#uwadv@irc.freenode.net logs for 8 Mar 2004 (GMT)

Archive Today Yesterday Tomorrow
Underworld Adventures homepage

[16:35:03] --> Dragsty has joined #uwadv
[16:36:50] <-- Dragsty has left IRC (Client Quit)
[22:06:56] --> vividos has joined #uwadv
[22:06:56] --- ChanServ gives channel operator status to vividos
[22:06:59] <vividos> hi
[22:07:29] <wjp> hi
[22:09:37] <vividos> seems to be a bug fix day today :)
[22:10:39] <wjp> speaking of which, I'm getting another compiler error :-)
[22:10:56] <wjp> did you commit all files?
[22:11:15] <vividos> not yet
[22:16:33] <vividos> I hope I fixed all valgrind warnings that occured
[22:16:59] <wjp> once it compiles again I'll run valgrind again
[22:17:53] <wjp> ok, that last commit made it pass the error I was getting
[22:20:53] <vividos> which one? :-)
[22:21:11] <wjp> critter.something was still a smartptr, but it was being assigned to a vector
[22:21:17] <vividos> ah right
[22:21:23] <vividos> ok should be finished committing all the stuff
[22:21:38] <wjp> loading all uw1 critter animations ... done, needed 53884 ms <-- with valgrind
[22:21:54] <wjp> loading all uw1 critter animations ... done, needed 811 ms <-- without :-)
[22:22:06] <vividos> nice :)
[22:22:48] <wjp> couple of warnings in SDL_mixer
[22:23:48] <vividos> do you have 1.2.5 yet?
[22:24:01] <wjp> hm, I just did something that triggered a warning
[22:24:07] <wjp> ==5227== Use of uninitialised value of size 8
[22:24:12] <wjp> ==5227== at 0x8096CF3: ua_window::calc_mousepos(SDL_Event&, unsigned&, unsigned&) (window.cpp:97)
[22:24:16] <wjp> ==5227== by 0x8087DF8: ua_ingame_move_arrows::process_event(SDL_Event&) (ingame_ctrls.cpp:728)
[22:24:50] <wjp> (and then the same error at line window.cpp:98 too
[22:25:09] <wjp> I think I pressed q or something
[22:26:21] <wjp> yeah, I have sdl-mixer 1.2.5
[22:26:38] <vividos> do you think it means the two references passed? I'll set them to 0 for each call to calc_mousepos()
[22:27:34] <wjp> hm
[22:27:58] <wjp> weird
[22:28:06] <vividos> or the SDL_Event struct isn't filled out completely
[22:28:49] <vividos> now I know
[22:29:19] <vividos> calc_mousepos() should only called for mouse events; key events do have unset xpos fields
[22:29:25] <vividos> fixing that ...
[22:32:04] <vividos> ok committed
[22:32:36] <vividos> I maybe should also commit that change in ua_smart_ptr to test with valgrind
[22:32:54] <wjp> which change is that?
[22:34:47] <vividos> I moved the members px and pn into a "struct data" member that is shared between all smart ptrs that point to the same object
[22:35:04] <vividos> effectively cutting down size of ua_smart_ptr in half
[22:36:40] <wjp> '--01.0'? :-)
[22:38:15] <vividos> typo; should be fixed in CVS
[22:38:32] <vividos> ua_smart_ptr change is also in CVS
[22:38:34] <wjp> yeah, I saw it in the cvs mail; interesting typo :-)
[22:39:05] <vividos> result of drinking coffee when you normally don't
[22:39:27] <wjp> hm, looks like this is causing quite a few files to be recompiled
[22:39:50] <vividos> yes, unfortunately
[22:44:13] <wjp> hm, not sure if I missed these the last time or if they're new:
[22:44:25] <wjp> ==7266== Conditional jump or move depends on uninitialised value(s)
[22:44:27] <wjp> [...]
[22:44:31] <wjp> ==7266== by 0x8059D80: ua_trace_printf(char const*, ...) (utils.cpp:127)
[22:44:36] <wjp> ==7266== by 0x80A4F89: ua_uwadv_game::init_sdl() (uwadv.cpp:506)
[22:46:27] <wjp> anyway, anything specific you want me to do to test the changes?
[22:46:47] <vividos> what does the output say for vendor: "%s\n renderer: %s\n version: %s" on the console?
[22:47:28] <wjp> looks ok
[22:47:29] <vividos> just run the program for testing; I mainly wanted to remove those warnings on mixing delete / delete []
[22:47:35] <wjp> probably confused by the nvidia drivers
[22:47:47] <vividos> maybe
[22:47:55] <wjp> (it might be getting those strings from a 'weird' place)
[22:48:17] <wjp> is falling broken?
[22:48:55] <vividos> yes
[22:48:58] <wjp> no valgrind warnings
[22:49:06] <vividos> I'm trying to reach Telemachos about that
[22:49:15] <vividos> ok good
[22:52:33] <vividos> should go to bed with that comfortable feeling that no errors are in uwadv ... :)
[22:52:54] <wjp> did I mention that that debugger string buffer thingie is still in there? ;-)
[22:53:45] <vividos> darn.. :)
[22:54:16] <wjp> of course since I hacked around that locally you could go to bed with the comfortable feeling that no errors are in my local copy of uwadv? ;-)
[22:54:52] <vividos> that's ok for me :-) g'night!
[22:54:57] <wjp> night :-)
[22:55:25] <-- vividos has left IRC ("Monday's bug fix day :-)")
[22:56:20] <wjp> (other than that problem with falling *cough* *cough* ;-P )