Skirnir

Ah damn, completly missed faramirs commit. Patman was right, I had forgotten to reextract my maps. Thanks for helping.

I have only the express version of visual studio 2010 and no debug option started when MoveMapGen.exe crashed. Maybe a missing feature from the professional version? Anyway extraction works now :-)

Hm I can't make it (the MoveMapGen.exe build from commit 33f31e27894b7f459b65) work with Win Vista 32 bit. If I run the MoveMapGen.exe with "--skipContinents false" it almost instantly crashes, extracting map 0 (I have no clue how to debug with Windows, there's no gdb :-) ). Does one have to use a 64 bit version, or is there anything I can do?

Regards

Skirnir

Github network graph is back on track :-O

Tested it just a minute ago.

:-)

Hi faramir118,

I just build the generator with your CMakeLists.txt. So far well done :-) But if you run it, it just says

'maps' directory is empty or does not exist
Press any key to close ...

Thanks again for all the work you and everyone else put into this great project. Now I'm waiting for a generator that doesn't crash with continents 1,530 and 570. :-) But take your time ...

Yours sincerely

Skirnir

I tried a bit more with CMakeList.txt the last errors remaining are:

make -j8
[ 15%] [ 29%] [ 33%] [ 41%] [ 46%] Built target libzlib
Built target libdebugutils
Built target libmangosrecast
Built target libmangosdetour
Built target libmangosvmaps
[ 91%] Built target libg3dlite
Scanning dependencies of target generator
[ 93%] [ 94%] Building CXX object CMakeFiles/generator.dir/MeshLoaderObj.o
Building CXX object CMakeFiles/generator.dir/InputGeom.o
Linking CXX executable generator
CMakeFiles/generator.dir/InputGeom.o: In function `InputGeom::loadMesh(char const*)':
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/InputGeom.cpp:128: undefined reference to `myMeshLoaderObj::~myMeshLoaderObj()'
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/InputGeom.cpp:134: undefined reference to `myMeshLoaderObj::myMeshLoaderObj()'
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/InputGeom.cpp:141: undefined reference to `myMeshLoaderObj::load(char const*)'
CMakeFiles/generator.dir/InputGeom.o: In function `InputGeom::load(char const*)':
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/InputGeom.cpp:187: undefined reference to `myMeshLoaderObj::~myMeshLoaderObj()'
CMakeFiles/generator.dir/InputGeom.o: In function `~InputGeom':
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/InputGeom.cpp:119: undefined reference to `myMeshLoaderObj::~myMeshLoaderObj()'
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/InputGeom.cpp:119: undefined reference to `myMeshLoaderObj::~myMeshLoaderObj()'
collect2: ld returned 1 exit status
make[2]: *** [generator] Error 1
make[1]: *** [CMakeFiles/generator.dir/all] Error 2
make: *** [all] Error 2

The CMakeList.txt is here: http://gist.github.com/433247

I'm not quite sure if that's anywhere near the "final" solution. Maybe someone with a bit more experience than myself could have a look?

Yours

Skirnir

I started playing around with a Linux Makefile and for whatever kind of reason I felt like lets try it with cmake. I created this CMakeLists.txt. I had to change "platform/Define.h" in MMapCommon.h to "Platform/Define.h", but still I only get a compile error.

[ 25%] Building CXX object CMakeFiles/generator.dir/MapBuilder.o
In file included from /home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MapBuilder.cpp:3:
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MMapCommon.h:13: warning: scoped enums only available with -std=c++0x or -std=gnu++0x
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MapBuilder.cpp: In member function ‘void MMAP::MapBuilder::loadVMap(uint32)’:
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MapBuilder.cpp:258: error: ‘class VMAP::VMapManager2’ has no member named ‘getInstanceMapTree’
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MapBuilder.cpp:263: error: ‘class VMAP::StaticMapTree’ has no member named ‘getModelInstances’
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MapBuilder.cpp:277: error: ‘class VMAP::ModelInstance’ has no member named ‘getWorldModel’
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MapBuilder.cpp:282: error: ‘class VMAP::WorldModel’ has no member named ‘getGroupModels’
/home/skirnir/Source/m/mmaps_rewrite/contrib/mmap/src/MapBuilder.cpp:303: error: ‘class VMAP::GroupModel’ has no member named ‘getMeshData’
make[2]: *** [CMakeFiles/generator.dir/MapBuilder.o] Error 1
make[1]: *** [CMakeFiles/generator.dir/all] Error 2
make: *** [all] Error 2

Gave it a try yesterday night (yeah had only 4 hours of sleep ...), so far: works as expected. Super fast vmap extraction, vmap size of ~500MB. Liquid type works, server worked smooth, thumbs up.

