Navi
-
- Gnoblar
- Posts: 22
- Joined: Thu Jan 04, 2007 9:39 pm
Hey ajs15822!
Sorry for the late response to you suggestion! I still can't scroll down a page clicking and dragging the scroll bar, or write on a NaviMaterial, even changing all injectMaterial functions The strangest thing is that I can select text on a NaviMaterial by dragging and dropping! Well... I'll keep digging
Sorry for the late response to you suggestion! I still can't scroll down a page clicking and dragging the scroll bar, or write on a NaviMaterial, even changing all injectMaterial functions The strangest thing is that I can select text on a NaviMaterial by dragging and dropping! Well... I'll keep digging
-
- Kobold
- Posts: 33
- Joined: Sun Jan 21, 2007 7:24 pm
rotate image
Hi guys,
we are currently testing Navi for creating an interface that we render-to-texture onto the screen of a virtual device in our environment.
On that interface we would like to display bitmap icons. That we managed to do easily with positioning divs. What I didn't find a solution for until now is rotating bitmaps icons. Seems there is no css way of doing that. Did I overlook something or ss there a different way in Navi?
Many thanks!
Markus
we are currently testing Navi for creating an interface that we render-to-texture onto the screen of a virtual device in our environment.
On that interface we would like to display bitmap icons. That we managed to do easily with positioning divs. What I didn't find a solution for until now is rotating bitmaps icons. Seems there is no css way of doing that. Did I overlook something or ss there a different way in Navi?
Many thanks!
Markus
- Kanma
- Halfling
- Posts: 61
- Joined: Sun Feb 22, 2004 3:41 pm
- Location: Martigny, Switzerland
- Contact:
Firefox is in fact licensed under a MPL/GPL/LGPL "triple license", so you can use it to make a closed source project.Caius wrote:I have a question, about licensing issues.
Navi is LGPL and Firefox is GPL.
Is it possible to make a closed source project that uses Navi, and that show webpages using firefox sdk?
-
- Gnoblar
- Posts: 1
- Joined: Sun Sep 02, 2007 4:24 pm
Problem with pageNaviData.get()
Hi,
in my project i changes from html to xhtml because of use of svg graphics.
With this change my javascript code for writing text isn`t functioning anymore. I clasped my javascript with //<!CDATA[...expression but without any luck.
</div>
<div id="start_beacon_ID2">
<script type="text/javascript">
//<![CDATA[
if(pageNaviData.doesExist("start_beacon"))
{
document.write(pageNaviData.get("start_beacon"));
}
//]]>
</script>
</div>
Has anyone an idear what could got wrong?
philipp
in my project i changes from html to xhtml because of use of svg graphics.
With this change my javascript code for writing text isn`t functioning anymore. I clasped my javascript with //<!CDATA[...expression but without any luck.
</div>
<div id="start_beacon_ID2">
<script type="text/javascript">
//<![CDATA[
if(pageNaviData.doesExist("start_beacon"))
{
document.write(pageNaviData.get("start_beacon"));
}
//]]>
</script>
</div>
Has anyone an idear what could got wrong?
philipp
- Kanma
- Halfling
- Posts: 61
- Joined: Sun Feb 22, 2004 3:41 pm
- Location: Martigny, Switzerland
- Contact:
-
- Kobold
- Posts: 33
- Joined: Sun Jan 21, 2007 7:24 pm
urgent: programmatic control of svg
I second the point about a separate forum!
Hey guys, if anybody of you has some experience with controlling svg through javascript on a navi that's rendered on a texture, we'd really appreciate some help. See the posting above from pilgrimm2000.
The javascript worked before introducing svg, after it didn't anymore.
Thanks!
Markus
Hey guys, if anybody of you has some experience with controlling svg through javascript on a navi that's rendered on a texture, we'd really appreciate some help. See the posting above from pilgrimm2000.
The javascript worked before introducing svg, after it didn't anymore.
Thanks!
Markus
- ajs15822
- OGRE Expert User
- Posts: 570
- Joined: Mon Jan 02, 2006 2:05 am
- Location: Texas
- x 2
- Contact:
"Hay, wat happened to ajs15822? Why isn't he replying as often?"
Wooo, been busy getting into the flow of a new semester @ my college. Once I find my rhythm, I'll be able to give Navi some more lovin'.
- The <canvas> tag.
