glob2-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [glob2-devel] Optimization suggestions


From: Bradley Arsenault
Subject: Re: [glob2-devel] Optimization suggestions
Date: Wed, 5 Sep 2007 16:51:12 -0400

On 9/5/07, Kai Antweiler <address@hidden> wrote:
> > For some reason I thought it might be a server-mode thing...
>
> ...The reason being that "glob2 --help" just says:
> -nox <game file name>    runs the game without using the X server

Yes, I asked Nuage last year to include more options.  Until then I used
a save game where the human player was about to loose and called
"glob2 --nox" from inside a bash for loop.


> Anyway, there seems to be a problem with that option - it triggers a
> segfault right away, here:
> 0x080f0ce5 in GameGUI (this=0xbf7f3a54) at src/GameGUI.cpp:132
> 132       minimap(globalContainer->gfx->getW()-128, 0, 128, 14,
> Minimap::HideFOW)
> - because globalContainer->gfx == NULL.

Not again.  Options that are not used become bugs really fast in glob2.
(This will happen to Stephs critters too.)

OK, I can't fix this because my valgrind isn't working.
You can change line 132 to:

  minimap(globalContainer->runNoX ? 0 :
globalContainer->gfx->getW()-128, 0, 128, 14, Minimap::HideFOW)

I don't know whether 0 is a good value.
Someone else has to fix this, or we have to wait until gentoo uses a
newer valgrind.
--
Kai Antweiler

This is because the GlobalContainer expects gfx to exist at initialization. In this case, your fix works, just because no mini map code is executed after initialization. A potential fix is to change the minimap in GameGUI from an object to a pointer, and only create it in the first place if gfx exists.


--
Extra cheese comes at a cost. Bradley Arsenault.
reply via email to

[Prev in Thread] Current Thread [Next in Thread]