Rockdude101

Thanks, I just figured out how to use ReflectionHelper.getPrivateValue to get what I needed: ThreadDownloadImageData dat=(ThreadDownloadImageData) Minecraft.getMinecraft().getTextureManager().getTexture(entity.getLocationSkin()); BufferedImage bufferedimage = ReflectionHelper.getPrivateValue(ThreadDownloadImageData.class, dat, "bufferedImage","field_110560_d"); int[] iA=bufferedimage.getRGB(0, 0, bufferedimage.getWidth(), bufferedimage.getHeight(), null, 0, bufferedimage.getWidth()); (This code, given an EntityPlayer, simply obtains the player's skin.) I just made the false assumption that you needed to create your own extension of the class that you want to reflect a variable from, but it turns out you just need the instance stored in an accessible variable somewhere. Once again, thank you for helping me resolve this problem!

Here's what I've found: after a ThreadDownloadImageData is created in SkinManager.loadSkin, TextureManager.loadTexture is called, which then calls ThreadDownloadImageData.loadTexture. This then calls its submethod SimpleTexture.loadTexture, which calls resourceManager.getResource to get an IResource which is then read into a BufferedImage, which is then finally used in TextureUtil.uploadTextureImageAllocate to bind the texture or something. After that, the ThreadDownloadImageData then reads its cache File as a BufferedImage and uses that to set its IImageBuffer. If for some reason the cache file is unavailable, then the ThreadDownloadImageData just creates a thread to download the image. After all this, the TextureManager then puts the TDID in its own Map, tied to the ResourceLocation provided, and TextureManager also has a method to easily get the TDID from the ResourceLocation. I am completely able to access that TDID using Minecraft.getMinecraft().getTextureManager().getTexture(entity.getLocationSkin()). The issue is, I need to get the BufferedImage buried within in order to get an int[] of the texture's pixels, but the BufferedImage is private. The only thing that I can get from the TDID is its gl texture id. This is why I'm trying to see if I can somehow get the player's skin into my own variation of ThreadDownloadImageData, but with the BufferedImage public or accessable via methods. If it is possible to use TextureUtils or something in order to get the BufferedImage from the gl texture id, or even rip the BufferedImage out of the TDID directly, then that is what I would like to know; if there is no simple way, then I might look into TGG's other method of rendering the skin into a framebuffer. In short, all I need to know is how to get the BufferedImage from a fully initialized ThreadDownloadImageData; if I can do that, then my problem is solved.

I tried to use the ReflectionHelper to get the BufferedImage public, but I don't think my implementation is correct, as the BufferedImage that I'm getting back is just returning null. Here's the code that uses my superclass of ThreadDownloadImageData: DNATThreadDownloadImageData itextureobject = (DNATThreadDownloadImageData)DNATThreadDownloadImageData.getDownloadImageSkin(entity.getLocationSkin(), entity.getName()); BufferedImage im=itextureobject.getBufferedImage(); And here's the superclass itself: public class DNATThreadDownloadImageData extends ThreadDownloadImageData { public DNATThreadDownloadImageData(File cacheFileIn, String imageUrlIn, ResourceLocation textureResourceLocation, IImageBuffer imageBufferIn) { super(cacheFileIn, imageUrlIn, textureResourceLocation, imageBufferIn); } public BufferedImage getBufferedImage(){ try { return (BufferedImage)ReflectionHelper.findField(ThreadDownloadImageData.class, "bufferedImage","field_110560_d").get(this); } catch (IllegalArgumentException e) { ModDNAT.logger.error("IllegalArgumentException occurred when attempting to get the BufferedImage from a DNATThreadDownloadImageData",e); return null; } catch (IllegalAccessException e) { ModDNAT.logger.error("IllegalAccessException occurred when attempting to get the BufferedImage from a DNATThreadDownloadImageData",e); return null; } } public static DNATThreadDownloadImageData getDownloadImageSkin(ResourceLocation resourceLocationIn, String username) { TextureManager texturemanager = Minecraft.getMinecraft().getTextureManager(); ITextureObject itextureobject = texturemanager.getTexture(resourceLocationIn); if (itextureobject == null || !(itextureobject instanceof DNATThreadDownloadImageData)) { itextureobject = new DNATThreadDownloadImageData((File)null, String.format("http://skins.minecraft.net/MinecraftSkins/%s.png", new Object[] {StringUtils.stripControlCodes(username)}), DefaultPlayerSkin.getDefaultSkin(AbstractClientPlayer.getOfflineUUID(username)), new ImageBufferDownload()); texturemanager.loadTexture(resourceLocationIn, itextureobject); } return (DNATThreadDownloadImageData)itextureobject; } } The bufferedImage field inside the the ThreadDownloadImageData part of the DNATThreadDownloadImageData is being initialized properly, so the error lies within DNATThreadDownloadImageData.getBufferedImage(). My guess is that the ".get(this)" part has to be changed, but I don't know what to change it to; I didn't even know that it was possible to access a private instance variable. Once again, thanks for the help so far, and let me know if you need any more info.

