Minecraft Forums

Quote from McTwist

I also have a good news and a bad news:
The bad news is that there is few that know about PixelMap and even fewer that even knows what it does, its capabilities and doesn't want anything to do with it. As a programmer that takes a hit on me, but I keep fighting because I trust in my skills and this project as it gives me experience for future projects!
The good news is that there will possible be a merge/fusion with Cartograph G. Details hasn't been cleared out yet, but I do know that we're currently working on several features that could be included along with it.

I use Overviewer for angled maps, and have been using Pixelmap for a top-down view since you started. If you do merge your work with Cartograph, please ensure the command line solution for generating top-down maps is just as simple, clean, customizable, and fast.

And thanks for all your work.

EDIT: I'm putting our Linux server back up now that 1.0 is out. Do you have up-to-date Linux source I can compile?

Quote from vinic

I use Overviewer for angled maps, and have been using Pixelmap for a top-down view since you started. If you do merge your work with Cartograph, please ensure the command line solution for generating top-down maps is just as simple, clean, customizable, and fast.

And thanks for all your work.

EDIT: I'm putting our Linux server back up now that 1.0 is out. Do you have up-to-date Linux source I can compile?

I have this:
Source
But it does not have the latest blocks, but at least the blocks for 1.8. As I know that the lack of angled view PixelMap have I needed to implement it so you don't have to use two programs to render a map. Not that I want to compete with the other mappers.

I am happy that you like the way PixelMap works, therefor I need to ensure it stays that way. The bad thing with merging PixelMap with Cartograph G is (without burning any bridges(!)) that Cartograph G will loose its value more than PixelMap will. This is because both have plenty of features implemented that is one alike. The only difference is the coding and some additional features Cartograph G possess. Currently I cannot tell which one that will be ripped off at most, but I will try to make the merge so even as possible.

Quote from Peacem4k3r

Mh is it possible to render maps which are bigger than 6 GB with your renderer?

c10t and cartograph both die trying to render my world :\

If you have a 32-bit OS that would be difficult. Especially if you require it to be one whole image. However, there is an old 64-bit version of PixelMap, or if you take the nightly source code and compile it in Linux you could tell it to compile in 64-bit.
On a side note, I think that there is mappers that creates so-called "Online maps" where it uses google maps to display the map. However, you require a web-browser and there is several images instead of just one, but the browser merges them together nicely.

@Peacem4k3r

You may have a problem ever getting one final PNG in one final render from any mapper as the amount of data and also the size of a final full size PNG or your world must be incredibly big, huge, massive, titanic...

Having the same problem with my (only :smile.gif: ) 1gig world I split my map into two smaller sub maps which PM then easily handled. Then in Photoshop I still could not combine them into a final full size png as Photoshop will only save 30000 x30000 pixel pngs and my world was 32416px by 15408px. So at full size I could only save a tiff (or RAW or DICOM) and the tiff was 83megs! Could not even save the 32416px by 15408px Photoshop file as a normal psd had to use a special "Large Image Format" Photoshop format.

Also my two half map pngs were 15 and and 18 megs, call it 30 megs combined, and your map has 10 times a much chunk data (by your save size) so a single full size png could possibly be 20 times bigger or 20 * 30 or 600megs... or even more!!

If you have too many tiles to easily re-combine, you could break up you map save (500ish megs seemed to work for me) render them out and at least you will have a more manageable amount of tiles to work with.

Quote from Peacem4k3r

Well, I do have an 64bit OS (Arch Linux), that wouldn't be the problem ;D

I know there are such mappers, and they did work well for my map, but I do need a complete image of my map. (Like Cartopgraph outputs). So if I succeed on compiling your program for 64 bit, I don't need to write a script to join all the tile images to one big image ( I rendered my map with a java cmdline renderer, the only one who succeeded rendering my 10 gb map :biggrin.gif: But as I said, I got many tile images )

You could read on the first page how to download and compile PixelMap for Linux, but I'll type it here as well:

$ cd /tmp
$ wget http://dl.dropbox.com/u/8304300/Minecraft/PixelMap_nightly.tar.gz
$ tar zxf PixelMap_nightly.tar.gz
$ cd PixelMap_nightly
$ make
$ make install

Then run ./pixelmap with proper command lines to retrieve your rendered map.

No-edit: I searched on the compiler and discovered that you need to make sure what sort of compiler you have. First, you require g++, but you need to make sure it compiles x64 programs and not only x32. Reading this you require a -m64 parameter to compile it. However, I've took the liberty to modify Makefile so it should compile x64 programs.
Makefile
I have not tested it as I don't have any Linux platform available.

