Menu
▾
▴
[ bzflag-Patches-408433 ] Radar changes
|
From: <no...@so...> - 2001-10-23 08:39:13
|
Patches item #408433, was opened at 2001-03-13 21:33 You can respond by visiting: http://sourceforge.net/tracker/?func=detail&atid=303248&aid=408433&group_id=3248 Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Jeremiah (cobraa1) Assigned to: Jeremiah (cobraa1) Summary: Radar changes Initial Comment: Does two things: 1. Makes a hollow diamond indicate height instead of a filled rectangle. That way, high players don't block large parts of the radar. 2. Made it much easier to navigate multi-level arenas, and to distinguish objects at different heights. Basically, here's what I did: I changed the radar so that objects close to your z-coordinate appear solid, and objects that are not close to your z-coordinate fade into the background. Have to have smoothing turned on in order to see this the best (smoothing turns on transparency on the radar). When smoothing is turned off, only the topmost object is seen, but it still changes color to indicate how close you are to it vertically. If anybody has any suggestions, adjustments, or ideas feel free to tell me. The bounding walls, players, and bullets are not affected by the changes listed in #2. You'll want to keep track of them at any height ;). -Jeremiah (CobraA1) ---------------------------------------------------------------------- >Comment By: Tim Riker (timriker) Date: 2001-10-23 01:39 Message: Logged In: YES user_id=8134 The segfault at startup is gone. changing settings in the Options menu still segfaults. Entering the help menu also segfaults. The radar look great, when it works! ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-10-22 10:29 Message: Logged In: YES user_id=25311 Excellent! I've also moved the initializing of usingCustom into the constructor of the class, so it's more OO friendly. Now all we need is word from TimRiker to see if the problem has gone away. ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-10-22 10:28 Message: Logged In: YES user_id=25311 Excellent! I've also moved the initializing of usingCustom into the constructor of the class, so it's more OO friendly. Now all we need is word from TimRiker to see if the problem has gone away. ---------------------------------------------------------------------- Comment By: David Trowbridge (davidtrowbridge) Date: 2001-10-22 09:09 Message: Logged In: YES user_id=66583 Small change to integrate with custom CTF/elevated bases (renders base after the rest of the radar, so even if a box is drawn above it the base position is still visible). Also allows radar to render rotated bases correctly. ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-10-21 21:34 Message: Logged In: YES user_id=25311 I think I've found it! Apparently it's not the braces - I watched the additions I made to the SceneRenderer class, and apparently usingCustom starts undefined (and stays that way)! I don't see why VC++ didn't catch that, but that seems to be the problem. I've updated my code for CVS, and included the fix. I'm also closing the bug report - somebody explained to me why the braces exist. ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-10-21 20:32 Message: Logged In: YES user_id=25311 There seems to be an anomaly in playing.cxx - there's a set of curly braces that don't seem to have a purpose. I've posted it as a bug. It is bug #473549. ---------------------------------------------------------------------- Comment By: David Trowbridge (davidtrowbridge) Date: 2001-10-21 17:25 Message: Logged In: YES user_id=66583 I just rebuilt CVS, so here's an updated patch (current with CVS as of 10/21/01 18:25 MDT ---------------------------------------------------------------------- Comment By: Tim Riker (timriker) Date: 2001-10-20 01:24 Message: Logged In: YES user_id=8134 nor current stack trace: (gdb) backtrace #0 AList::getLength (this=0xc483c2a1) at ../../include/AList.h:232 #1 0x080e33fa in BzfWindowCBAList::append (this=0xc483c2a1, item=@0xbfffebb4) at ../../include/BzfWindow.h:30 #2 0x080b1c46 in BzfWindow::addResizeCallback (this=0xc483c289, _cb=0x804f0f0 <ControlPanel::resizeCallback(void *)>, data=0xbffff594) at BzfWindow.cxx:72 #3 0x0804d7f6 in ControlPanel::ControlPanel (this=0xbffff594, _mainWindow=@0x402cc18a, renderer=@0xbffff76c) at ControlPanel.cxx:65 #4 0x0807d025 in startPlaying (_display=0x81016d8, renderer=@0xbffff76c, _resources=@0x80fbec0, _info=0x80fbd60) at playing.cxx:4320 #5 0x08092c59 in main (argc=1, argv=0xbffffcb4) at bzflag.cxx:1178 #6 0x402846ef in __libc_start_main () from /lib/libc.so.6 ---------------------------------------------------------------------- Comment By: Ken Franks (turnkey) Date: 2001-10-16 23:00 Message: Logged In: YES user_id=351537 I need instruction on how to install this patch after I get it. Thanks Ken ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-09-25 15:40 Message: Logged In: YES user_id=25311 CVS changed again; here's the updated version of the patch. TimRiker, could you send me an updated version of the debug dump? The line numbers on the old dump don't match. Could you also add "Updated for current CVS" to canned responses? ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-08-28 19:51 Message: Logged In: YES user_id=25311 I've re-updated the file for CVS, including timriker's modification. I'll try to find where the segfault occurs based on timriker's info, but I don't have Linux so I can't guarantee I'll be able to find it and fix it. ---------------------------------------------------------------------- Comment By: Tim Riker (timriker) Date: 2001-07-11 17:55 Message: Logged In: YES user_id=8134 ok, a windows thing. It's "Obstacle.h" not "obstacle.h" I fixed the patch and reattached it. When I apply if on Linux it segfaults. So I have not applied it to CVS. timr@localhost:~/work/bzflag$ gdb bin/bzflag GNU gdb 5.0 Copyright 2000 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i686-pc-linux-gnu"... (gdb) run Starting program: /home/timr/work/bzflag/bin/bzflag [New Thread 1024 (LWP 1546)] joystick 'joystick' not supported... Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 1024 (LWP 1546)] AList::getLength (this=0xffc38173) at ../../include/AList.h:232 232 return length; (gdb) backtrace #0 AList::getLength (this=0xffc38173) at ../../include/AList.h:232 #1 0x80e27aa in BzfWindowCBAList::append (this=0xffc38173, item=@0xbfffe7f4) at ../../include/BzfWindow.h:30 #2 0x80b14f6 in BzfWindow::addResizeCallback (this=0xffc3815b, _cb=0x804edb0 <ControlPanel::resizeCallback(void *)>, data=0xbffff1d4) at BzfWindow.cxx:72 #3 0x804d4b6 in ControlPanel::ControlPanel (this=0xbffff1d4, _mainWindow=@0x402bbe4d, renderer=@0xbffff56c) at ControlPanel.cxx:65 #4 0x807c575 in startPlaying (_display=0x8100758, renderer=@0xbffff56c, _resources=@0x80fb120, _info=0x80fafc0) at playing.cxx:4218 #5 0x80924e2 in main (argc=1, argv=0xbffffabc) at bzflag.cxx:1172 #6 0x4026b2eb in __libc_start_main () from /lib/libc.so.6 ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-07-11 16:15 Message: Logged In: YES user_id=25311 Unfortunately, I don't have access to the CVS tree right now, and I haven't gotten X-windows to work on my computer yet. I may not be able to do anything until I get back to college, where I have sufficient resources to start working on BZFlag again. Are you sure obstacle.h was changed? Try using the one on the CVS tree. ---------------------------------------------------------------------- Comment By: Tim Riker (timriker) Date: 2001-07-11 03:22 Message: Logged In: YES user_id=8134 Patch is missing obstacle.h so it does not compile. Please attach that file here as well. Tab formatting is all messed up. I attach a correctly formatted diff here. ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-04-22 11:34 Message: Logged In: YES user_id=25311 Updated for current version of CVS again. ---------------------------------------------------------------------- Comment By: Gert Brinkmann (gbrinkmann) Date: 2001-04-09 05:20 Message: Logged In: YES user_id=48712 Hi Jeremiah, when driving through the world, especially when turning, the buildings on the radar are flickering. Also when dropping or collecting a flag they flash dark. (That has been in all patch-versions already.) The last thing is: Being much higher than a building results in the building to disapear from the radar. I thought that this was fixed so that they have a minimum brightness all the time? (i am using low details; Linux Suse6.3, Xfree3.3.6) Ciao, Gert ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-04-07 12:39 Message: Logged In: YES user_id=25311 Updated for CVS again. The diamons is drawn all of the time; it's just that it dissapears because it gets smaller than the rectangle when at ground level. I've set VC++ for tab=8 and indent=2, so there's less problems when I change the code in the future. I still have to look at the formatting of the existing code sometime. Did a cvs update -C bzflag.dsp on my computer. Deleted CR at end of flag.cxx, so diff'ing doesnt'add the extra section. ---------------------------------------------------------------------- Comment By: Gert Brinkmann (gbrinkmann) Date: 2001-03-29 23:17 Message: Logged In: YES user_id=48712 A suggestion to the "hollow diamond indicate height instead of a filled rectangle" topic: Could you draw this diamond all the time even if the tanks are on the ground? (I think, at the moment it is only drawn when the tank is higher than the ground, isn't it?). The reason for this is that the blue tank very often is difficult to see on the radar. The white diamond around it would make it more visible. Gert ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-03-29 18:00 Message: Logged In: YES user_id=25311 Made Chris's changes: ---------------------- jeremiah, i've included a patch to your patch. it has the following changes: * uses the existing formatting style * exploits polymorphism (just one colorScale() and transScale() each) * makes new functions into methods on RadarRenderer * sets a minimum transparency