I've looked over the ThreadDownloadImageData class to see how I can obtain the data inside, but the BufferedImage contained within is private and so far I haven't succeeded in using my own implementations of classes to get a variation of the ThreadDownloadImageData with the image public. At the moment, the only thing I can think of doing is just creating an entire new cache of player skins, since Minecraft's TextureManager just had the private image data classes. Do you know of any other ways of pulling the int[] or BufferedImage from the player skin cache, or, if making my own cache is the only option, could you tell me how to go about doing that?

So, looking at the DynamicTexture object, I see two ways I could potentially get an int[] of the player's skin: either getting a DynamicTexture object itself and calling DynamicObject.getTextureData, or using mc.getTextureManager().getDynamicTextureLocation(String,DynamicTexture) to get a ResourceLocation that I assume might be useable with TextureUtil.readImageData. However, I can't find any functions that could possibly return a DynamicTexture, so the latter option seems to be the only option. Looking at calls of this function, it seems that the first String is the name of some texture: "modlogo","lightMap","texturepackicon", etc. So, should I be putting the player's texture UUID here to get the skin? Of course, I might be looking at the wrong place, and if there is just a way to pull a BufferedImage of the player's cached skin from somewhere, that works too. Please let me know what I should be doing with the DynamicTexture; I understand that the player's skin is likely stored in this format, but I'm not sure how even Minecraft references it. Thanks for all of the help so far.

As a part of my mod, I need to be able to take control of rendering the player, so I handled the necessary event and redirected the call to RenderPlayer.doRender to my own renderer's doRender method. As far as I know, this code is perfectly fine, but I'm including it in case the event itself is necessary for the following problem: @SubscribeEvent public void onPlayerRender(RenderPlayerEvent.Pre event){ event.setCanceled(true); float f=event.getEntityPlayer().prevRotationYaw + (event.getEntityPlayer().rotationYaw - event.getEntityPlayer().prevRotationYaw) * event.getPartialRenderTick(); DNATRenderPlayer.instance.doRender((AbstractClientPlayer) event.getEntityPlayer(), event.getX(), event.getY(), event.getZ(), f, event.getPartialRenderTick()); } Now, within DNATRenderPlayer.doRender, everything is the same as RenderPlayer.doRender, except with its calls to RenderLivingBase.doRender and Render.doRender being pasted directly into the method instead of being called using super.doRender. This is so I can call my own versions of RenderLiving.renderModel and Render.bindEntityTexture. My current goal is to get the player's skin as an int[] so that I can tweak it and then bind the modified version. Currently, this is my code for this: protected boolean bindEntityTexture(AbstractClientPlayer entity) { ResourceLocation resourcelocation = entity.getLocationSkin(); if (resourcelocation == null) { return false; } else if(/*just testing for some entity capabilities here*/) { int[] iA; try { iA=TextureUtil.readImageData(Minecraft.getMinecraft().getResourceManager(),entity.getLocationSkin()); } catch (Exception e) { ModDNAT.logger.error("Exception occurred when attempting to read player ResourceLocation.",e); return false; } //manipulation of iA occurs here return true; }else{ this.bindTexture(resourcelocation); return true; } } However, a FileNotFoundException happens on the line where iA is set to whatever TextureUtil.readImageData returns. Upon digging through the stack trace, I found that FallbackResourceManager.getResource is throwing it: public IResource getResource(ResourceLocation location) throws IOException { this.checkResourcePath(location); IResourcePack iresourcepack = null; ResourceLocation resourcelocation = getLocationMcmeta(location); for (int i = this.resourcePacks.size() - 1; i >= 0; --i) { IResourcePack iresourcepack1 = (IResourcePack)this.resourcePacks.get(i); if (iresourcepack == null && iresourcepack1.resourceExists(resourcelocation)) { iresourcepack = iresourcepack1; } if (iresourcepack1.resourceExists(location)) { InputStream inputstream = null; if (iresourcepack != null) { inputstream = this.getInputStream(resourcelocation, iresourcepack); } return new SimpleResource(iresourcepack1.getPackName(), location, this.getInputStream(location, iresourcepack1), inputstream, this.frmMetadataSerializer); } } And after looking at the function of this code, it seems that it is supposed to iterate through FallbackResourceManager.resourcePacks and find the texture that is called; though apparently, it's not there. After some more digging, I did find a file in C:/Users/Me/.gradle/caches/minecraft/assets/skins that matches the number after the FileNotFoundException explanation in the stacktrace, so I assume this is my cached skin, it's just not being found for some reason. Just in case, here's the stacktrace: So, in short, this line: iA=TextureUtil.readImageData(Minecraft.getMinecraft().getResourceManager(),entity.getLocationSkin()); is not actually finding the skin, even though it exists. I should also note that I did have this same problem in a previous thread (http://www.minecraftforge.net/forum/index.php?topic=24897.0), but that was within a GUI and outside of the actual game, so I was able to resolve it my just handling the download code myself. Of course, this is pretty impractial for rendering. Please let me know if you need any more code, and thanks in advance for any help.