Quote from Peacem4k3r

That big :ohmy.gif:

I rendered a 7 GB map in tiles and I got a total size of 62 MB (png Tiles), but you're right, no Image program allows the user to save pngs/jpgs that big x_x

@McTwist

I'm trying to do that right now :tongue.gif: Thanks for your help :3

My map was that big (35km) as I did a lot of exploring in basically one direction, because of a very large ocean, only about 10% of the area was filled in overall.

What were the dimensions of your worlds 7g and 10g?

BTW: I said '20*30' should have been '10*30' for 300megs (was thinking of the possible 20 sub maps if divided)

Is this updated for 1.0.0 yet?
Great progress.

Quote from C4C3

Is this updated for 1.0.0 yet?
Great progress.

Quite. I only need to add the new 1.0 blocks and then it will work as expected. As said in previous post, I've only managed to release a nightly build, but I hope to release a stable version sooner or later. Currently I cannot continue developing as my computer is broken. Still, on Christmas holiday I though of retrieving the project from the HDD and maybe put some effort in it.

I love PixelMap, been using it for ages. Was particularly good for finding strongholds, just use cavemode and look for the purple blocks! Thought I'd drop in and offer my encouragement while I'm on and have an account. I remember you once sought feedback, here was one that would have but didn't, and there must have been more.

I even made a java program that takes images from pixelmap and re-interprets the colours to assist with mine planning, but I'm not releasing it until I get around to resolving that dependancy/complication, and I decide whether such a tool is constitutional. Gutted that someone's already called something geocraft too!

Cheers, and good luck, I'll update when the time comes, but luckily your sturdy program doesn't necessitate such actions.

Thank you for giving your support! I am not sure what sort of functionality you're looking for, so could you please give a further explanation on the subject?

In other news, I haven't been able to update due of lack of computer. My computer died a month ago and I haven't been able to get my hands on a new one. Hopefully I will be up and ready until the end of this month.

Any word on updates? It's been a few months, and I've really been looking forward to seeing what you were going to do with this. Plus, there just isn't a good option for saving a top-down image of a 1.2.4 map right now.

I've been busy. Seriously busy. I hope I will continue to maintain this in two or three weeks, but it all depends on if I get other more important things that needs to be done.

This tool is great! I've tried to add anvil support but I can't seem to build on Windows with the libpng binaries included in the source tarball, and I'm having trouble building libpng and zlib myself.

Not complaining, just explaining. At any rate, here's the changes to the source that I think would work, but can't test. Maybe someone else will have more luck.

I tried to change as little as possible, keeping almost everything in AnvilChunk.hpp and AnvilChunk.cpp. Just thought I'd post what I had, in case it's at least a little useful to someone else.

GUI.hpp, line 37:

GUI_ANVIL,

GUI.cpp, line 48:

Add(RADIO, 175, 5, 65, 20, GUI_ANVIL, "Anvil");
Set(GUI_ANVIL, TRUE);

Level.hpp, line 11:

#include "AnvilChunk.hpp"

Level.cpp, line 47:

else if (chunk->pref.version == BETA || chunk->pref.version == ANVIL)

Level.cpp, line 182:

  if(prefs.version == BETA)
  {
   if(file.length() > 3 && file.substr(file.length() - 3, 3).compare("mcr") == 0)
   {
	files.push(file);
	total++;
   }
  }
  else if(prefs.version == ANVIL)
  {
   if(file.length() > 3 && file.substr(file.length() - 3, 3).compare("mca") == 0)
   {
	files.push(file);
	total++;
   }
  }

Level.cpp, line 208:

if (prefs.version == BETA || prefs.version == ANVIL)

Level.cpp, line 293:

  else if (prefs.version == BETA || prefs.version == ANVIL)

Level.cpp, line 302:

	if(prefs.version == BETA)
	 temp = new Chunk();
	else if(prefs.version == ANVIL)
	 temp = new AnvilChunk();

Level.cpp, line 356:

bool useRegion = (prefs.version == BETA || prefs.version == ANVIL);

Parameters.cpp, line 304:

gui->Set(GUI_ANVIL, worldVersion == ANVIL);

Parameters.cpp, line 733:

else if(gui->Get(GUI_ANVIL, B))
{
  worldVersion = ANVIL;
}

Parameters.cpp, line 960:

else if (worldVersion == BETA || worldVersion == ANVIL)

ChunkPrefs.hpp, line 19:

const int SECTIONBLOCKS = MAPX * MAPY * 16;
const int MAXSECTIONS = 256 / 16;

ChunkPrefs.hpp, line 45:

ANVIL

nbt.hpp, line 26:

TAG_IntArray,

nbt.hpp, line 46:

  int * ias;

nbt.hpp, line 78:

int ReadIntArray(int **);

nbt.cpp, line 152:

  case TAG_IntArray:
#if SAFE_MODE < 2
   break;
#endif
#if SAFE_MODE >= 1
   delete[] tag->ias;
#endif
   break;

nbt.cpp, line 242:

  case TAG_IntArray:
   tagdata->size = ReadIntArray(&tagdata->ias);
   break;

nbt.cpp, line 344:

// Read an int array
int NBT_Reader::ReadIntArray(int ** p)
{
   int length = 0;
   ReadInt(&length);
#if SAFE_MODE >= 2
   *p = new int[length];
   memcpy(*p, pointer, length * sizeof(int));
#else
   *p = (int*)pointer;
#endif
   pointer += length * sizeof(int);
   return length * sizeof(int);
}

Chunk.hpp, line 42:

virtual int GetHeight(int x, int y, int z = MAPZ, int minZ = 0);
virtual Color GetColor(int x, int y, int z, bool side = false);

Chunk.hpp, line 78:

protected:
virtual void Allocate();
virtual void Deallocate();
virtual bool ParseNBT(void *);

Chunk.hpp, line 98:

virtual void PassExclude(int & x, int & y, int & z, Color & color);

Chunk.hpp, line 105:

virtual void PassNight(int & x, int & y, int & z, Color & color);

AnvilChunk.hpp:
http://pastebin.com/K2wHqznZ

AnvilChunk.cpp:
http://pastebin.com/b6i9qHYs

I feel happy that there's people that still care about this software. I still haven't gathered enough time to implement Anvil, but I'm working on it!
It is quite interesting edits you made there, but I sure wonder what made your compiler not wanting to compile the source I provided. May you please tell me how you are compiling it(Program and version) and what errors is made? I could at least try help you out on that so you could release your modified version to be used for the community for now.

Quote from McTwist

I feel happy that there's people that still care about this software. I still haven't gathered enough time to implement Anvil, but I'm working on it!
It is quite interesting edits you made there, but I sure wonder what made your compiler not wanting to compile the source I provided. May you please tell me how you are compiling it(Program and version) and what errors is made? I could at least try help you out on that so you could release your modified version to be used for the community for now.

I don't think I had any problems with your program. PixelMap was crashing somewhere inside libpng. From searching Google, it looks like the solution is building libpng and zlib myself using the same version of Visual Studio / configuration.

I'm not explaining it very well, but I'll try some more later to see if I can get a Windows binary posted for people to try.

The Win32 binaries for libpng and zlib I compiled myself right through the source I downloaded from their site. The only thing I had to change was what sort of binary to create: dll debug, dll, static debug or static. It could be that your problem is that you're using a too old, or too new, version of Visual Studio. I am using VS 2008, as I never got VS 2010 working with PixelMap for some strange reason.

I got it to compile and run, had to change quite a bit with the code I posted previously. However, the program only seems to run if I start debugging from inside Visual Studio. If I try to run PixelMap manually, it just crashes.

Once again, I'm not complaining or criticizing, I'm just not doing a very good job of figuring this out. :/ So, here's the source code that should work if anyone else wanted to figure out how to build it. http://www.mediafire.com/file/6xe2pdeoef1xh3n/PixelMap20120414.zip.

If you do end up implementing Anvil support, I'd recommend not using what I wrote, at least not without refactoring it completely. My code is quite a mess, unfortunately. Sorry I wasn't more help!

I got it to work, but I haven't got enough time to test it with Anvil already. However, I will provide you with the information you and others require to successfully compile and run the program. I will also provide with a download here for those interested in it.
PixelMap Anvil

Now over to the errors I found. First off, I couldn't compile right after I opened it up, as there was still some dependencies that was not added to the project, therefor I added them.
Secondly, when I tried to run the program it crashed. This was because the dll libraries was in an another folder. I moved them to the proper folder and tried again with success. The same goes for the release dll's, so those was added in the same folder as the program.

To summarize what caused the error when trying to run the program is that due it have two dependencies, zlib and libpng, on Windows they are left in respective dll. When debugging, these dlls first needs to be debug dlls, and secondly place in the same folder as the project itself(Root folder). When you're making a release you need to move those specified dlls to the same folder as the program. In some cases it is enough to leave them in System32, but that's a different story...

I really appreciate your time, McTwist. I gave it another go and managed to get libpng, zlib, and PixelMap to play nicely using mingw instead of Visual Studio. I identified and fixed a few more bugs in what I had written too. At any rate, here's a build for Windows along with the updated source. http://www.mediafire.com/file/pj5it37hkp5w57m/PixelMap20120415-bin.zip. Now I'm actually done, and wish you the best.

Please try this version, it shouldn't have that dependency: http://www.mediafire.com/file/pj5it37hkp5w57m/PixelMap20120415-bin.zip