Project maintainer wanted #2

civctp2 · 2018-10-12T22:40:52Z

Contribution to the project are on hold, until maintainer is found.
For details please refer to this discussion: ptitSeb/ctp2#3 (comment)

git checkout 3285b19~1 -- ./ctp2_code/ui/aui_sdl/aui_sdlblitter.cpp gdb: Thread 5 "ctp2" received signal SIGSEGV, Segmentation fault. (gdb) bt #0 0x00007ffff76d9fef in ?? () from /usr/lib/x86_64-linux-gnu/libSDL-1.2.so.0 civctp2#1 0x00007ffff76c96b9 in ?? () from /usr/lib/x86_64-linux-gnu/libSDL-1.2.so.0 civctp2#2 0x00007ffff76de4ea in SDL_LowerBlit () from /usr/lib/x86_64-linux-gnu/libSDL-1.2.so.0 civctp2#3 0x00007ffff76de77a in SDL_UpperBlit () from /usr/lib/x86_64-linux-gnu/libSDL-1.2.so.0 civctp2#4 0x00000000004f2ed7 in aui_SDLBlitter::Blt16To16 (this=0x1f8fb70, destSurf=0x2078a80, destRect=0x7fffe94dfaa0, srcSurf=0x2076830, srcRect=0x7fffe94dfa90, flags=17) at ui/aui_sdl/aui_sdlblitter.cpp:56

…ce.cpp: #0 0x00007ffff725126d in __lll_lock_wait () from /lib/x86_64-linux-gnu/libpthread.so.0 civctp2#1 0x00007ffff724ae42 in pthread_mutex_lock () from /lib/x86_64-linux-gnu/libpthread.so.0 civctp2#2 0x00007ffff77044ce in SDL_mutexP () from /usr/lib/x86_64-linux-gnu/libSDL-1.2.so.0 civctp2#3 0x00000000004f62d1 in aui_SDLSurface::Lock (this=0x1f8eec0, rect=0x7fffffffe8b0, buffer=0x7fffffffe6f0, flags=0) at ui/aui_sdl/aui_sdlsurface.cpp:126 civctp2#4 0x000000000048b39d in aui_Blitter::Blt16To16 (this=0x1f8f2f0, destSurf=0x1f8eec0, destRect=0x7fffffffe8b0, srcSurf=0x1f8eff0, srcRect=0x7fffffffe8a0, flags=17) at ui/aui_common/aui_blitter.cpp:292

#0 0x0000000000c5bbf6 in CityData::SpecialistCount (this=0x0, type=POP_SCIENTIST) at CityData.cpp:8074 civctp2#1 0x000000000063d3ff in CityWindow::WorkerSpinnerCallback (control=0x2bc8950, action=1, data=0, cookie=0x2bc8950) at ui/interface/citywindow.cpp:1330 civctp2#2 0x00000000005784cd in ctp2_Spinner::ActionCallback (control=0x2bc8950, state=1, data=0, cookie=0x2bc8950) at ui/aui_ctp2/ctp2_spinner.cpp:84 civctp2#3 0x00000000004c1069 in aui_Ranger::RepositionThumb (this=0x2bc8950, callActionFunc=1) at ui/aui_common/aui_ranger.cpp:638 civctp2#4 0x00000000004c0789 in aui_Ranger::IncrementUpY (this=0x2bc8950) at ui/aui_common/aui_ranger.cpp:483 civctp2#5 0x00000000004c2be2 in aui_Ranger::ForceScroll (this=0x2bc8950, deltaX=0, deltaY=1) at ui/aui_common/aui_ranger.cpp:1147 civctp2#6 0x00000000004f5994 in HandleMouseWheel (delta=-1) at ui/aui_sdl/aui_sdlmouse.cpp:40 civctp2#7 0x00000000004f5c21 in aui_SDLMouse::GetInput (this=0x1fa6300) at ui/aui_sdl/aui_sdlmouse.cpp:91

DEVoytas · 2018-11-29T17:23:35Z

Since @ptitSeb said in this comment: #21 (comment) that we will not be able to act as a temporary maintainer, I propose two alternative solutions (until real maintainer is found):

either @LynxAbraxas agrees for this role (he crated many very useful PRs so he is most active and motivated person in this project at the moment)
I will start merging pending PRs, but since I (as stated many times before) do not feel authorized to make such decisions, I would only merge those, that we all agree on (but take into account I may have periods of lower activity)

@LynxAbraxas, @ptitSeb what do you think about it? Do yo have any other ideas?

ptitSeb · 2018-11-29T19:41:15Z

I'm ok with that @DEVoytas .

LynxAbraxas · 2018-11-30T22:45:38Z

Yes, pleas go ahead @DEVoytas! My current New Years resolution for 2019 is to not work on this any more (for at least a year). It has infected me again...

LynxAbraxas · 2018-12-21T17:09:51Z

@MartinGuehmann: Would you mind to be maintainer? You have profound knowledge of civctp2 source and of its evolution.

MartinGuehmann · 2018-12-22T02:05:47Z

For sure, I can do this, I just have to get familiar with GitHub.

So the tasks are merging the branches. We have the two mentioned here and there is also trunk it is 11 commits ahead of master. I don't exactly know what these commits are about.

DEVoytas · 2018-12-26T03:24:42Z

For sure, I can do this, I just have to get familiar with GitHub.

@MartinGuehmann that is great! Now we need to figure out how can I pass civctp2 account credentials to you. Maybe email?

PS.
I understand the need of getting familiar with GitHub, but are you more or less familiar with git in general?

MartinGuehmann · 2018-12-26T19:30:38Z

@MartinGuehmann that is great! Now we need to figure out how can I pass civctp2 account credentials to you. Maybe email?

@DEVoytas, I guess email is fine. You can make GitHub display the email address, publically. But of course in the age of spam-bots, that's not a good idea. I can't make GitHub display it only to some people, so let's do it like this:

guehmann
at
t-online.de

PS.
I understand the need of getting familiar with GitHub, but are you more or less familiar with git in general?

I played around with the Git stuff in the last days. So it is more complex than Subvision, but that shouldn't be much of a problem.