Tested with Debian Testing x86_64 i7, 4gb ram.

Regards

Skirnir

Edit: I forgot to say, I merged this branch with mangos master: http://github.com/Lynx3d/mangos/commit/0be80326d330f995e70536a03e28a6a17c390e05

@faramir118: Thanks, I'm compiling, but you forgot to add theses lines to configure.ac: manual patch for configure.ac

   src/shared/vmap/Makefile
+   src/shared/pathfinding/Makefile
+   src/shared/pathfinding/Detour/Makefile
+   src/shared/pathfinding/Recast/Makefile
  src/shared/SystemConfig.h

Btw. Thanks for your work, I highly appreciate it =-)

Regards

Skir.

Got a compile warning and error with Debian Lenny x64 and gcc (Debian 4.3.2-1.1) 4.3.2

http://paste2.org/p/778552

Regards

Skirnir

Read gigelfs comment, searched through the code. Quickly ... ~1 minute:

// Added by AHBot, Unknown use (Commented)
   /*if(sWorld.getConfig(CONFIG_ALLOW_TWO_SIDE_INTERACTION_AUCTION))
       AH->auctioneer = 23442;
   else
       AH->auctioneer = GUID_LOPART(auctioneer);*/

You commented a line you didn't understand. That was (a) a bad idea and (b) causes at least gigelfs trouble, don't know about the others, stopped following this thread. That's why I don't know if uncommenting will solve all your problems. Haven't checked the rest of your branch, as I still use the old Naicisum branch with MinimalLoad modification by chucky.black.

Skirnir

XTZGZoReX: I was going to try this patch today, but I ran into a compile error:

../../../src/game/MapManager.cpp: In member function ‘void MapManager::Update(uint32)’:

../../../src/game/MapManager.cpp:417: error: no matching function for call to ‘MapManager::_deleteMap(std::tr1::__detail::_Hashtable_iterator<std::Pair<const unsigned int, Map*>, false, false>, std::tr1::unordered_map<unsigned int, Map*, std::tr1::hash<unsigned int>, std::equal_to<unsigned int>, std::allocator<std::Pair<const unsigned int, Map*> > >&, NULL)’

../../../src/game/MapManager.h:119: note: candidates are: void MapManager::_deleteMap(std::tr1::__detail::_Hashtable_iterator<std::Pair<const unsigned int, Map*>, false, false>&, std::tr1::unordered_map<unsigned int, Map*, std::tr1::hash<unsigned int>, std::equal_to<unsigned int>, std::allocator<std::Pair<const unsigned int, Map*> > >&, ACE_Thread_Mutex*)

make[3]: *** [MapManager.o]

I had to patch World.h (a config option was added) and MiscHandler.cpp (a comment was added) manually.

So I am following instructions exactly as posted and it goes wrong and now I'm being flamed for trying.

I suppose as long as you are ok Skirnir, stuff everyone else who isn't a mastermind of git.

[...]

I won't waste any more of your time as Skirnir has pointed out I am posting to big-up myself or something.

Ah cmon, first it was a joke and later it wasn't even directed at you. Just try watching what the commandline says. sometimes it's not that wrong :-) And that wasn't even a "git mistake" with svn it woulda been the same. It's just 50% of this thread is filled with how can I apply this patch, how do I configure it. If the posts inside this thread rly don't help you, get on IRC and I'll guide you personally.

As you said "have fun" ;-)

Skirnir

@xeross155: If your really wondering what I was referring to, I'll explain it to you (I know your busy to post a reply in every thread at this forum ;-) ). EARTHWALKER checked out your repository to C:\\mangos\\mangos. That was told him by his console output. But instead of changing to that dir, he stayed in C:\\mangos and just blindly followed the given instructions. Now the comandline output told him, that there is no git repo in the directory where he is. But instead of reading the comandline (and realizing his mistake) he just posted it here to blow this thread some more. Just having a closer look to what he is doing might have saved us from this little discussion.