- SVG (best, imo)
- Flash
The issue you're facing here is that 'document.write()' can invalidate the XHTML DOM. Thus, you must instead update the contents (.innerHTML) of individual elements via javascript.
In v1.4, all pages that include 'Navi.js' will send an empty piece of NaviData to the application named 'ready' when the DOM is ready to be manipulated/changed. Thus, you just bind your initialization function to 'ready' and intialize page elements using 'NaviManager::naviEvaluateJS'.
The code in the current SVN is pretty much what v1.4 will be when it is released. The only delay is my writing new documentation/tutorials/wiki articles for the new release.
If yall would have it, I would actually like to establish a forum on my own server to have more freedom of organization and creative reign.
Wooo, been busy getting into the flow of a new semester @ my college. Once I find my rhythm, I'll be able to give Navi some more lovin'.
I'll try to put together a test case this weekend and see if I can't offer more help. Until then, good luck!RedEagle wrote:Hey ajs15822!
Sorry for the late response to you suggestion! I still can't scroll down a page clicking and dragging the scroll bar, or write on a NaviMaterial, even changing all injectMaterial functions. The strangest thing is that I can select text on a NaviMaterial by dragging and dropping! Well... I'll keep digging.
To do vector manipulation you have three options:mklann wrote:Hi guys,
we are currently testing Navi for creating an interface that we render-to-texture onto the screen of a virtual device in our environment.
On that interface we would like to display bitmap icons. That we managed to do easily with positioning divs. What I didn't find a solution for until now is rotating bitmaps icons. Seems there is no css way of doing that. Did I overlook something or ss there a different way in Navi?
Many thanks!
Markus
- The <canvas> tag.
- SVG (best, imo)
- Flash
Hey there philipp, as of NaviLibrary v1.4, that method of page initialization has been deprecated in favor of application side-initialization.pilgrimm2000 wrote:Hi,
in my project i changes from html to xhtml because of use of svg graphics.
With this change my javascript code for writing text isn`t functioning anymore. I clasped my javascript with //<!CDATA[...expression but without any luck.
</div>
<div id="start_beacon_ID2">
<script type="text/javascript">
//<![CDATA[
if(pageNaviData.doesExist("start_beacon"))
{
document.write(pageNaviData.get("start_beacon"));
}
//]]>
</script>
</div>
Has anyone an idear what could got wrong?
philipp
The issue you're facing here is that 'document.write()' can invalidate the XHTML DOM. Thus, you must instead update the contents (.innerHTML) of individual elements via javascript.
In v1.4, all pages that include 'Navi.js' will send an empty piece of NaviData to the application named 'ready' when the DOM is ready to be manipulated/changed. Thus, you just bind your initialization function to 'ready' and intialize page elements using 'NaviManager::naviEvaluateJS'.
Code: Select all
NaviManager::Get().bind("myNavi", "ready", NaviDelegate(this, &MyApplication::initializeMyNavi));
// somewhere else:
void MyApplication::initializeMyNavi(const NaviData &naviData)
{
std::string myText = "This text was generated by the application!";
NaviManager::Get().naviEvaluateJS("myNavi", "$('start_beacon_ID2').setHTML('" + myText + "')");
}
The agenda for v1.5 is integrating LLMozLib's latest code dump (of which there is some experimental support for Linux/Mac). So sorry, not yet.kcbanner wrote:Is there a working linux build of Navi yet?
Kanma wrote:Just a suggestion: maybe you should ask Sinbad for a dedicated subforum in the addons one?
kcbanner wrote:I agree, one thread for everyone's probs doesn't work lol
Humm, maybe . I hadn't realized this little project had grown so much.mklann wrote:I second the point about a separate forum!
If yall would have it, I would actually like to establish a forum on my own server to have more freedom of organization and creative reign.
A month or two ago, I was really into SVG + Javascript. I'll see what I can do about putting a few examples up on the wiki.mklann wrote:Hey guys, if anybody of you has some experience with controlling svg through javascript on a navi that's rendered on a texture, we'd really appreciate some help. See the posting above from pilgrimm2000.
The javascript worked before introducing svg, after it didn't anymore.
Thanks!
Markus
-
- Kobold
- Posts: 33
- Joined: Sun Jan 21, 2007 7:24 pm
thanks and svg
Hi ajs15822,
thanks a lot for the reply - we'll try your suggestion to resolve the update issue right away.