The first goal is to bring master into a releasable stage, at least for Windows. Of course master should also compile on Linux, but I haven't managed to do this, yet. It stuck at wldgen.cpp. I haven't checked the compiler. I just used what came with Kubuntu.

Right now, I know that "Now it compile on 64Bits platforms" (Git doesn't have these nice revision numbers) breaks compiling on Windows, but up then everything is fine. This version behaves the same as Revision 1111, if you compile it with VC 2008. The behavior of the AI changes a little on VC 2017 and a few fixes for compiling. But so far I can assume it is just the different build environment. The same happened with compiling on VC6. My feeling was on VC6 the AI behaved a little bit better. Therefore, I always released VC6 builds.

The current head of master is broken. The AI does not behave not only differently, but it is differently in a broken way. It doesn't explore, and it doesn't improve tiles, properly. The road network is broken and no commerce improvements and also some other things. I assume that these intptr _t, maybe a problem, but there were also some other things, I have seen. But that should be easy to sort out.

More difficult are the changes from the Roland Taverner branch. They don't seem to break the AI, but they make the game crash. And that is quite some code that was changed. We have in simplifying the code, which the new standard allows, probably I take this, because these iterators are quite hard to read. Replacing the raw pointers is a bigger task. And I am not sure whether we should do this. Since the debug version has the leak reporter, we should find the leaks.

LynxAbraxas · 2018-12-26T20:15:26Z

So the tasks are merging the branches. We have the two mentioned here and there is also trunk it is 11 commits ahead of master. I don't exactly know what these commits are about.

I think a lot of contributions are similar and originate from the fact that they were done out of sync. Some were on SVN trunk, others on non-deterministic git ports. I worked with the contribs of @ptitSeb because they make it possible to compile master on linux and provide a way to avoid FAT installs which is the main reason that ctp2 can now be not only compiled inside a docker environment but also played from a docker image (DI), which I find most appealing (see my other posts for details).

The intptr _t cased inappropriate game behaviour, see e.g. #33. Reverting back to the former code and using -fpermissive for compilation seems to solve the problem, see #35 and #36.

All issues and PRs reported by me are for the compilation with SDL under Linux inside the docker environment defined by the Dockerfile in https://github.com/LynxAbraxas/ctp2DF.

With the changes comprised of all my PRs master4ctp2DF (some debugging, more additional features), the main problem that I face when playing with the DI version of master, are the crashes appearing at random intervals (sometimes within the first minute after game start, sometimes after hours of playing), see #30.

MartinGuehmann · 2018-12-26T21:03:20Z

Originally, I would have gone for the branch of Roland Taverner, but that one is more difficult to fix. It also makes the code compile under Windows with VC 2017, but the project files are a bit ugly, so I ended up to convert my own VC 2008 project files. Basically for VC 2017 compilation, I also did the same changes as in that branch. It seems to work everything except the intro move, just sound but no picture.

Anyway with the fix of #36, the AI seems to be fine, almost like before, just for minor differences. The max template function uses float, even so it's supposed to be double.

MartinGuehmann · 2019-01-09T19:11:14Z

For sure, I can do this, I just have to get familiar with GitHub.

@MartinGuehmann that is great! Now we need to figure out how can I pass civctp2 account credentials to you. Maybe email?

PS.
I understand the need of getting familiar with GitHub, but are you more or less familiar with git in general?

@DEVoytas, I am still waiting. The alternative of sending me the credentials, you could allow me to commit to the civctp2 repository. Something, I would do this way, anyway.

DEVoytas · 2019-01-16T10:21:14Z

Hi @MartinGuehmann, apologies for the delay. I've just sent account details to your email, so please check if you can log in now.

civctp2 · 2019-01-16T17:54:38Z

@DEVoytas, obviously I can log in, thanks. I added myself as a collaborator. Let's see how that works out.

DEVoytas · 2019-01-17T17:20:38Z

@civctp2 , @MartinGuehmann great! So I guess we can now merge following PR: #56 and close this issue: #2.

MartinGuehmann · 2019-01-18T22:20:19Z

#56 is merged in, and this issue is now closed.

LynxAbraxas · 2019-01-22T08:38:46Z

Thanks @MartinGuehmann for taking the job as maintainer and thanks @DEVoytas for all the initiatives. It's really great to see the ctp2 source code project revived.

@MartinGuehmann: From the commit history I see you have been busy a lot already. Thanks for that. I would recommend though that you also create PRs for the commits you make (create your own fork of civctp2 at https://github.com/MartinGuehmann, push to that and create PRs to https://github.com/civctp2/civctp2/ that can be discussed before being merged at GH by you or civctp2 user) and want to integrate in to ctp2@master such that ideally master only contains PR merge commits, as is the recommended git/GH way of contributing. This way it is easier for other contributors to keep track of what is where and the change can be discussed in its own thread (PR). Currently, I cannot find the video-tutorial (from someone working at GH) I once got those ideas from but you can have a look at https://help.github.com/articles/what-is-a-good-git-workflow/ or https://gist.github.com/Chaser324/ce0505fbed06b947d962, perhaps @DEVoytas has some other great links.

LynxAbraxas · 2019-01-22T08:52:27Z

For example: if 6d817f4 had been in a separate branch for a PR, commit 55ecfc4 would probably have followed directly onto it or both would have been squeezed into one and tested by the build system and only merged into master when the builds succeeded.
If you look at https://github.com/civctp2/civctp2/commits/master you can see that e.g.
55ecfc4
501d56c
f228215
did not compile on the build system (red cross), such commits should not be part of master but be on branch with a PR.
Contribution rules on some projects even demand that every commit on any branch compiles and passes all tests, although I am a fan (so far) for not squeezing all commits to preserve the info of what did not work/suffice, see e.g. #13 (comment).

LynxAbraxas · 2019-01-22T09:30:11Z

@MartinGuehmann Concerning the above, after goning through some of the new commit to master I am beginning to wonder if you accidentally pushed your local testing master to civctp2-master?
For example I only created an issues (#30) but never created a PR for debug-random-segfaults but it got included into civctp2-master here:
fcfefc6
Which is not good because it contains commits from other PRs that I needed to include to test if they were the origin for the random segfaults. Anyway this branch as a PR would be WIP and not ready for merging.