Regards

Skirnir

Initialized empty Git repository in C:/mangos/mangos/

c:/mangos

[...]

fatal: Not a git repository (or any of the parent directories): .git

Don't give up on yourself, start learning how to read today! ;-)

So Mr. Derex, three weeks have passed, is it time to release? Or are you still waiting for Windows/FreeBSD tests?

Regards

Skirnir

I think they removed it because nothing used MinimalLoadFromDB anymore, can't be too hard to use the normal loadFromDB in ahbot.

Feel free go show us the way ;-) As said just some posts ago my C++ knowledge is ... well there ain't no C++ knowledge. What you suggest sounds like the way MaNGOS devs feel would be acceptable for the master branch. Look at Neo2003's comment in this thread or VladimirMangos's (iirc last) comment in the old ahbot thread. Up to this day no one, not even Naicisum, has done something in this way. The easiest way right now, would be to modify minimalload to use the new fields of characters table, instead the data blob. Sadly I feel that I can't do it. So from my point of view right now it's farewell ahbot ...

Regards

Skirnir

it still works if you don't remove the data blob yet.

Somehow it won't start filling my auction house any idea why that happens ?

hi xeross155,

the thing is, MaNGOS devs removed the data blob for reasons (maybe beyond my horizon, but they have some). That whole minimalload thing was removed since no one used it anymore or maybe even ever and I don't like carrying around all that stuff. Maybe it'll work for some weeks or months, but then you need that new structure for something else, like dual talents whatsoever (I don't know, maybe you never need it ...), then you will start looking for another ugly solution. I personally just want to avoid that.

If your tables aren't filled, you either have messed up with your database configuration or with your mangosd.conf file. Import ahbot.sql files from the patch, play with some of the configuration options, create an account, log into world once, fill the account information inside your mangosd.conf file and restart the server.

Regards

Skirnir

sLog.outError("Player #%d have broken data in `data` field. Can't be loaded for character list.",GUID_LOPART(guid)); <- this one, of course filled out. "Player ..."

and with numbers I ment this part:

- m_name = fields[2].GetCppString();

+ m_name = fields[1].GetCppString();

- Relocate(fields[3].GetFloat(),fields[4].GetFloat(),fields[5].GetFloat());

- SetLocationMapId(fields[6].GetUInt32());

+ Relocate(fields[2].GetFloat(),fields[3].GetFloat(),fields[4].GetFloat());

+ SetLocationMapId(fields[5].GetUInt32());

// the instance id is not needed at character enum

- m_Played_time[PLAYED_TIME_TOTAL] = fields[7].GetUInt32();

- m_Played_time[PLAYED_TIME_LEVEL] = fields[8].GetUInt32();

+ m_Played_time[PLAYED_TIME_TOTAL] = fields[6].GetUInt32();

+ m_Played_time[PLAYED_TIME_LEVEL] = fields[7].GetUInt32();

obvious it didn't work out as expected :-)

I did and I didn't run into crashes like before, but I get a constant warning ... some lines down Player.cpp Minimal Load function. I removed that part and changed numbers according to numbers of characters table and now it instantly crashes.

As said before compilation takes some time, windows and precompiled headers can't beat 'em. I won't mess around with code I don't understand, maybe I'll have a deeper look tomorrow.

Well guys, this commit: this breaks ahbot. At my console this pops up before a crash:

SQL: SELECT guid, data, name, position_x, position_y, position_z, map, totaltime, leveltime, at_login, zone, level FROM characters WHERE guid = '30'
query ERROR: Unknown column 'data' in 'field list'

this is caused by removing the 'data' blob in mangos. Minimal load function (ahbot uses it, but mangos already removed it's code a while ago) searches for the data table.

{
       //                                        0     1     2     3           4           5           6    7          8          9         10    11
       result = CharacterDatabase.PQuery("SELECT guid, data, name, position_x, position_y, position_z, map, totaltime, leveltime, at_login, zone, level FROM characters WHERE guid = '%u'",guid);
       if (!result)
           return false;
   }

The second field is the "data" blob. Now there is no data blob, what would be required would be, make this line use all now available fields, but I'm on linux compiling takes still "ages" and I'm not quite sure if this would fix the problem or just creates new problems.

With mangos revs higher than 9630 you can't use ahbot any longer.