About svg: what would be really helpful is a very simple example of rotating with the keyboard an svg bitmap object on a rendered-to-texture Navi. If the bitmap doesn't work we can also redraw our icons as svg.
If we manage to do this with svg, we'll be switching to Navi in our application and I'd be happy to provide some documented samples.
Best,
Markus
thanks a lot for the reply - we'll try your suggestion to resolve the update issue right away.
About svg: what would be really helpful is a very simple example of rotating with the keyboard an svg bitmap object on a rendered-to-texture Navi. If the bitmap doesn't work we can also redraw our icons as svg.
If we manage to do this with svg, we'll be switching to Navi in our application and I'd be happy to provide some documented samples.
Best,
Markus
- madmark
- Goblin
- Posts: 292
- Joined: Fri Jul 22, 2005 8:14 pm
- Location: Midwest USA
- Contact:
Updated NaviLua for anyone using it.
Info here.
Mostly a release to conform to the API changes in version of the SVN repository. There are a few bugfixes as well.
Also: I can set up a forum for Navi users on my WBI worksite if all y'alls would like, but the idea of getting an official Ogre addon forum is waaaaaay better.
Info here.
Mostly a release to conform to the API changes in version of the SVN repository. There are a few bugfixes as well.
Also: I can set up a forum for Navi users on my WBI worksite if all y'alls would like, but the idea of getting an official Ogre addon forum is waaaaaay better.
-
- OGRE Expert User
- Posts: 557
- Joined: Wed May 05, 2004 3:19 pm
- Location: Portland, OR, USA
- Contact:
I wasn't sure if NaviLua was "up to date" with Navi (since Navi's API seems to change quite a bit and quickly still) but I *want* to use it, so I'll take a look.madmark wrote:Updated NaviLua for anyone using it.
Info here.
Mostly a release to conform to the API changes in version of the SVN repository. There are a few bugfixes as well.
Also: I can set up a forum for Navi users on my WBI worksite if all y'alls would like, but the idea of getting an official Ogre addon forum is waaaaaay better.
I also vote for an Ogre Addon forum since it's much more convenient for me to just go to this website to check the forums. I dislike having to go to multiple websites to check on various projects (which are all Ogre related).
Chaster
- madmark
- Goblin
- Posts: 292
- Joined: Fri Jul 22, 2005 8:14 pm
- Location: Midwest USA
- Contact:
Now working on getting my stuff working under SVN version 16 and I have an issue with page NaviData.
The example on page 14 of this thread does not seem to workAlways displays "NaviData Not Present" even though I am certain that there isNaviNata...
UPDATE: Changed Navi.js to init $PND to null and changed the test above against null and have the same result. I am getting the "ready" NaviData passed back so that domReady event is firing, just not defining a $PND.
I also displayed the url and it has valid NaviData
The example on page 14 of this thread does not seem to work
Code: Select all
if($defined($PND.get))
{
document.write("NaviData Present<br\>");
if($defined($PND.get("buttons")))
{
document.write("Button string: " + $PND.get("buttons")+"<br\>");
}
else
{
document.write("NO Button string!!!<br\>");
}
}
else
{
document.write("NaviData Not Present<br\>");
}
UPDATE: Changed Navi.js to init $PND to null and changed the test above against null and have the same result. I am getting the "ready" NaviData passed back so that domReady event is firing, just not defining a $PND.
I also displayed the url and it has valid NaviData
Last edited by madmark on Wed Sep 05, 2007 8:56 pm, edited 2 times in total.
-
- Halfling
- Posts: 55
- Joined: Sat Jul 15, 2006 4:03 pm
- Contact:
haven't read the entire thread, here some questions:
-i have no GUI with openGL. Is this an ogre problem, somehow related to my graphics card/its drivers or what? edit: updated drivers, no effect
-what about fonts? are system's fonts used? can I add custom fonts? if yes, what formats are supported?
-is there still the limitation with powers-of-2 sizes?
-i have no GUI with openGL. Is this an ogre problem, somehow related to my graphics card/its drivers or what? edit: updated drivers, no effect
-what about fonts? are system's fonts used? can I add custom fonts? if yes, what formats are supported?
-is there still the limitation with powers-of-2 sizes?
- ajs15822
- OGRE Expert User
- Posts: 570
- Joined: Mon Jan 02, 2006 2:05 am
- Location: Texas
- x 2
- Contact:
@mklann: Sure, I'll try to write up an example for you this evening.