How did you stumble up on debug-random-segfaults?

LynxAbraxas · 2019-01-22T09:41:41Z

If civctp2-master should stay as is, I will test it on GL with https://github.com/LynxAbraxas/ctp2DF.

However it was accidentally pushed I would vote for resetting master to 74c4f07 and then do clean PR merges for the various changes. I do realize that this would be some work considering how busy @MartinGuehmann was ;-) but probably would be worth the effort due to the fact that some things got included that should not be and some multiple times and currently it is very hard to follow through the late history chaos.

74c4f07 was the latest commit when I last fetched https://github.com/civctp2/civctp2 at the beginning of the year.
It seems some of my branches/PRs got merged outside GH such that I got no message about the new status and therefore could not comment on new commits, nor did I get informed by all the great work that was done by @MartinGuehmann lately. Please merge on GH or others wont be able to follow.

DEVoytas · 2019-01-22T19:09:10Z

Thanks @MartinGuehmann for taking the job as maintainer and thanks @DEVoytas for all the initiatives.

Big thanks to both of you as well @LynxAbraxas and @MartinGuehmann as well!

I would recommend though that you also create PRs for the commits you make (create your own fork of civctp2

I agree, actually already mentioned that here: #57

This way it is easier for other contributors to keep track of what is where and the change can be discussed in its own thread (PR)

Also agree. Final decision is always up to the maintainer, but it is always good practice to have some discussion and let others give their opinion about all changes that are meant to be merged.

I would vote for resetting master to 74c4f07 and then do clean PR merges for the various changes.

Normally rewriting the history of a public branch it is very strongly discouraged (see https://stackoverflow.com/questions/11149499/what-happens-when-you-rewrite-history-of-a-public-branch#11149606 for reasons why). However given that not many people has cloned the repo so far (judging by the number of forks on GH) it may actually be worth, it in order to have cleaner history (which does not really look so good right now).
From network view: https://github.com/civctp2/civctp2/network it looks like no one based their work on most recent master, and there is rather low chance that someone cloned the repo and does some work outside of GH.
So on one hand this is really bad practice and the other benefits seem to outweigh the downsides in this case... I cannot make my mind on this so I guess I will abstain from giving my opinion,

MartinGuehmann · 2019-01-22T19:09:47Z

@MartinGuehmann: From the commit history I see you have been busy a lot already. Thanks for that. I would recommend though that you also create PRs for the commits you make (create your own fork of civctp2 at https://github.com/MartinGuehmann, push to that and create PRs to https://github.com/civctp2/civctp2/ that can be discussed before being merged at GH by you or civctp2 user) and want to integrate in to ctp2@master such that ideally master only contains PR merge commits, as is the recommended git/GH way of contributing. This way it is easier for other contributors to keep track of what is where and the change can be discussed in its own thread (PR).

@LynxAbraxas, I created my own fork and pushed the first stuff there. Actually, also fixing my own bug, but well. Right now it is also my master branch, probably also not the best location, but for the start it should be good enough.

MartinGuehmann · 2019-01-22T19:21:11Z

Contribution rules on some projects even demand that every commit on any branch compiles and passes all tests, although I am a fan (so far) for not squeezing all commits to preserve the info of what did not work/suffice, see e.g. #13 (comment).

I wouldn't squeeze anything, either. It doesn't feel like a good idea to tinker around with it. We should keep in mind that this project has a non-Git history. This means not every commit compiles, anyway. So, I would rather demand that only the head passes all tests. The thing is it is about coding and move things, so I would like to put the bureaucracy to a minimum. Obviously, we need some, but it shouldn't be not more then necessary.

MartinGuehmann · 2019-01-22T20:36:33Z

@MartinGuehmann Concerning the above, after goning through some of the new commit to master I am beginning to wonder if you accidentally pushed your local testing master to civctp2-master?

Well, it is my local master. This just reflects the subversion workflow. Where you rather download the things from the server, merge them into your working copy and then upload once you are done. We still have issue #1, which is an issue because it was branched. As long the person who feels responsible for the branch is around and cares, we have no problem the branch won't go out of sync. But once, the person disappears, nobody would keep it in sync, and then things go bad. To merge the Linux branch properly, I think I have to do the brute force method: To go through every file and merge manually what I think is was merging. That is quite some work, so for me it is a nightmare to have branches laying around in this project. Of course, it would have been our burden to make the game compile on Windows again, with the Linux changes that may occasionally break things on Windows. But it would have been the minor costs.

For example I only created an issues (#30) but never created a PR for debug-random-segfaults but it got included into civctp2-master here:
fcfefc6
Which is not good because it contains commits from other PRs that I needed to include to test if they were the origin for the random segfaults. Anyway this branch as a PR would be WIP and not ready for merging.

How did you stumble up on debug-random-segfaults?

As with the things that should have been merged in, I fetched them from the server. With all these pull requests lying around it got a bit confusing. I couldn't merge them on GitHub, either. The debug-random-segfaults is one file to revert, that should be easy to do, because as far as I can see only one file came from it. Now, I have the following in master:

DEVoytas/pm_readme_update #56
LynxAbraxas/activateAllGoods #44
LynxAbraxas/spriteCreation #46
LynxAbraxas/spriteEditor4Goods #45
LynxAbraxas/spriteEditor #41
LynxAbraxas/debugKeyHandling #40
LynxAbraxas/goodMushrooms #47
LynxAbraxas/debug-random-segfault
LynxAbraxas/compile4debug #39 x
LynxAbraxas/revert-intptr_t #36
LynxAbraxas/debugDiplomacy #35
LynxAbraxas/SDLvideo4master #34
LynxAbraxas/mouse-wheel #26
LynxAbraxas/goodsAnimDebug #24
LynxAbraxas/parallelBuild #17
LynxAbraxas/debugLoadGame #18

PR #39 did not build, but I fixed it. And then I have only debug-random-segfault to remove. What I can already say about is that the problem is not there. But it is something that has been fixed in the Linux branch.