So far I only see people using highly customized MaNGOS versions having this problems. If you use a core patched by someone else, throwing in dozens of custom patches and you run into a freeze, I'd say it's your very own fault. Patches like ahbot are made for a clean MaNGOS core (w/o any patches). In this thread we have a bunch of users using latest MaNGOS with ahbot w/o problems on different platforms. The way it is right now, I see a direct link between patches and crashes.

Anyway some of you will never agree to this point, it's not that popular to point at oneself if asked for whom to be blamed. So blow up this thread for 3 or 4 pages, complaining about freezes, that way this thread at least can easily be spotted ...

Skirnir

PS: "it freezes 30 - 40 seconds right after the first person logs into the world" this def. doesn't look like a ahbot failure :-\\

@Kreegoth thanks mate, I also don't know whats up. My core is

MaNGOS/0.16.0 (* * Revision 9611 - b951e50c9ad67d0c65d997cd725b381eb85d4036) for Linux_x64 (little-endian)

and everything works fine. My auctionhouse is full. I can buy with alliance/horde in their auctionhouses and in the neutral one. I can bid on items, I can place items for auctions and the ahbot buys them, after some delay time (tested with alliance dwarf in ironforge auctionhouse and in bootybay auctionhouse, you get informed that your item was sold and after some delay time you get mail with money. If your item wasn't sold you get it instantly back. Mail sending isn't broken.) I had my server running at two days for about two hours, well only with myself connected, but absolutely no crash or anything funky caused by Naicisums ahbot.

So I repeat myself, it's either another patch file that kills the server or a database issue, either caused by your database deliverer or by an earlier version of ahbot. So you could clean your auctionhouse table, and your iteminstance table from items that belong to ahbots_account. And as last idea a crash can also can be caused by a hardware failure. Ahbot takes some CPU load and writes a bunch of stuff into mysql. But since not even a mangos cmd log or anything else was handed in ... what shall we do? My crystal ball is on reparation ... :-)

I used Naicisum's repo and no patches for ahbot but now I do finnaly have a problem with the latest core rev, ahbot is crashing the server...the Developers have finnaly figgured out how to break it.

Are you 100% sure that crashes come from ahbot? I'm rather bad with C++ but I couldn't see anything changing or affecting ahbots behavior. My server with MaNGOS 9609 runs smooth. No error messages, no high CPU usage, I even compile wine right now and I have no probs. So if you haven't messed up your ahbot configs I think another custom patch of yours is causing the problems.

Regards

Skirnir

PS: I hope Naicisum will come back, no reason to declare him "gone".

@temporary: m8, this isn't a "I want to learn what is a patch file" or "how can I apply a patch file" thread. Therefor MaNGOS has a special forum section and many questions can be solved with good old google. For this time I wont be rude and just tell you to f*cking google it: The plus in front of a patch file tells the programm "patch" that this line shall be created, a minus tells the programm to delete a line. Now if you open a patch file with an editor you can see a bunch of pluses and minuses with some line without plus and minus. In case of the sql file: the complete sql file has to be created, so every line of this file has a plus in front of it. You can get this sql file manually if you cut out the lines with a plus in front, throw it into a new txt file, and erase this pluses. Afterwards you name that file "characters_auctionhousebot.sql" and have done manually what the programm patch does automated.

If you have any further questions on how to apply a patch file or what a patch file does or can do, please use the appropriate forum section and stop monologizing in this thread.

Hth and thanks

Skirnir

X-Savior: Here is the promised patch: ahbot.patch. I have just applied it to a clean mangos branch and compiled it on Debian Lenny x64 with MaNGOS rev9567. I quickly tested if ahbot inserts items and that part seems to work, so I have no doubt about the rest.

I have only included the character_auctionhousebot.sql file, as the update_character_auctionhousebot.sql file isn't needed (was discussed earlier this thread and if you already have that file applied to your "test" database, nothing has changed!). I've applied the patch by dropping "ahbot.patch" into my mangos source dir, checked out a clean branch and "git apply ahbot.patch". If that doesn't work for you, somehow git sometimes complains about corrupted file endings (I rly don't know why it does so) you could try it with: "patch -p1 < ahbot.patch" ... you know the procedure^^

Hope it works out (else I'd failed again)

Regards

Skirnir