so buildings never disappear ------------------- Also corrected compiler warnings and errors, and updated for CVS. ---------------------------------------------------------------------- Comment By: Gert Brinkmann (gbrinkmann) Date: 2001-03-29 02:50 Message: Logged In: YES user_id=48712 I retried the patch. Now i did a "make clean" before making it and it works. Thanks. The effect of shading the buildings will be good, i think, when it will be finnished. At the moment only buildings and teleporters at the same or higher level as the tanks level can be seen. This is irritating, the shading of the buldings should be computed relatively to the ground's height and not the tank's height, i think. Buildings in the same height as the tank is at are flickering when moving the tank. I guess that this will be gone, too, when the groundlevel-distance is used for computing the shading-intensity. Ciao, Gert ---------------------------------------------------------------------- Comment By: Tim Riker (timriker) Date: 2001-03-28 23:40 Message: Logged In: YES user_id=8134 indenting does not follow the tab=8 indent=2 SceneRenderer.cxx: In method `SceneRenderer::SceneRenderer(MainWindow &)': ../../include/SceneRenderer.h:197: warning: member initializers for `int SceneRenderer::useEnhancedRadarOn' ../../include/SceneRenderer.h:198: warning: and `boolean SceneRenderer::useFogHack' SceneRenderer.cxx:106: warning: will be re-ordered to match declaration order src/bzflag/menus.cxx - standard options should preceed debug options. I guess I will move the debug block at some point, so don't worry about this one. please avoid extraneous mods like the extra cr at end of src/common/Flag.cxx win32/bzflag.dsp incorrectly reverts VERSION= note: I did get the reported segfault when compiling under Linux. People should keep in mind that header dependancies are NOT calculated by the makefiles. a "make clean" and then a "make DEBUG=1" seems to fix most of the issue. I'm still getting seg faults when entering the options menu but it's inconsistant. I'll do more testing/debugging later. ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2001-03-28 07:37 Message: Logged In: NO I got the same problem, as gbrinkmann on my box (RedHat7, XFree4.0.2, Kernel 2.4.2, Geforce GTS with newest nvidia driver ) now going downgrading ...;-) bye ---------------------------------------------------------------------- Comment By: Gert Brinkmann (gbrinkmann) Date: 2001-03-26 23:51 Message: Logged In: YES user_id=48712 Oh sorry, "then it freezes" is missing in my reply. I can switch to another console to (CTRL-ALT-Fn) and kill bzflag. Ciao, Gert ---------------------------------------------------------------------- Comment By: Gert Brinkmann (gbrinkmann) Date: 2001-03-26 23:48 Message: Logged In: YES user_id=48712 I had the recent CVS version of bzflag (2001/03/27), patched and recompiled it with this patch. The result ist that i cannot start bzflag anymore. It opens its screen and the last output i see is "loading fixedmr" (was it?) in the console-area. My System: Linux, Suse6.3, RivaTNT, XFree3.3.6 (do you need more information?) ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-03-26 14:18 Message: Logged In: YES user_id=25311 Updated for latest CVS again. TESTERS WANTED! ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-03-20 12:18 Message: Logged In: YES user_id=25311 Doesn't create & destroy the list anymore, also updated for latest CVS. ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-03-17 15:00 Message: Logged In: YES user_id=25311 -Corrected loading colors from config file - now, when the "custom" option is selected, BZFlag uses the colors from the config file. -Corrected opporite colors problem. -Did what I said I was going to do in #3. -Still needs testing! ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-03-17 13:05 Message: Logged In: YES user_id=25311 Added option menu item for the enhanced radar for those who don't want it. I have a few more things to sort out: 1. The radar colors option shows the opposite of what you have selected sometimes. 2. The colors loaded from the config file no longer work. 3. I plan on making the enhanced radar slightly more intelligent: I plan on making it so that as long as your z- coordinate is between the top & bottom of the object, it is completely opaque and colored. 4. It has to be tested more. ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-03-16 14:52 Message: Logged In: YES user_id=25311 Tried to fix memory leak. Tell me if it works. ---------------------------------------------------------------------- Comment By: Jeremiah (cobraa1) Date: 2001-03-15 15:21 Message: Logged In: YES user_id=25311 Changed to include an additional option menu item to adjust brightness of radar. I plan on also adding an option to switch between the old & new radar looks. ---------------------------------------------------------------------- You can respond by visiting: http://sourceforge.net/tracker/?func=detail&atid=303248&aid=408433&group_id=3248 |
