Page 1 of 1

0.4.8 RC: planet suitability check closes game

Posted: Sun Jun 17, 2018 1:49 pm
by jsena
Mac OS 12.6
v0.4.8 2018-06-14.3f22a75
Right-click on a planet for planet suitability, game closes (doesn't crash as far as I can see; crash reporter doesn't come up)

Pretty sure this has been happening for a while, but frankly I forgot to report it!

Re: 0.4.8 RC: planet suitability check closes game

Posted: Sun Jun 17, 2018 3:08 pm
by Geoff the Medio
Doesn't crash for me on master on Windows.

Re: 0.4.8 RC: planet suitability check closes game

Posted: Mon Jun 18, 2018 3:12 pm
by Vezzra
@jsena: I'm on the exact same OSX version as you (10.12.6 Sierra), and I can't reproduce the crash. Pedia window opens just fine.

Can you post the freeorion.log from immediately after such a crash? Preferably open a github issue for this problem and upload the log there.

Re: 0.4.8 RC: planet suitability check closes game

Posted: Mon Jun 18, 2018 7:07 pm
by Vezzra
Vezzra wrote:...I can't reproduce the crash.
Erm, I've to retract that statement. I can reproduce this issue, turned out it's important for which planet you try to open the suitability report first. I tried with my homeworld first, and everything worked just fine. Also opening the suitability report for other (unowned planets) worked. However, when the first thing you do after starting a new game is opening the suitability report for one of the unowned planets in your home system -> crash.

Furthermore, the issue only occurs with builds of the release branch, and apparently only on OSX (I couldn't reproduce the issue on Windows). So, that means the issue only happens if all of the following are true:
  • You do not open the suitability report of an owned colony first
  • You are on OSX
  • You use a build of the release branch
Just peachy. :? Don't you just love bugs like these. Always a joy to hunt down... :(

Issue on github: #2166

Re: 0.4.8 RC: planet suitability check closes game

Posted: Fri Jun 22, 2018 4:17 pm
by Vezzra
Vezzra wrote:You use a build of the release branch
After many hours of trying to narrow down when this bug has been introduced I'm can't say that for sure anymore. The bug turns out to be very elusive, apparently it's harder to trigger on some builds than on others. Which means, just because I haven't been able to reproduce the issue on current master (by following the steps described in the issue on github) doesn't mean it isn't there.