We hope this to be the last testing release before the real release.

See the Testing Release Wiki page for general info about how to obtain the release, and an archive of all testing releases since the last stable one.

Changes

• new versioning scheme RC12 (see viewtopic.php?p=501948#p501948)
• fixed logging on windows
• record WARNING and higher log levels to stderr, normal log to stdout
• fixed some pathfinder issues
• many bug-fixes

Bugs and issues
Report in this thread or on Mantis.
Remember to attach infolog.txt as file, if you crash!

is it just me or the links on the wiki point nowhere (yet) ?

I'm having problems getting this test release to compile on Ubuntu 11.04. Here are the last few lines showing in my terminal when it quits:

I've compiled many, many test releases/masters/develops/etc until about a month ago with no problems. I'll try provide more info if needed. Sorry if this belonged on Mantis, wasn't sure where it was appropriate.

true.. sorry, and thanks!
i fixed the windows links, i think only the app bundle is left now.
i see no app bundle for latest versions. i think it is cause the mac buildbot still uses master (only), but we moved to the develop branch now.
tell me if the bundle is ready, then i'll adjust the link if required.

Well the master branch bundle has been working until now, so the only way to check if develop works is to try it out...
If the final release exe and bundles are made from the master branch once merged, maybe it isn't necessary to switch the Mac slave to develop branch, still, testing latest fixes could be nice to avoid bad surprises.

FYI, the bundle will *not* contain SpringLobby, which will be bundled and made available separately, to allow using multiple versions of spring at the same time.

all buildbots should auto-compile every push to both develop and master.
master will contain only releases, so every commit has to be compiled anyway. develop should be built for the same reason why we did build master every push, so far (cause it is simply Matrix Reloaded ... i mean, master Renamed).

app bundle link fixed too now, thanks daftalx (i did not think about forcing the build :/ ).

I don't mean to clog up the forums if I'm the only one having the issue (maybe I did something dumb ), but I'm still having the same issue compiling that I mentioned in my previous post even after moving to a fresh install of Ubuntu 11.10. It occurs while trying to compile the multithreaded version of spring. I'm positive I successfully compiled the last few test releases including the recent one from September 6 with no problems by using the instructions on the wiki's. Is it something I'm not doing correctly, or is it a reproducible/known problem for everyone right now?

Need an error log or something, otherwise it is impossible to help.

Fyi, i can build (make spring) this version on archlinux which have gcc 4.6.1.

Sorry, I'll try to be more helpful. In the third post of this thread I pasted what was showing up in the terminal when the compiling failed, which I'll try to explain in more detail below...

Using

then

then

This displays in my terminal when it fails (the last few lines):

When building multithreaded instead (after cleaning everything out from the previous attempts) using...

These are the last few lines in my terminal:

My GCC version is 4:4.6.1-2ubuntu5 which I would think is basically 4.6.1, but I am not very familiar with this. I will gladly post the outcomes of any commands or any logfiles you guys might need, I just don't know what would be helpful to you or where I might need to go to get them. Thanks in advance & sorry if this problem was my mistake.

EDIT:
It just occurred to me that it seems to get hung up on files called "GameVersion.cpp" and "VersionGenerated.h". I've noticed the chatter around the forums lately surrounding a new versioning system; perhaps there was a typo in some changes made lately for that?

I get a crash when trying to launch the portable version with ba 7.60 and tropical:
http://pastebin.com/g058y4Q8

[f=0054304]
[f=0055076]
[f=0055453]
[f=0076859]
[f=0104774]
[f=0109879]
[f=0111593]
[f=0111669]
[f=0112946]
[f=0124448]

nothing important, just a path finding thing and the amount of time tanks are spending flying appears to have been increased. overall seems to work faster than the latest release version



units spawn from outside of the map when factory exit is jammed, sometimes they are spawned wrecked

Rename the file VersionGenerated.h.template to VersionGenerated.h.

When I did that, it made it all the way to 72%, but failed with...

so instead of only renaming VersionGenerated.h.template, I copied it and renamed that copy VersionGenerated.h. Having them both in there finally got it to compile. Looking forward to testing this version. You guys have a great project going here, best real time strategy I've ever played. Thanks!

EDIT:
Looks like I celebrated prematurely...Although it complied, I cannot get the game to successfully launch past the main menu using the generated executable or the latest spring lobby. Here is my infolog.txt if it is helpful: http://pastebin.com/3x1UspPr. I noticed glancing through the log that the version number still is not being generated correctly as it is displaying as "Spring @SPRING_VERSION_ENGINE@ (@SPRING_VERSION_ENGINE_ADDITIONAL@ MT-Sim)". I don't mean to use you guys as my personal troubleshooters (sorry about that), but this problem is way over my head.

cmake is supposed to read the actual version info from git, and then configure the template with the actual values, and write the outcome into the build-dir. if that does not work, it is most likely a problem with our CMake files (and yes, it is related to recent version changes, kind of).
We have to find out why the file is not configured for you, or why the respective build dir sub-dir is not in the include paths.
i am may have time later today, but maybe also not.
i'll have little time these days in general.

I'll be glad to try and help in any way that you guys might need me to. I don't have a ton of time during the week either, but I can still check in for a few minutes now and then. Thanks for your replies, hopefully things will get sorted out soon.

Change this to:



(VersionGenerated.h only actually seems to get generated for the default build target with 'make' or 'make all', not for any others)

in all the test versions i tried my fps are lower than what i get in current version. fps are only 1/2 to 3/4 of normal, so it is still playable but noticeable.
Settings are the same (at least tried to put same)
Is because of debug builds or something?

thanks Kloot, i hope it is fixed by this commit (can currently not test)

https://github.com/spring/spring/commit ... 3c05a80231

It works! Just successfully built the latest using: Version from infolog: Made it all the way into the game too...no crashes while launching. Thanks to hoijui, Kloot, and zerver for your time and fixes.