On the Linux branch I have:
LynxAbraxas:linux-segfault-debug #25
LynxAbraxas:docker4linux #37

docker4linux already contained LynxAbraxas:linux4ctpDF. So to have these things should be fine, I added on top the VS files for browsing the project on Windows. I assume this is also uncontroversial.

Then I also updated the VS project files on the RolandTaverner's branch to make the game compile according to the manual.

MartinGuehmann · 2019-01-22T21:09:19Z

Obviously, we have to talk about something, so I reopened the issue. Ideally, I would split this part off, but I don't know how.

If civctp2-master should stay as is, I will test it on GL with https://github.com/LynxAbraxas/ctp2DF.

However it was accidentally pushed I would vote for resetting master to 74c4f07 and then do clean PR merges for the various changes. I do realize that this would be some work considering how busy @MartinGuehmann was ;-) but probably would be worth the effort due to the fact that some things got included that should not be and some multiple times and currently it is very hard to follow through the late history chaos.

Well, if you wanna do this. ;) For removing what nobody has requested to include, I would just need to change ../ctp2_code/ui/aui_sdl/aui_sdlblitter.cpp. And that's it. As far as I understand it turns segfault into deadlocks. That's exactly what I get. I prefer the segfaults this way I don't have to kill CTP2 myself. But whether deadlocks or segfault, either way CTP2 for Linux from master isn't in a good state.

74c4f07 was the latest commit when I last fetched https://github.com/civctp2/civctp2 at the beginning of the year.
It seems some of my branches/PRs got merged outside GH such that I got no message about the new status and therefore could not comment on new commits, nor did I get informed by all the great work that was done by @MartinGuehmann lately. Please merge on GH or others wont be able to follow.

I think, I just remove what should not go into master and agree on the rules for merging. And agree on the rules to adhere for the future.

No direct push to master.
The master branch is only updated via pull requests.
Pull requests can only be merged in if they pass Travis-CI.
Any head commit should build on Linux, ideally also on Windows but Travis-CI does not check this right now.
No, history flattening, unless you still request to do it for the current master. But that would be the only exception.

For the pull requests, I would like to merge them in as soon as possible, at max a week, unless there is more work to do on them. So what would be the minimum time to keep them open, I guess at least a day. Or should I merge them in as soon as they have passed the checks?

LynxAbraxas · 2019-01-23T20:11:02Z

I wouldn't squeeze anything, either. It doesn't feel like a good idea to tinker around with it. We should keep in mind that this project has a non-Git history. This means not every commit compiles, anyway. So, I would rather demand that only the head passes all tests.

My preferred way for my own projects is to tell the history of trail-and-error in side-branches (only squashing/reordering commits if I forgot something or made a mistake) allowing commits not passing tests, but being strict that any commit to master does, ideally master only to be made up of merge commits that used --no-ff. If you then look at the git history with e.g. gitk you can get the the squeezed history on the short branch side and the full story on the long. With a right click on another commit you can even get the diff over the range of commits.

LynxAbraxas · 2019-01-23T20:29:21Z

Regarding merging PRs: From othe projects I'm used to the procedure that a contributor is asked to base their changes on current master. Then when all test pass for that PR and the PR is approved (by maintainer or someone with similar status) it might be merged as is or rebased on the current master (if it has moved on) and then merged with --no-ff. One reason for the rebasing is to avoid branches embracing others, which is at an extreme for the current master and makes it very difficult to go through.

For example

PR #39 did not build, but I fixed it.

it would be great to have your commits fixing this directly listed in #39. For that you can add my fork https://github.com/LynxAbraxas/civctp2/ as a remote to your local clone fetch from that remote and then checkout https://github.com/LynxAbraxas/civctp2/tree/compile4debug and place your commit on top. Then either create a PR to my repos or ask me to cherry-pick your commits. Then when it finally compiles and test fine (with travis for linux and docker, in future ideally also for windows) and it is approved it will be merged into master.
If you look at the history with gitk this would mean that 3dbf32d would come before the merge cd089f8

LynxAbraxas · 2019-01-23T20:41:14Z

For commits like 19fc4ac it is best to use git cherry-pick. Even better would be to create a new topic-branch on master, cherry-pick to that, then create a PR.

LynxAbraxas · 2019-01-23T23:25:05Z

Well, if you wanna do this. ;) For removing what nobody has requested to include, I would just need to change ../ctp2_code/ui/aui_sdl/aui_sdlblitter.cpp.

Concerning resetting master: From looking at the current history, there is more in from me than there should, but I agree that reverting the changes from https://github.com/LynxAbraxas/civctp2/tree/debug-random-segfaults is probably the shorter option. I leave it to @MartinGuehmann as maintainer to decide.
Currently, I can't say if the order the branches were merge introduced some inconsistencies. However, since https://github.com/LynxAbraxas/civctp2/tree/master4ctp2DF (which is for https://github.com/LynxAbraxas/ctp2DF until PRs would be merged and not meant for merging directly (no PR)) is now in, most inconsistencies I faced when combining content of the PR branches are resolved there.

With #63 compilation with SDL_ffmpeg on GitLab works again but not all test of https://github.com/LynxAbraxas/ctp2DF succeed.

MartinGuehmann · 2019-01-24T18:38:56Z

Concerning resetting master: From looking at the current history, there is more in from me than there should, but I agree that reverting the changes from https://github.com/LynxAbraxas/civctp2/tree/debug-random-segfaults is probably the shorter option. I leave it to @MartinGuehmann as maintainer to decide.

Actually, since I did not understand the acronym WIP (I commented in #57 about it), I am more seriously thinking about. But still it is some work, it may be less work than I thought, initially. But it is still something I should not decide on so easily.

With #63 compilation with SDL_ffmpeg on GitLab works again but not all test of https://github.com/LynxAbraxas/ctp2DF succeed.

Well, that are regressions, which would have to be resolved anyway, whether I would rewrite history or not.

LynxAbraxas · 2019-01-24T22:19:01Z

Well, that are regressions, which would have to be resolved anyway, whether I would rewrite history or not.

