0.4.6 RC1 play testing and packaging

Describe your experience with the latest version of FreeOrion to help us improve it.
Forum rules
Always mention the exact version of FreeOrion you are testing.

When reporting an issue regarding the AI, if possible provide the relevant AI log file and a save game file that demonstrates the issue.
Message
Author
raptor
Space Squid
Posts: 75
Joined: Sat Jun 02, 2012 11:29 pm

0.4.6 RC1 play testing and packaging

#1 Post by raptor » Sat Sep 10, 2016 5:50 pm

Hi,

I've packaged (for openSUSE) and played 0.4.6 RC1 and I have come across only two possible issues so far:
  1. With packaging, RPATH problems have always seemed to crop up, but I think I've found the simplest solution now - a simple patch to the main CMakeLists.txt:

    [The extension diff has been deactivated and can no longer be displayed.]

  2. A minor graphical issue where a dialog does not seem to follow the same button style as the rest of the game. See this graphic:
    snapshot21.png
    snapshot21.png (349.91 KiB) Viewed 727 times
Otherwise looks good!

User avatar
adrian_broher
Programmer
Posts: 1072
Joined: Fri Mar 01, 2013 9:52 am
Location: Germany

Re: 0.4.6 RC1 play testing and packaging

#2 Post by adrian_broher » Sat Sep 10, 2016 7:58 pm

Could you clarify the issues you have with the current RPATH?

Doing a regular install with the build as it is nows sets the RPATH properly for me.

E.g.

CMAKE_INSTALL_PREFIX=~/Projekte/freeorion/freeorion.install

Leads to the RPATH:

~/Projekte/freeorion/freeorion.install/lib/freeorion


$ ls ~/Projekte/freeorion/freeorion.install/lib/freeorion

libfreeorioncommon.so libfreeorionparse.so libGiGiSDL.so libGiGi.so
Resident code gremlin
Attached patches are released under GPL 2.0 or later.
Git author: Marcel Metz

raptor
Space Squid
Posts: 75
Joined: Sat Jun 02, 2012 11:29 pm

Re: 0.4.6 RC1 play testing and packaging

#3 Post by raptor » Tue Sep 13, 2016 1:50 am

Hi,

This is the error I get in the build system console:

Code: Select all

[  118s] -- Installing: /home/abuild/rpmbuild/BUILDROOT/freeorion-0.4.6rc1-1.1.x86_64/usr/lib64/freeorion/libfreeorioncommon.so
[  118s] CMake Error at cmake_install.cmake:49 (file):
[  118s]   file RPATH_CHANGE could not write new RPATH:
[  118s] 
[  118s]     /usr/lib64/freeorion
[  118s] 
[  118s]   to the file:
[  118s] 
[  118s]     /home/abuild/rpmbuild/BUILDROOT/freeorion-0.4.6rc1-1.1.x86_64/usr/lib64/freeorion/libfreeorioncommon.so
[  118s] 
[  118s]   No valid ELF RPATH or RUNPATH entry exists in the file;
In the past, I would set CMAKE_SKIP_BUILD_RPATH to TRUE and then add in an ld.so.conf file in the system, in order to get around this problem. Now I just patch the CMakeLists.txt.

I suspect that this may be an issue with the particular version of openSUSE + CMake, but I'm not sure. In any event, the default for CMAKE_BUILD_WITH_INSTALL_RPATH is already FALSE, so maybe just removing the line from CMakeLists.txt would be good, and then packagers could override it without a patch.

User avatar
adrian_broher
Programmer
Posts: 1072
Joined: Fri Mar 01, 2013 9:52 am
Location: Germany

Re: 0.4.6 RC1 play testing and packaging

#4 Post by adrian_broher » Tue Sep 13, 2016 7:40 am

I dropped both CMAKE_SKIP_BUILD_RPATH and CMAKE_BUILD_WITH_INSTALL_RPATH from the CMakeLists.txt.

