Spring RTS Engine

The filename is correct, the .dylib reported in infolog.txt exists at that exact location, and the double slash shouldn't be a real problem for an unix-like OS.

From the fact that unitsync as used by my lobby is able to find the AI in its proper place, i assume the interface "kind of" works?

Probably some kind of runtime failure, but how do i test/confirm that?

Anarchid wrote:The filename is correct, the .dylib reported in infolog.txt exists at that exact location, and the double slash shouldn't be a real problem for an unix-like OS.

does ldd exist on osx? this maybe gives a hint?!

Googling tells me otool -L is the mac alternative to ldd, so i did the obvious:

$ otool -L libAIInterface.dylib
libAIInterface.dylib:
@loader_path/../0.1/libstdc++.6.dylib (compatibility version 7.0.0, current version 7.21.0)
/usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 1213.0.0)
@loader_path/../0.1/libgcc_s.1.dylib (compatibility version 1.0.0, current version 1.0.0)

Is there anything non-obvious for me to do?

A bit alarming: knorke is complaining about having something like exactly the same problem on Windows in ZKGBAI thread on zk.info: http://pastebin.com/ZcCRnS6b

Anarchid wrote:A bit alarming: knorke is complaining about having something like exactly the same problem on Windows in ZKGBAI thread on zk.info: http://pastebin.com/ZcCRnS6b

your infolog.txt please! very likely it fails at a different stage!

http://pastebin.com/ZcCRnS6b wrote:Microsoft Windows 7 Ultimate Edition, 64-bit

32-bit Spring (AI interface) but only a 64-bit JVM installation, maybe?

your infolog.txt please! very likely it fails at a different stage!

I'll post it here after New Year because i'm away from that machine again. You should have the log in chat history though.

for the reference.

Anarchid wrote:I'll post it here after New Year because i'm away from that machine again. You should have the log in chat history though.

http://hastebin.com/sukoxukule.xml

-> the ai interface fails to load at an early stage.

also possible ai interface logging is broken for release builds, mantised: https://springrts.com/mantis/view.php?id=5020

Anarchid wrote:A bit alarming: knorke is complaining about having something like exactly the same problem on Windows in ZKGBAI thread on zk.info: http://pastebin.com/ZcCRnS6b

interessting choice of words. If http://zero-k.info/Forum/Thread/20617?p ... 852#148852 reads like complaining to you then I better avoid this appearently touchy topic.

I'm using Mac OS 10.8
Can somebody offer a build?
If not, is somewhere a guide how to do that myself?

Tamriel wrote:...
If not, is somewhere a guide how to do that myself?

Is this page that I found in the help pages of any help for your case?
https://springrts.com/wiki/Spring_on_MacOSX

Sorry, I don't have 10.8 on any machine to build from. As for build instructions.. I don't know, the whole build process for osx apps is fairly broken. I actually made my own build script thats great for my own environment. If you are familiar with bash and know your way around a package manager like macports, I'd be happy to show it to you.

Is there a build of the 101.0 RC for OS X I can play with for testing purposes?

I put a 10.11 build of Spring revision 100.0.1-1183-g1081d06 on the first post. Not sure if that is the commit that coincides with RC3.

I have a bundle for 101.0 up (OSX 10.11+)

Hey Darklord42, congratulations on the packages, great job.

So, engine 100.0 worked flawlessly, except for AA AI crashing after one computer dies, but alas 101.0 build is barely working. I couldn't get any AI working (except for the null AIs) all the ones I tried crashed the game right after load.

The shaders seem to have changed (F4 leaves almost the whole map under fog, for instance), but some basic things like the faster-rate LOS update are simply not there, which makes me wonder if there were some old files merged with the new ones?

Faster-rate LOS update requires OpenGL 3+. Mac OS X doesn't have OpenGL compatibility profiles for 3+ version and thus operates only in 2.1.
See https://www.opengl.org/wiki/Core_And_Co ... n_Contexts, "Platform Issue (MacOSX)" section.
Also Mac drivers are too strict and may result in errors like "invalid framebuffer" if something done not in apple way.

Hmm, I saw some trolling about Vulcan, maybe it is better to rewrite the renderer in Heavy Metal? :)

lamer wrote:
Hmm, I saw some trolling about Vulcan, maybe it is better to rewrite the renderer in Heavy Metal? :)

No.

Metal?? Won't happen, I know. In any case, 101's rendering / shader issues on OSX (not to mention the AI crashes, probably unrelated) are so grave right now that I wouldn't recommend it to any player, except maybe for watching replays.

Just sad.

I'll just hope the mods keep supporting 100. BA 9.35 for instance only officially supports 101, although I did get it running in 100.

Hi darklord42!

Can you make a build from develop branch? I would like to test/report bugs.
I have some crashes in 101.. but I guess it's too late to report 101 bugs because 102 is RC already

Wow 102 already! Ok give me a little bit I'll have it up for you.
[edit] done!

This is the most current code, as of writing, since there isn't a rc1 tag
http://www.mediafire.com/download/2ba7d ... 0.11.3.dmg