With git, this can be avoided when carefully doing the merge commit. Git allows to have two or multiple way merges (with special logic) which combine changes from both branches that are combined, you can even add further changes (overlooked first) with git commit --amend to a merge commit. Git's automatic merging mostly decides well which changes have precedence over other others, but not in all cases. It would have been nice to have the "regression" of #63 be part of the merge commit that introduced it. With the order of merging the PRs into master as in https://github.com/LynxAbraxas/civctp2/commits/master4ctp2DF?after=20eb4b309803f2d06fa6c09b658ed0e554d53e33+69 git's precedence decision avoided such a regression.

DEVoytas · 2019-01-25T17:34:14Z

@LynxAbraxas @MartinGuehmann the discussion here seem to be no more related to finding project maintainer. I suggest closing the issue and continue it in dedicated issue I created here: #65
Please read it and see if you agree.
If there is any specific information it this issue that should not be lost/overlooked, it is very easy to link to specific comment or quote it.

MartinGuehmann · 2019-01-25T18:50:28Z

I haven't figured out how to link to a specific post, but let's do as @DEVoytas suggest not here.

…tes (sender civ == receiver civ)", causes "No such object": This reverts commit fe3b57c. Messagebox(GameObj.cpp Error): No such object Continue? (ctp2:385): Gtk-CRITICAL **: 18:45:55.493: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.495: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.496: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.498: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.506: IA__gtk_main_quit: assertion 'main_loops != NULL' failed Thread 1 "ctp2" received signal SIGSEGV, Segmentation fault. 0x00005555557902ba in UnitData::GetOwner() const () (gdb) bt #0 0x00005555557902ba in UnitData::GetOwner() const () civctp2#1 0x00005555560917d4 in Unit::GetOwner() const () civctp2#2 0x000055555608c8b0 in TradeRoute::RemoveAllReferences(CAUSE_KILL_TRADE_ROUTE) () civctp2#3 0x000055555608c658 in TradeRoute::KillRoute(CAUSE_KILL_TRADE_ROUTE) () civctp2#4 0x000055555578fe8e in TradeRoute::Kill(CAUSE_KILL_TRADE_ROUTE) () civctp2#5 0x000055555600931e in CityData::~CityData() () civctp2#6 0x0000555556009504 in CityData::~CityData() () civctp2#7 0x00005555560985b8 in UnitData::~UnitData() () civctp2#8 0x000055555609869c in UnitData::~UnitData() () civctp2#9 0x0000555556043a05 in GameObj_Delete(GameObj**, unsigned int) () civctp2#10 0x0000555556052891 in ObjPool::Del(ID const&) () civctp2#11 0x00005555560912ee in Unit::RemoveAllReferences(CAUSE_REMOVE_ARMY, int) () civctp2#12 0x0000555556090ae2 in Unit::KillUnit(CAUSE_REMOVE_ARMY, int) () civctp2#13 0x00005555558bcfd4 in Unit::Kill(CAUSE_REMOVE_ARMY, int) () civctp2#14 0x00005555560179fc in CityData::CityToPark(int) () civctp2#15 0x00005555560a0dd2 in UnitData::CityToPark(int) () civctp2#16 0x000055555609446f in Unit::CityToPark(int) () civctp2#17 0x00005555560c6171 in CreateParkEvent::GEVHookCallback(GAME_EVENT, GameEventArgList*) () civctp2#18 0x0000555555fc0500 in GameEventHook::Activate(GameEventArgList*, int, int&) const () civctp2#19 0x0000555555fbd9c4 in GameEventManager::ActivateHook(GAME_EVENT, GameEventArgList*, int, int&) () civctp2#20 0x0000555555fbff27 in GameEvent::Process() () civctp2#21 0x0000555555fbd6b4 in GameEventManager::ProcessHead() () civctp2#22 0x0000555555fbd5e3 in GameEventManager::Process() () civctp2#23 0x0000555555780f4b in CivApp::Process() () civctp2#24 0x0000555555776cec in CivMain(int, char**) () civctp2#25 0x0000555555776889 in main () # Please enter the commit message for your changes. Lines starting # with '#' will be kept; you may remove them yourself if you want to. # An empty message aborts the commit. # # Date: Sat Dec 28 19:51:22 2019 +0100 # # On branch master4testing # Your branch is ahead of 'origin/master4testing' by 1 commit. # (use "git push" to publish your local commits) # # Changes to be committed: # modified: ctp2_code/gs/gameobj/TradeRoute.cpp #

…tes (sender civ == receiver civ)", causes "No such object": This reverts commit fe3b57c. Messagebox(GameObj.cpp Error): No such object Continue? (ctp2:385): Gtk-CRITICAL **: 18:45:55.493: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.495: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.496: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.498: IA__gtk_main_quit: assertion 'main_loops != NULL' failed (ctp2:385): Gtk-CRITICAL **: 18:45:55.506: IA__gtk_main_quit: assertion 'main_loops != NULL' failed Thread 1 "ctp2" received signal SIGSEGV, Segmentation fault. 0x00005555557902ba in UnitData::GetOwner() const () (gdb) bt #0 0x00005555557902ba in UnitData::GetOwner() const () civctp2#1 0x00005555560917d4 in Unit::GetOwner() const () civctp2#2 0x000055555608c8b0 in TradeRoute::RemoveAllReferences(CAUSE_KILL_TRADE_ROUTE) () civctp2#3 0x000055555608c658 in TradeRoute::KillRoute(CAUSE_KILL_TRADE_ROUTE) () civctp2#4 0x000055555578fe8e in TradeRoute::Kill(CAUSE_KILL_TRADE_ROUTE) () civctp2#5 0x000055555600931e in CityData::~CityData() () civctp2#6 0x0000555556009504 in CityData::~CityData() () civctp2#7 0x00005555560985b8 in UnitData::~UnitData() () civctp2#8 0x000055555609869c in UnitData::~UnitData() () civctp2#9 0x0000555556043a05 in GameObj_Delete(GameObj**, unsigned int) () civctp2#10 0x0000555556052891 in ObjPool::Del(ID const&) () civctp2#11 0x00005555560912ee in Unit::RemoveAllReferences(CAUSE_REMOVE_ARMY, int) () civctp2#12 0x0000555556090ae2 in Unit::KillUnit(CAUSE_REMOVE_ARMY, int) () civctp2#13 0x00005555558bcfd4 in Unit::Kill(CAUSE_REMOVE_ARMY, int) () civctp2#14 0x00005555560179fc in CityData::CityToPark(int) () civctp2#15 0x00005555560a0dd2 in UnitData::CityToPark(int) () civctp2#16 0x000055555609446f in Unit::CityToPark(int) () civctp2#17 0x00005555560c6171 in CreateParkEvent::GEVHookCallback(GAME_EVENT, GameEventArgList*) () civctp2#18 0x0000555555fc0500 in GameEventHook::Activate(GameEventArgList*, int, int&) const () civctp2#19 0x0000555555fbd9c4 in GameEventManager::ActivateHook(GAME_EVENT, GameEventArgList*, int, int&) () civctp2#20 0x0000555555fbff27 in GameEvent::Process() () civctp2#21 0x0000555555fbd6b4 in GameEventManager::ProcessHead() () civctp2#22 0x0000555555fbd5e3 in GameEventManager::Process() () civctp2#23 0x0000555555780f4b in CivApp::Process() () civctp2#24 0x0000555555776cec in CivMain(int, char**) () civctp2#25 0x0000555555776889 in main ()