So your problem should be fixed with 6875fd23.


But is still suggest you run the spec build with
CMAKE_BUILD_WITH_INSTALL_RPATH=ON
and leave the ld.so.conf file alone as already suggested in this comment.
Resident code gremlin
Attached patches are released under GPL 2.0 or later.
Git author: Marcel Metz

User avatar
adrian_broher
Programmer
Posts: 1072
Joined: Fri Mar 01, 2013 9:52 am
Location: Germany

Re: 0.4.6 RC1 play testing and packaging

#5 Post by adrian_broher » Tue Sep 13, 2016 7:59 am

https://build.opensuse.org/package/view ... orion.spec

Also some dependencies have changed:
  • no bullet
  • no dejavu-fonts
  • google-roboto-fonts (fedora)
  • cmake >= 2.8.5
  • boost >= 1.54.0
Resident code gremlin
Attached patches are released under GPL 2.0 or later.
Git author: Marcel Metz

raptor
Space Squid
Posts: 75
Joined: Sat Jun 02, 2012 11:29 pm

Re: 0.4.6 RC1 play testing and packaging

#6 Post by raptor » Tue Sep 13, 2016 2:05 pm

Thanks for the dependency changes!

I will make all of your suggested changes to the openSUSE build once 0.4.6 comes out.

Thanks again.

raptor
Space Squid
Posts: 75
Joined: Sat Jun 02, 2012 11:29 pm

Re: 0.4.6 RC1 play testing and packaging

#7 Post by raptor » Thu Sep 15, 2016 3:25 am

Hi,

I believe I found a third issue:

3. A planet with 'temporal anomoly' does not seem to apply the penalty to the 'planet suitability' window:
snapshot01.png
snapshot01.png (418.33 KiB) Viewed 569 times
I have had my exobots die out twice already before I noticed it.

Also, is my issue #2 in the first post a real issue?

dbenage-cx
Programmer
Posts: 389
Joined: Sun Feb 14, 2016 12:08 am

Re: 0.4.6 RC1 play testing and packaging

#8 Post by dbenage-cx » Thu Sep 15, 2016 4:23 am

Hi raptor,

Could you provide a savefile for the temporal anomaly issue? These display fine for me.

Yes, the button style does seem a bit off on that dialog.

User avatar
Vezzra
Release Manager, Design
Posts: 4648
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: 0.4.6 RC1 play testing and packaging

#9 Post by Vezzra » Thu Sep 15, 2016 9:01 am

@raptor: Can you also open an issue on github?

@everyone: This is definitely another one of those bugs we make RCs for, so usually that should be fixed for the release. Which means deferring the next RC until a fix is in etc. However, given how rare that special is and, judging by dbenage's comment, apparently an edge case that doesn't always turn up with that special, and considering that we're beginning to cutting it really close regarding my fast approaching autumn break, we probably should consider alternatives.

So, IMO our options are:
  1. Defer RC2 until this (and other issues tha might come up) is fixed, basically blocking the release for it.
  2. Defer RC2 until tomorrow, and if no fix can be provided until then, just leave the bug as it is and have 0.4.6 with another known issue.
  3. Defer RC2 until tomorrow, and if no fix can be provided until then, comment out the Temporal Anomaly special in the release branch.
I definitely prefer option c. Any other votes on this?

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12268
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

Re: 0.4.6 RC1 play testing and packaging

#10 Post by Geoff the Medio » Thu Sep 15, 2016 9:16 am

What is the other special on that planet? What does the effect accounting for population say if you colonize?

User avatar
Vezzra
Release Manager, Design
Posts: 4648
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: 0.4.6 RC1 play testing and packaging

#11 Post by Vezzra » Thu Sep 15, 2016 1:09 pm

Can't reproduce on OSX. In my brief test game the effect of the Temporal Anomaly special is correctly incorporated into the suitability report, everything works exactly as intended.

