modeling then I.
|
|
|
|
|   |
|
rollin | Wed Dec-08-10 10:19 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#63. "RE: Source Code Work"
In response to Reply # 62
|
yes I've seen them
but I don't see any use to increase them atm. Interesting only of someone wants to try to create a new graphics quality level
|
|
|
|
|           |
|
rollin | Mon Mar-14-11 08:05 AM |
Member since Aug 25th 2010
114 posts
|   |
|
#70. "RE: Source Code Work"
In response to Reply # 69
|
|
|
|             |
|
rollin | Mon Sep-05-11 09:18 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#71. "RE: Source Code Work"
In response to Reply # 70
|
Hi Magic,
do you also notice an increase in crashes with the more-objects-higher-viewdistance? Working with mc2x r16 atm and not sure what is causing the crashes.
You have told that you have fixed some issues with the memory and poly/objects heaps.. Did this fixed the issues for you? And do you have no more crashes connected with that?
Would be nice to get some insight in what you have found out on that issue!
cheers rollin
|
|
|
|
|               | |
|                 |
|
rollin | Wed Sep-07-11 06:18 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#73. "RE: Source Code Work"
In response to Reply # 72
|
Hi, thanks for the valuable info!
I want to make some tests but I have a simple problem now. If I want to load my map with the mco editor but the editor crashes. Funny is: even if I want to load some of the maps that come with mco it crashes (e_1_x for example)
It would be great if missing objects are replaced by a dummy object. That way I just have to go over all dummy objects and change them back to the object they should be.
Strange is that the editor has the functionality to do so as long as the object is not visible (== hidden behind the fog)
would be great to have this dummy-object-function
|
|
|
|
|                   | |
|                     |
|
rollin | Thu Sep-08-11 07:31 PM |
Member since Aug 25th 2010
114 posts
|   |
|
#75. "RE: Source Code Work"
In response to Reply # 74
|
yes got it working. thanks for the tip!!
btw a suggestion: can you "fade out" the shadows? like make the transparency dependent on the view distance and disable them when they reach the max range
about the dummy object hting. I'll have to setup a source code and visual studio first. my other rig where I did all the code work is not here where I live now. Will try to get this running.
thx again for your help!!
|
|
|
|
|                       | |
|