…ct": This reverts commit 81ae687. Thread 1 "ctp2" hit Catchpoint 1 (signal SIGSEGV), 0x0000555555790236 in UnitData::GetOwner (this=0x0) at ../ctp2_code/gs/gameobj/UnitData.h:296 296 PLAYER_INDEX GetOwner () const { return m_owner; }; #0 0x0000555555790236 in UnitData::GetOwner (this=0x0) at ../ctp2_code/gs/gameobj/UnitData.h:296 civctp2#1 0x000055555609176a in Unit::GetOwner (this=0x7fffffffe254) at Unit.cpp:397 civctp2#2 0x000055555608c847 in TradeRoute::RemoveAllReferences (this=0x7fffffffe2b4, cause=CAUSE_KILL_TRADE_ROUTE_CITY_DIED) at TradeRoute.cpp:70 civctp2#3 0x000055555608c5c8 in TradeRoute::KillRoute (this=0x55555f993d5c, cause=CAUSE_KILL_TRADE_ROUTE_CITY_DIED) at TradeRoute.cpp:35 civctp2#4 0x000055555578fe0a in TradeRoute::Kill (this=0x55555f993d5c, cause=CAUSE_KILL_TRADE_ROUTE_CITY_DIED) at ../ctp2_code/gs/gameobj/TradeRoute.h:51 civctp2#5 0x000055555600928e in CityData::~CityData (this=0x55555f9935ac, __in_chrg=<optimized out>) at CityData.cpp:582 civctp2#6 0x0000555556009474 in CityData::~CityData (this=0x55555f9935ac, __in_chrg=<optimized out>) at CityData.cpp:605 civctp2#7 0x000055555609854e in UnitData::~UnitData (this=0x55555f99327c, __in_chrg=<optimized out>) at UnitData.cpp:384 civctp2#8 0x0000555556098632 in UnitData::~UnitData (this=0x55555f99327c, __in_chrg=<optimized out>) at UnitData.cpp:389 civctp2#9 0x0000555556043975 in GameObj_Delete (p=0x55555f74099c, id=268435500) at GameObj.cpp:190 civctp2#10 0x0000555556052801 in ObjPool::Del (this=0x55555f74082c, id=...) at ObjPool.cpp:106 civctp2#11 0x0000555556091284 in Unit::RemoveAllReferences (this=0x7fffffffe474, cause=CAUSE_REMOVE_ARMY_PARKRANGER, killedBy=1) at Unit.cpp:293 civctp2#12 0x0000555556090a78 in Unit::KillUnit (this=0x55555f9936b0, cause=CAUSE_REMOVE_ARMY_PARKRANGER, killedBy=1) at Unit.cpp:126 civctp2#13 0x00005555558bcf50 in Unit::Kill (this=0x55555f9936b0, cause=CAUSE_REMOVE_ARMY_PARKRANGER, killedBy=1) at ../ctp2_code/gs/gameobj/Unit.h:108 civctp2#14 0x000055555601796c in CityData::CityToPark (this=0x55555f9935ac, agressor=1) at CityData.cpp:5981 civctp2#15 0x00005555560a0d68 in UnitData::CityToPark (this=0x55555f99327c, agressor=1) at UnitData.cpp:3191 civctp2#16 0x0000555556094405 in Unit::CityToPark (this=0x7fffffffe5a0, agressor=1) at Unit.cpp:1522 civctp2#17 0x00005555560c6107 in CreateParkEvent::GEVHookCallback (this=0x55555674d980 <s_CreateParkEvent>, gameEventType=GEV_CreatePark, args=0x555582a0543c) at CityEvent.cpp:662 civctp2#18 0x0000555555fc0470 in GameEventHook::Activate (this=0x55555ef016cc, args=0x555582a0543c, startIndex=0, resumeIndex=@0x555582a062bc: 1) at GameEventHook.cpp:254 civctp2#19 0x0000555555fbd934 in GameEventManager::ActivateHook (this=0x55555eee50ac, type=GEV_CreatePark, args=0x555582a0543c, startIndex=-1, resumeIndex=@0x555582a062bc: 1) at GameEventManager.cpp:511 civctp2#20 0x0000555555fbfe97 in GameEvent::Process (this=0x555582a062ac) at GameEvent.cpp:112 civctp2#21 0x0000555555fbd624 in GameEventManager::ProcessHead (this=0x55555eee50ac) at GameEventManager.cpp:440 civctp2#22 0x0000555555fbd553 in GameEventManager::Process (this=0x55555eee50ac) at GameEventManager.cpp:415 civctp2#23 0x0000555555780ec7 in CivApp::Process (this=0x555556caec9c) at ctp/civapp.cpp:2563 civctp2#24 0x0000555555776c68 in CivMain (iCmdShow=2, pSzCmdLine=0x7fffffffecb8) at ctp/civ3_main.cpp:1694 civctp2#25 0x0000555555776805 in main (argc=2, argv=0x7fffffffecb8) at ctp/civ3_main.cpp:1440

