@krzys_h : Sure. Well, that's probably because I didn't ever did a bigger version of this render: https://github.com/colobot/colobot-data/...intro1.png

because I'm still using this one and it's boring http://i.imgur.com/rqDrMrl.png

@RaptorParkowsky Later make a version branded with full Colobot title at the top for my new wallpaper pleeeeaseee

Sparkles or Firework. What was first? Discover it by yourself: http://i.imgur.com/6UgxIsB.png - Now only makeshift textures and materials for Firework model left to do and I can render it in 2K.

@tomangelo I don't get it

@kevinvr that is a known problem with 0.1.8, use the latest dev version or wait for next release

will reload and try again without any old saves.

don't know why but now missions crash as soon as I start shooting ants , wasps and stuff.

@tomangelo put it on github, I'll forget otherwise

And did make clean before compiling

And repo and MSYS are both up-to-date

Don't want to bring bad news, but still got SIGSEGV in CBOT code when loading levels http://pastebin.com/LcDcxN7L

https://github.com/colobot/colobot/blob/...#L365-L370 no, it's NOT okay like that

It will be a lot work though.

I think we should find some automated way to place such safeguards in CBot code or at least detect where these unsafe calls are used and correct them manually.

But it is still undefined behavior. Code should check if pointer is null and only call the method if it's not null.

This happens when you call a non-static, non-virtual method on null pointer. Since they are usually implemented as normal functions with hidden pointer, nullptr is simply passed to it and called as usual.

Comparing this to nullptr means there's a severe problem in your code because something like this should never happen in the first place. It's a play on undefined behavior.

Quite interesting, comparing to nullptr is very common in code, but it was never optimized out from it. Or at least it didn't make such impossible to overlook results.