@madmark: w00t! Thanks a lot!
@chaster: If yall really want Navi in an Ogre Subforum, I have no objections.
@madmark:
Well, you opened up a can of worms: 'if($defined($PND.get))'. You're checking whether or not the member property 'get' is defined or not (not the actual result of get()).
I understand what you wish to do (check whether or not NaviData has been passed to the page) and have realized a flaw in my implementation. Your inline javascript (<script> tags inside the <body>) is executed before the DOM is actually ready, thus $PND, at that point in time, is equal to 'Class.empty' (see line 7358 of Navi.js).
I think a much better compromise would be to initialize $PND on that line to '$ND('')' instead, so that in your inline javascript, you could just do: "if($PND.getName().length)" in place of your previous statement.
Humm, this also means I will have to move the chunk of code below line 7358 into the global scope (that is, it is executed before the DOM is ready) so that you could actually work with $PND using inline javascript.
I'll update the SVN in a bit.
Still, the reason I hadn't caught this earlier is because this method of using $PND and 'document.write' to alter the contents of the page is deprecated. You should use application-side javascript evaluation to initialize page elements, (see my reply to pilgrimm2000 above).
Sorry for the confusion, some new tutorials are coming that should clear all this up.
@pra
- On my laptop's videocard, Ogre's dynamic textures in OpenGL do not work. Running the NaviDemo or the Ogre DynTex demo under OpenGL gives me non-visible materials. Haven't had the time to go on a witch hunt yet.
- System fonts are used, there is no way at this time to make Gecko use custom fonts that aren't registered with the OS. If you wish to use a custom font, you'll have to install it along with your executable.
- Nope, that limitation is non-existent with Navi v1.4.
@madmark: w00t! Thanks a lot!
@chaster: If yall really want Navi in an Ogre Subforum, I have no objections.
@madmark:
Well, you opened up a can of worms: 'if($defined($PND.get))'. You're checking whether or not the member property 'get' is defined or not (not the actual result of get()).
I understand what you wish to do (check whether or not NaviData has been passed to the page) and have realized a flaw in my implementation. Your inline javascript (<script> tags inside the <body>) is executed before the DOM is actually ready, thus $PND, at that point in time, is equal to 'Class.empty' (see line 7358 of Navi.js).
I think a much better compromise would be to initialize $PND on that line to '$ND('')' instead, so that in your inline javascript, you could just do: "if($PND.getName().length)" in place of your previous statement.
Humm, this also means I will have to move the chunk of code below line 7358 into the global scope (that is, it is executed before the DOM is ready) so that you could actually work with $PND using inline javascript.
I'll update the SVN in a bit.
Still, the reason I hadn't caught this earlier is because this method of using $PND and 'document.write' to alter the contents of the page is deprecated. You should use application-side javascript evaluation to initialize page elements, (see my reply to pilgrimm2000 above).
Sorry for the confusion, some new tutorials are coming that should clear all this up.
@pra
- On my laptop's videocard, Ogre's dynamic textures in OpenGL do not work. Running the NaviDemo or the Ogre DynTex demo under OpenGL gives me non-visible materials. Haven't had the time to go on a witch hunt yet.
- System fonts are used, there is no way at this time to make Gecko use custom fonts that aren't registered with the OS. If you wish to use a custom font, you'll have to install it along with your executable.
- Nope, that limitation is non-existent with Navi v1.4.
- madmark
- Goblin
- Posts: 292
- Joined: Fri Jul 22, 2005 8:14 pm
- Location: Midwest USA
- Contact:
Figured out my problem!
The $PND is created at the domReady event while I am using inline JavaScript to set up my menu. My code was being run before the URL was decoded.
Once I moved the code that builds the $PND var out of that function, all was well.
Any suggestions on how to work around this or should I just go with this modified version of Navi.js. That would sort of suck...
The $PND is created at the domReady event while I am using inline JavaScript to set up my menu. My code was being run before the URL was decoded.
Once I moved the code that builds the $PND var out of that function, all was well.
Any suggestions on how to work around this or should I just go with this modified version of Navi.js. That would sort of suck...
- ajs15822
- OGRE Expert User
- Posts: 570
- Joined: Mon Jan 02, 2006 2:05 am
- Location: Texas
- x 2
- Contact:
I fixed that bug and refactored some of the initialization code for $PND. It's in the trunk right now.madmark wrote:Figured out my problem!