Thread 1 "ctp2" hit Catchpoint 2 (signal SIGINT), 0x00007ffff724a727 in raise () from /lib/x86_64-linux-gnu/libpthread.so.0 Assertion (start < m_code + m_codeSize) Failed in File:SlicSegment.cpp, Line:649 Thread 1 "ctp2" hit Catchpoint 2 (signal SIGINT), 0x00007ffff724a727 in raise () from /lib/x86_64-linux-gnu/libpthread.so.0 Thread 1 "ctp2" hit Catchpoint 3 (signal SIGILL), SlicModFunc::SlicModFunc (this=0x5555605ae6fc, name=0x555556864ee1 "mod_CanPlayerHaveAdvance") at SlicModFunction.cpp:21 21 tag = va_arg(vl, SLIC_TAG); #0 SlicModFunc::SlicModFunc (this=0x5555605ae6fc, name=0x555556864ee1 "mod_CanPlayerHaveAdvance") at SlicModFunction.cpp:21 civctp2#1 0x00005555566159c9 in SlicEngine::AddModFuncs (this=0x55555e2e4fac) at SlicEngine.cpp:3069 civctp2#2 0x0000555556604f29 in SlicEngine::PostSerialize (this=0x55555e2e4fac) at SlicEngine.cpp:457 civctp2#3 0x00005555566a4f0d in gameinit_Initialize (mWidth=-1, mHeight=-1, archive=...) at gameinit.cpp:1683 civctp2#4 0x0000555555b82af5 in CivApp::InitializeGame (this=0x5555570b4bdc, archive=...) at ctp/civapp.cpp:1717 civctp2#5 0x00005555563cd57c in GameFile::Restore (this=0x7fffffffe9e7, filepath=0x55555e916bf4 "/opt/ctp2/Scenarios/AlexanderTheGreat/scen0000/savegame.c2g") at GameFile.cpp:848 civctp2#6 0x00005555563cb4ca in GameFile::RestoreGame (name=0x55555e916bf4 "/opt/ctp2/Scenarios/AlexanderTheGreat/scen0000/savegame.c2g") at GameFile.cpp:186 civctp2#7 0x0000555555b856b6 in CivApp::LoadSavedGame (this=0x5555570b4bdc, name=0x55555e916bf4 "/opt/ctp2/Scenarios/AlexanderTheGreat/scen0000/savegame.c2g") at ctp/civapp.cpp:2651 civctp2#8 0x0000555555b862de in LoadSaveGameAction::Execute (this=0x55555e916bec, control=0x0, action=0, data=0) at ctp/civapp.cpp:2931 civctp2#9 0x0000555555c4b232 in aui_UI::HandleActions (this=0x555557c790dc) at ui/aui_common/aui_ui.cpp:1862 civctp2#10 0x0000555555c4b188 in aui_UI::Process (this=0x555557c790dc) at ui/aui_common/aui_ui.cpp:1847 civctp2#11 0x0000555555b84e02 in CivApp::ProcessUI (this=0x5555570b4bdc, target_milliseconds=30, used_milliseconds=@0x7fffffffeb1c: 0) at ctp/civapp.cpp:2395 civctp2#12 0x0000555555b852a8 in CivApp::Process (this=0x5555570b4bdc) at ctp/civapp.cpp:2548 civctp2#13 0x0000555555b7b2cc in CivMain (iCmdShow=2, pSzCmdLine=0x7fffffffecc8) at ctp/civ3_main.cpp:1694 civctp2#14 0x0000555555b7ae69 in main (argc=2, argv=0x7fffffffecc8) at ctp/civ3_main.cpp:1440 # Please enter the commit message for your changes. Lines starting # with '#' will be kept; you may remove them yourself if you want to. # An empty message aborts the commit. # # Date: Wed May 20 11:54:39 2020 +0200 # # interactive rebase in progress; onto 88ba108 # Last command done (1 command done): # edit 4548e58f gdb caught SIGILL (illegal instruction): # Next command to do (1 remaining command): # pick b6491837 no backtrace on SIGINT (too many) # You are currently splitting a commit while rebasing branch 'dbgAlex' on '88ba1081'. # # Changes to be committed: # modified: debug/gdb.cmds #

