velinath

I see you've already got this where you need it, but I wanted to offer just direct management through systemctl as an option too for other folks that may visit here in the future. I did some stuff with CI for a server a couple years ago and the compile.sh script in this repository includes service scripts. They should still work.

I just rebuilt my compile server on Debian 9 and it works well. Using the standard installed packages in the wiki, you won't need to modify your build script at all. Aokromes: Very much appreciate the update to the wiki page, that helps to clarify.

I'm running into the same issue as above. Removed Boost 1.58 and installed 1.65.1 to see if that'd have any effect, but no dice. Compiling using Clang 5.

For reference - for others doing this, I only needed to install libboost-all-dev libmysqlclient-dev mysql-client to run TC on its own application server. This can be considered solved.

Trying to get back into messing with TC a bit more, and I'm planning to actually set up proper CI and keep things more separated between a build server and an application server. As part of that, I want to make sure I'm packaging what I need and nothing else for deployment. What do I need - or not need - installed on the application server, assuming that I'm moving over binaries from a build server?

Cool. I'll hop in and mess around with the CentOS/RHEL install instructions in the wiki then. I've got a dev Fedora box that I'll stick Clang on and give it a shot. Appreciate the advice and help! edit: After testing on my Fedora box, all that needs to happen is: yum remove gcc gcc-c++ && yum install clang and clang will just start working, with no further changes. It's also way faster. Thanks again for the advice.

Clearing out the build directory seems to have done it. There was stuff in there from as far back as June of 2015, so immediate assumption is that it had something to do with old stuff that needed to get rebuilt, or something that cmake 2 left behind. I'll add a full clear of the build directory, not just clearing the one file, to the build script. Thanks very much for the help on this one. I was banging my head against it for hours and didn't think to try the basic "clear out the old stuff" step. Really appreciated. I don't have a wiki account, but the RHEL docs could be updated - in the interest of ease of use - to: yum install epel-release -y && yum install cmake3 rather than what's there. Plus, that way, no need to build cmake 3 from source. As a side note - and now is when I'm starting to stray from original scope - but as someone who has zero experience with clang as a compiler, is it worth making the switch?

Having issues with latest ( b82ffb8 at the time of writing). My compile script can be found here, starting on line 57 and ending on line 68 for the purposes of relevance. I looked at #14657, which seemed to be someone with the same issue, but no love there. I'm not running Arch. CentOS 7.2 x64, cmake3 version 3.5.1 (I also have cmake 2.8.11 installed, but I'm explicitly referencing cmake3 in my build script), gcc 4.8.5. I successfully built the server on March 15th and this is the first time I've tried since (my last successful build was 632e667). SELinux is set to permissive. When running make, I get the following error: [ 5%] Building C object dep/jemalloc/CMakeFiles/jemalloc.dir/src/arena.c.o In file included from /home/XXXXX/TrinityCore/dep/jemalloc/include/jemalloc/internal/jemalloc_internal.h:568:0, from /home/XXXXX/TrinityCore/dep/jemalloc/src/arena.c:2:/home/XXXXX/TrinityCore/dep/jemalloc/include/jemalloc/internal/util.h:88:49: error: expected ';', ',', or ')' before 'nptr' uintmax_t malloc_strtoumax(const char *restrict nptr, make[2]: *** [dep/jemalloc/CMakeFiles/jemalloc.dir/src/arena.c.o] Error 1 make[1]: *** [dep/jemalloc/CMakeFiles/jemalloc.dir/all] Error 2 make: *** [all] Error 2

Latest TrinityCore revision (6.x) adds these settings to the worldserver.conf. Hasn't been merged into 3.3.5 yet but if you add those settings manually it will work. Here are the relevant settings: TrinityCore/TrinityCore@e5c9997 For the glyphs, can you link your per-class weights from worldserver.conf? Maybe someone will see something there. The rest...I'm not enough of an expert to help unfortunately.

Spent some time looking through the source for mentions of reputation with ID 1119 or just the name "Hodir" in files, but no luck as far as understanding this. I feel like I'm not grasping something obvious here, but it just doesn't make sense to me. I haven't found any other faction that behaves like this in rep calculations. Are there any other factions that behave like this?

Ok, got it. So why do other reputations define 0/3000 Neutral as 0, and seemingly only Sons of Hodir uses the behaviour that you're describing? I guess that's what I'm most confused by, since I expected to see consistent behaviour across all factions in the table. edit: clarity

Sorry, don't understand - what lower reputations? I didn't think there were prerequisites for that reputation, if that's what you mean. I looked into it a little further and 0/3000 Neutral for Hodir is 42000, but 0/3000 Neutral for every other reputation is 0. I apologize for not understanding! :(

Hello! I had a question regarding the Sons of Hodir reputation entry in the database. It didn't seem like an issue since it's working correctly in game, so I didn't want to open anything on GitHub, but I do want to understand how it works just so I can wrap my brain around it. On my server, I have a character who is somewhere into Honored with the Sons of Hodir. When I look in the character_reputation table in the database, though, their faction standing is set to 51175, which if I'm not mistaken is somewhere above the reputation cap. Other factions seem to work as I would expect them to, which is why I'm confused. Am I just misunderstanding how reputations are calculated or am I missing something obvious here?