I created a small workaround using a new GUI that redirects to the GUI that I originally wanted to show. Since the GuiMainMenu and any subclasses seem to be exempt from this graphical bug, I made my redirection GUI extend it allow GuiMainMenu's drawScreen method to run once, before changing the GUI. I pasted the code of the GUI below. Anyways, thanks for your help.

Disabling the loading screen does fix the problem, though this does remove the loading bars, which are always a nice thing to have available to the client, especially in big modpacks. Is there a way to fix this problem without having to disable configuration settings, such as setting up an event to call the GUI as soon as the loading screen closes? Or is there something present in the GuiMainMenu file that could allow my GUI to overcome this bug?

For whatever reason, a GUI that I created is not rendering after being called during a GuiOpenEvent handled by my mod's event handler. The GUI is being set by simply overwriting event.gui within the function, and this worked perfectly in 1.7.10. In 1.8, however, the GUI just isn't drawing. The GUI is functional and has working, clickable buttons, but nothing is showing; the window is completely white. Here's a few bullets of the facts that I have found so far: - With the GuiScreen I'm using, initGui is called when the GUI "loads", and drawScreen is repeatedly called after initialization. - The blank, white appearance isn't exclusive to my GUI; I also tested it with a GuiYesNo (which is what my GUI is somewhat based off) and a few other of my button-and-text-based GUIs. The only instance of a GUI showing was with one of my GUIs that extends GuiMainMenu. That's all I have discovered so far, and what I've found seems to suggest that the way I'm calling my GUI is improper in some way. Below is the source code for my mod file, event handler, and the GUI I'm trying to show. Thanks in advance for any help.

This thread has been inactive for quite a while, but I managed to find a solution. Basically, I used some of the code from ThreadDownloadImageData.func_152433_a() to get a BufferedImage, and simply ran the code from TextureUtils.readImageData(), without the first line, to get the int[] I wanted. Here's a link to the pastebin: http://pastebin.com/fzWmheB9 Also, let me know if there are any major mistakes in the code.

Compiling it didn't seem to work. I used some logging to figure out that GameProfile game does have my username, Rockdude101, inside its instance, but the profile Map still isn't returning true. I changed the if statement to always be true, and sure enough, there was a NullPointerException at this line in SkinManager.func_152789_a: final ResourceLocation resourcelocation = new ResourceLocation("skins/" + p_152789_1_.getHash()); The problem spot in my code is probably this line, in that case: Map profmap = mc.func_152342_ad().func_152788_a(game); I'm finding it very difficult to understand this piece of code, due to the complex structure of it, so any insight on this would be any helpful.

The code fits in perfectly, thanks for the help! There's only one problem; profmap.get(Type.SKIN) keeps returning false, even when I the flag to change the username to my own (which also works fine, I might add). Changing the --username and --password flags to my credentials also returns false. I'm not sure whether there is a problem in my code, or the skin simply doesn't download due to the fact that it's in a development environment. In case there's something wrong, I've made a pastebin link: http://pastebin.com/P1Vvph6V Also, I'll be sure to make a mental note of the answers to the other two questions. Thanks for all the help so far!

So, I'm building a GUI, and for the mod I'm creating, I need to access the current player's skin as a ResourceLocation object. The getLocationSkin() function within the AbstractClientPlayer class is returning nulls, most likely because the player object hasn't been initialized (my GUI is accessed from the main menu). My question is: how can I get a ResourceLocation of the current player's skin, even from the main menu? The reason I want a ResourceLocation is so that I can use TextureUtil.readImageData(ResourceManager,ResourceLocation) to break it up into an int array. Also two other simple side questions: how do you access the current player's name within code (e.g. ForgeDevName) and is it possible to change the default name to a different name for testing purposes? Thanks for any help.

Works perfectly, thanks for all the help

So overriding the GUI just changes the displayed GUI outright? Let me try this real quick... EDIT: Ok, it displays the MainMenuWarning GUI perfectly. The only problem is now when I open the new MainMenu from MainMenuWarning.actionPerformed using mc.currentscreen=new DNATGuiMainMenu(), it crashed because of a NullPointerException. Here's the stacktrace if it helps: https://gist.github.com/anonymous/1407cf49ad0a59adfbba I have an idea of what might be causing the problem, so I'm going to look into that. Thanks for the help with the other GUI though! EDIT 2: It appears to be a problem within the GuiMainMenu.renderSkybox(int,int,float) method, on the line this.mc.getFramebuffer().unbindFramebutter(). This probably means that the Framebuffer in the main Minecraft instance is still not initialized; so the new question is, how can I initialize this safely?