The $PND is created at the domReady event while I am using inline JavaScript to set up my menu. My code was being run before the URL was decoded.
Once I moved the code that builds the $PND var out of that function, all was well.
Any suggestions on how to work around this or should I just go with this modified version of Navi.js. That would sort of suck...
-
- Halfling
- Posts: 55
- Joined: Sat Jul 15, 2006 4:03 pm
- Contact:
so it's the card's faultajs15822 wrote:@pra
- On my laptop's videocard, Ogre's dynamic textures in OpenGL do not work. Running the NaviDemo or the Ogre DynTex demo under OpenGL gives me non-visible materials. Haven't had the time to go on a witch hunt yet.
DynTex works fine, though, the only demo I have problems with is the DereferredShader (or something like that), everything else works fine since the driver update...
well, I hope there's a way to fix it...
what do you think, would a JS function to replace chars with corresponding images use too much performance?^^- System fonts are used, there is no way at this time to make Gecko use custom fonts that aren't registered with the OS. If you wish to use a custom font, you'll have to install it along with your executable.
or, maybe there's an ogre function to create a single image with the desired text in the desired font (putting single images together or stuff)
that's good- Nope, that limitation is non-existent with Navi v1.4.
oh, and another thing: do different navis share the javascript vars and functions, or are they like different tabs in firefox?
-
- Halfling
- Posts: 55
- Joined: Sat Jul 15, 2006 4:03 pm
- Contact:
- ajs15822
- OGRE Expert User
- Posts: 570
- Joined: Mon Jan 02, 2006 2:05 am
- Location: Texas
- x 2
- Contact:
Nope, neither of those methods will work . This is really an issue with Gecko: Gecko doesn't support CSS2 @font-face (dynamic fonts) yet, and who knows when, if ever, they will. Thus, if you really really must have custom fonts, you'll just have to install it along with your distributable.pra wrote:what do you think, would a JS function to replace chars with corresponding images use too much performance?^^
or, maybe there's an ogre function to create a single image with the desired text in the desired font (putting single images together or stuff)
Each Navi has its own individual script context, very much like separate tabs.oh, and another thing: do different navis share the javascript vars and functions, or are they like different tabs in firefox?
Going off on a tangent here, I'm getting a little annoyed at the current API for evaluating/executing Javascript in Navi. This is ugly:
Code: Select all
string myString = "somevalue";
NaviManager::Get().naviEvaluateJS("myNavi", "$('someID').setHTML('" + myString + "')");
string txtboxVal = NaviManager::Get().naviEvaluateJS("myNavi", "$('textboxID').getValue()");
Code: Select all
JSContext myContext = NaviManager::Get().getNaviJSContext("myNavi");
myContext.eval("$('someID').setStyle('color', 'blue')");
myContext.setHTML("someID", "somevalue");
string txtboxVal = myContext.getValue("textboxID");
- ajs15822
- OGRE Expert User
- Posts: 570
- Joined: Mon Jan 02, 2006 2:05 am
- Location: Texas
- x 2
- Contact:
v1.4 = SVNpra wrote:err, i was a n00b and haven't noticed that the current version is 1.3 -.-ajs15822 wrote:- Nope, that limitation is non-existent with Navi v1.4.
and here I do have problems if not power-of-2...
It's pretty much solid for right now, I'm just waiting to release it until I finish writing new tutorials/docs/articles.
-
- Halfling
- Posts: 55
- Joined: Sat Jul 15, 2006 4:03 pm
- Contact:
I don't understand. I meant a function that takes a string, and replaces the chars with IMG tags. ie. 'omfg' => '<img src="o.png"><img src="m.png"><img src="f.png"><img src="g.png">'. I think this may be useful sometimes, thinking of a big main menu, perhaps...ajs15822 wrote:Nope, neither of those methods will work . This is really an issue with Gecko: Gecko doesn't support CSS2 @font-face (dynamic fonts) yet, and who knows when, if ever, they will. Thus, if you really really must have custom fonts, you'll just have to install it along with your distributable.