Thread 1 "ctp2" hit Catchpoint 2 (signal SIGINT), 0x00007ffff724a727 in raise () from /lib/x86_64-linux-gnu/libpthread.so.0 Assertion (start < m_code + m_codeSize) Failed in File:SlicSegment.cpp, Line:649 Thread 1 "ctp2" hit Catchpoint 2 (signal SIGINT), 0x00007ffff724a727 in raise () from /lib/x86_64-linux-gnu/libpthread.so.0 Thread 1 "ctp2" hit Catchpoint 3 (signal SIGILL), SlicModFunc::SlicModFunc (this=0x5555605ae6fc, name=0x555556864ee1 "mod_CanPlayerHaveAdvance") at SlicModFunction.cpp:21 21 tag = va_arg(vl, SLIC_TAG); #0 SlicModFunc::SlicModFunc (this=0x5555605ae6fc, name=0x555556864ee1 "mod_CanPlayerHaveAdvance") at SlicModFunction.cpp:21 civctp2#1 0x00005555566159c9 in SlicEngine::AddModFuncs (this=0x55555e2e4fac) at SlicEngine.cpp:3069 civctp2#2 0x0000555556604f29 in SlicEngine::PostSerialize (this=0x55555e2e4fac) at SlicEngine.cpp:457 civctp2#3 0x00005555566a4f0d in gameinit_Initialize (mWidth=-1, mHeight=-1, archive=...) at gameinit.cpp:1683 civctp2#4 0x0000555555b82af5 in CivApp::InitializeGame (this=0x5555570b4bdc, archive=...) at ctp/civapp.cpp:1717 civctp2#5 0x00005555563cd57c in GameFile::Restore (this=0x7fffffffe9e7, filepath=0x55555e916bf4 "/opt/ctp2/Scenarios/AlexanderTheGreat/scen0000/savegame.c2g") at GameFile.cpp:848 civctp2#6 0x00005555563cb4ca in GameFile::RestoreGame (name=0x55555e916bf4 "/opt/ctp2/Scenarios/AlexanderTheGreat/scen0000/savegame.c2g") at GameFile.cpp:186 civctp2#7 0x0000555555b856b6 in CivApp::LoadSavedGame (this=0x5555570b4bdc, name=0x55555e916bf4 "/opt/ctp2/Scenarios/AlexanderTheGreat/scen0000/savegame.c2g") at ctp/civapp.cpp:2651 civctp2#8 0x0000555555b862de in LoadSaveGameAction::Execute (this=0x55555e916bec, control=0x0, action=0, data=0) at ctp/civapp.cpp:2931 civctp2#9 0x0000555555c4b232 in aui_UI::HandleActions (this=0x555557c790dc) at ui/aui_common/aui_ui.cpp:1862 civctp2#10 0x0000555555c4b188 in aui_UI::Process (this=0x555557c790dc) at ui/aui_common/aui_ui.cpp:1847 civctp2#11 0x0000555555b84e02 in CivApp::ProcessUI (this=0x5555570b4bdc, target_milliseconds=30, used_milliseconds=@0x7fffffffeb1c: 0) at ctp/civapp.cpp:2395 civctp2#12 0x0000555555b852a8 in CivApp::Process (this=0x5555570b4bdc) at ctp/civapp.cpp:2548 civctp2#13 0x0000555555b7b2cc in CivMain (iCmdShow=2, pSzCmdLine=0x7fffffffecc8) at ctp/civ3_main.cpp:1694 civctp2#14 0x0000555555b7ae69 in main (argc=2, argv=0x7fffffffecc8) at ctp/civ3_main.cpp:1440

DEVoytas mentioned this issue Oct 12, 2018

How was the (git) history flattened? ptitSeb/ctp2#3

Closed

civctp2 added the help wanted Extra attention is needed label Oct 12, 2018

DEVoytas mentioned this issue Oct 13, 2018

Merge @ptitSeb's and @RolandTaverner's branches to master #4

Closed

This was referenced Nov 29, 2018

Dockerfile and .travis.yml to test docker build #23

Merged

Debug diplomacy #35

Merged

MartinGuehmann closed this as completed Jan 18, 2019

MartinGuehmann reopened this Jan 22, 2019

MartinGuehmann removed the help wanted Extra attention is needed label Jan 22, 2019

LynxAbraxas mentioned this issue Jan 23, 2019

Merging linux branch into master #1

Closed

DEVoytas mentioned this issue Jan 25, 2019

Workflow and other general project discussions #65

Open

MartinGuehmann closed this as completed Jan 25, 2019

MartinGuehmann mentioned this issue Feb 1, 2019

irregular segfaults during game play #30

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Project maintainer wanted #2

Project maintainer wanted #2

civctp2 commented Oct 12, 2018

DEVoytas commented Nov 29, 2018

ptitSeb commented Nov 29, 2018

LynxAbraxas commented Nov 30, 2018

LynxAbraxas commented Dec 21, 2018

MartinGuehmann commented Dec 22, 2018

DEVoytas commented Dec 26, 2018

MartinGuehmann commented Dec 26, 2018

LynxAbraxas commented Dec 26, 2018

MartinGuehmann commented Dec 26, 2018

MartinGuehmann commented Jan 9, 2019

DEVoytas commented Jan 16, 2019

civctp2 commented Jan 16, 2019

DEVoytas commented Jan 17, 2019

MartinGuehmann commented Jan 18, 2019

LynxAbraxas commented Jan 22, 2019 •

edited

Loading

LynxAbraxas commented Jan 22, 2019

LynxAbraxas commented Jan 22, 2019 •

edited

Loading

LynxAbraxas commented Jan 22, 2019 •

edited

Loading

DEVoytas commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

LynxAbraxas commented Jan 23, 2019

LynxAbraxas commented Jan 23, 2019

LynxAbraxas commented Jan 23, 2019

LynxAbraxas commented Jan 23, 2019

MartinGuehmann commented Jan 24, 2019

LynxAbraxas commented Jan 24, 2019

DEVoytas commented Jan 25, 2019

MartinGuehmann commented Jan 25, 2019

Project maintainer wanted #2

Project maintainer wanted #2

Comments

civctp2 commented Oct 12, 2018

DEVoytas commented Nov 29, 2018

ptitSeb commented Nov 29, 2018

LynxAbraxas commented Nov 30, 2018

LynxAbraxas commented Dec 21, 2018

MartinGuehmann commented Dec 22, 2018

DEVoytas commented Dec 26, 2018

MartinGuehmann commented Dec 26, 2018

LynxAbraxas commented Dec 26, 2018

MartinGuehmann commented Dec 26, 2018

MartinGuehmann commented Jan 9, 2019

DEVoytas commented Jan 16, 2019

civctp2 commented Jan 16, 2019

DEVoytas commented Jan 17, 2019

MartinGuehmann commented Jan 18, 2019

LynxAbraxas commented Jan 22, 2019 • edited Loading

LynxAbraxas commented Jan 22, 2019

LynxAbraxas commented Jan 22, 2019 • edited Loading

LynxAbraxas commented Jan 22, 2019 • edited Loading

DEVoytas commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

MartinGuehmann commented Jan 22, 2019

LynxAbraxas commented Jan 23, 2019

LynxAbraxas commented Jan 23, 2019

LynxAbraxas commented Jan 23, 2019

LynxAbraxas commented Jan 23, 2019

MartinGuehmann commented Jan 24, 2019

LynxAbraxas commented Jan 24, 2019

DEVoytas commented Jan 25, 2019

MartinGuehmann commented Jan 25, 2019

LynxAbraxas commented Jan 22, 2019 •

edited

Loading

LynxAbraxas commented Jan 22, 2019 •

edited

Loading

LynxAbraxas commented Jan 22, 2019 •

edited

Loading