We really need that savegame.

User avatar
MatGB
Creative Contributor
Posts: 3310
Joined: Fri Jun 28, 2013 11:45 pm

Re: 0.4.6 RC1 play testing and packaging

#12 Post by MatGB » Thu Sep 15, 2016 1:57 pm

Geoff the Medio wrote:What is the other special on that planet? What does the effect accounting for population say if you colonize?
The other special is one of the tech-created stealth granting one, volcanic ash clouds I think, given both planets have it and there's no slave displaying fairly sure it's from the tech not the monster.

While I haven't tested this right now, in the past when I did the suitability reports were accurate for me.

Hmm, also, suitability reports can get messed up on the turn an evacuation or concentration camp completes, given that's a conquered desert homeworld I'm assuming a Trith population has been removed? That can sometimes cause errors.
Mat Bowles

Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

User avatar
Vezzra
Release Manager, Design
Posts: 4648
Joined: Wed Nov 16, 2011 12:56 pm
Location: Sol III

Re: 0.4.6 RC1 play testing and packaging

#13 Post by Vezzra » Thu Sep 15, 2016 4:59 pm

MatGB wrote:Hmm, also, suitability reports can get messed up on the turn an evacuation or concentration camp completes, given that's a conquered desert homeworld I'm assuming a Trith population has been removed? That can sometimes cause errors.
The planet in question is not the desert one (which is "Eternal Helix Alpha II"), but the radiated one above ("Helix Alpha I").

Nevertheless, I have the sneaking suspicion that you're right on spot here. I just did some testing, and one thing that I can reproduce consistently is that on the turn right after a colony got depopulated (in my test cases the pop died off because not fit for the environment) the suitability report always has been messed up. That clears up after another turn.

So my guess is that raptor depopulated the planets, and on the turn he got the message that the planet had been depopulated, checked the suitability report and queued an Exobot colony building. The colony got established and died off, he got the pop died off message and queued another Exobot colony (maybe after checking the suitability report once again).

The probem is, he might have checked the report right after depopulation each time, and so got always the wrong numbers.

@raptor: Is this was happened?

Assuming it is, I'm inclined to scrap the options I listed above, just accept this as a known issue for now and go ahead with RC2 tomorrow (Friday). Commenting out the Temporal Anomaly special is pointless in this case, as it actually has nothing to do with that special (the special probably just compounded the effect because of the high malus it applies, and thus made the underlying issue more obvious). And blocking the release for an issue that most likely has been around for a long time now and is probably quite difficult to fix is also something I don't want to do.

Is that decision seconded, do you guys object? Otherwise I'll proceed as proposed.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 12268
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

Re: 0.4.6 RC1 play testing and packaging

#14 Post by Geoff the Medio » Thu Sep 15, 2016 5:55 pm

Vezzra wrote:Is that decision seconded, do you guys object? Otherwise I'll proceed as proposed.
Sounds good.

raptor
Space Squid
Posts: 75
Joined: Sat Jun 02, 2012 11:29 pm

Re: 0.4.6 RC1 play testing and packaging

#15 Post by raptor » Fri Sep 16, 2016 4:58 am

Hi,

I am sorry for not responding sooner (school has me busy).

I do not have that save file - I ended up playing a bit afterwards (more than 10 turns) after taking the screenshot. I did do an evacuation of the third, desert planet in the system a few turns (7?) before - it was Trith, and I like to get rid of xenophobic species. However, now the planet shows the appropriate numbers:
snapshot02.png
snapshot02.png (399.39 KiB) Viewed 488 times
I do think I found a 4th issue that I've hit twice now:

4. In some instances the save file I previously used does not show up in the GUI so I cannot load it, yet I can still overwrite it:
snapshot03.png
snapshot03.png (299.27 KiB) Viewed 488 times
I will not do any more turns until I get further instructions this time.

Post Reply