what about a simple wrapper for certain stuff? likeGoing off on a tangent here, I'm getting a little annoyed at the current API for evaluating/executing Javascript in Navi. This is ugly:Maybe it might be cooler if we could just grab an object from NaviManager that lets you execute/evaluate JavaScript on a certain Navi. It also might have a few commonly-used JavaScript methods exposed:Code: Select all
string myString = "somevalue"; NaviManager::Get().naviEvaluateJS("myNavi", "$('someID').setHTML('" + myString + "')"); string txtboxVal = NaviManager::Get().naviEvaluateJS("myNavi", "$('textboxID').getValue()");
Dunno, tell me what you think.Code: Select all
JSContext myContext = NaviManager::Get().getNaviJSContext("myNavi"); myContext.eval("$('someID').setStyle('color', 'blue')"); myContext.setHTML("someID", "somevalue"); string txtboxVal = myContext.getValue("textboxID");
Code: Select all
Navi myNavi = naviMgr->getById("someObjectsId");
myNavi = naviMgr->getByName("someObjectsName",2); //same as document.getElementsByName("someObjectsName")[2]
myNavi = naviMgr->getByTag("div",5); //same as document.getElementsByTagName("div")[5]
String someStuff = myNavi->getAttribute("innerHTML");
myNavi->setAttribute("innerHTML","some stuff");
String bgcolor = myNavi->getStyleValue("background-color");
myNavi->setStyleValue("position","absolute");
EDIT: downloaded navi from svn, tried width=1020, but it just crashes. in debug mode, it breaks in Navi.cpp, line 426,
Code: Select all
pDest[y*pitch+destx] = naviCache[y*pitch+destx] = B;
more edit: it is possible to make navis "movable" and then move them with right-drag. is it possible to define what exactly you have to do to move them, or even make dragging only possible if clicking a certain region?
and onDblClick seems not to work...
and another edit: openGL seems not to work if I use alpha masks. as soon as I comment the line with "setNaviMask" out, i can see the navi
btw, is there a tutorial for mootools usage? it seems dragging with mootools is much more peformant than my own approach^^
- madmark
- Goblin
- Posts: 292
- Joined: Fri Jul 22, 2005 8:14 pm
- Location: Midwest USA
- Contact:
If you have made the changes to support the inline stuff, is it still deprecated? Not complaining, I see the reasoning, but I have a bunch of scripts that do this in my current project that I really don't want to go back and "do right"...ajs15822 wrote:Still, the reason I hadn't caught this earlier is because this method of using $PND and 'document.write' to alter the contents of the page is deprecated. You should use application-side javascript evaluation to initialize page elements, (see my reply to pilgrimm2000 above).
Wow, must have missed this. Good news indeed!ajs15822 wrote: - Nope, that limitation is non-existent with Navi v1.4.
- danharibo
- Minaton
- Posts: 997
- Joined: Sat Feb 25, 2006 8:14 pm
- Location: Wales, United Kingdom
- Contact:
OGRE EXCEPTION(7:InternalErrorException): LLMozLib failed initialization, could not Startup NaviManager. in NaviManager::Startup at .\Source\NaviManager.cpp (line 93)
Code: Select all
//Test Navi...
using namespace NaviLibrary;
NaviManager::Get().Startup(mWindow);
NaviManager::Get().createNavi("myFirstNavi", "local://BuildLog.html", NaviPosition(Center), 400, 320);
- Noman
- OGRE Retired Team Member
- Posts: 714
- Joined: Mon Jan 31, 2005 7:21 pm
- Location: Israel
- x 2
- Contact:
I just updated to the latest SVN navi, and im getting an error :
Unhandled exception at 0x0043fcd8 in TeamEditorDebug.exe: 0xC0000005: Access violation writing location 0x03765020.
navi.cpp (void Navi::update), line 426
y was 469, x was 0, pitch was 4096 (resolution is 1024x768, so it makes sense)
Any idea what it could be?
Unhandled exception at 0x0043fcd8 in TeamEditorDebug.exe: 0xC0000005: Access violation writing location 0x03765020.
navi.cpp (void Navi::update), line 426
Code: Select all
size_t destx = x * destPixelSize;
pDest[y*pitch+destx] = naviCache[y*pitch+destx] = B; <------------------
pDest[y*pitch+destx+1] = naviCache[y*pitch+destx+1] = G;
pDest[y*pitch+destx+2] = naviCache[y*pitch+destx+2] = R;
pDest[y*pitch+destx+3] = A * fadeMod;
naviCache[y*pitch+destx+3] = A;
Any idea what it could be?