Re: [Warzone-dev] Re: [Warzone-commits] r686 - /trunk/src/projectile.c
On Sun, 28 Jan 2007 20:59:17 -0800, The Watermelon [EMAIL PROTECTED] wrote: On 1/28/07, Giel van Schijndel [EMAIL PROTECTED] wrote: Author: muggenhor Date: Sun Jan 28 19:19:32 2007 New Revision: 686 URL: http://svn.gna.org/viewcvs/warzone?rev=686view=rev Log: Modification 8 by Watermelon's bugfix patch: * Changed establishTargetHeight in projectile.c to use pIMD rather than displayImd for structures to fix some weird height problems(hopefully). * Giel: Slightly changed the patch's modifications to increase code readability Refer: Message-ID: [EMAIL PROTECTED] Modified: trunk/src/projectile.c some of your change swapped the correct order and causes crash: * height = asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymax - asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymin;* * psDroid = (DROID*)psTarget;* it tried to access psDroid pointer before it gets initialized as (DROID*)psTarget,the original/correct order is: * psDroid = (DROID*)psTarget; height = asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymax - asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymin;* It seems Ari fixed this in r688. Would have been nice if the compiler would just kick my ass on this sorts of cases (dereferencing uninitialized pointers). -- Giel ___ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev
Re: [Warzone-dev] Re: [Warzone-commits] r686 - /trunk/src/projectile.c
Am Montag, 29. Januar 2007 schrieb Giel van Schijndel: On Sun, 28 Jan 2007 20:59:17 -0800, The Watermelon [EMAIL PROTECTED] wrote: On 1/28/07, Giel van Schijndel [EMAIL PROTECTED] wrote: Author: muggenhor Date: Sun Jan 28 19:19:32 2007 New Revision: 686 URL: http://svn.gna.org/viewcvs/warzone?rev=686view=rev Log: Modification 8 by Watermelon's bugfix patch: * Changed establishTargetHeight in projectile.c to use pIMD rather than displayImd for structures to fix some weird height problems(hopefully). * Giel: Slightly changed the patch's modifications to increase code readability Refer: Message-ID: [EMAIL PROTECTED] Modified: trunk/src/projectile.c some of your change swapped the correct order and causes crash: * height = asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymax - asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymin;* * psDroid = (DROID*)psTarget;* it tried to access psDroid pointer before it gets initialized as (DROID*)psTarget,the original/correct order is: * psDroid = (DROID*)psTarget; height = asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymax - asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymin;* It seems Ari fixed this in r688. Would have been nice if the compiler would just kick my ass on this sorts of cases (dereferencing uninitialized pointers). I thought GCC does with -Wall... pgpUvsnKFNrds.pgp Description: PGP signature ___ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev
Re: [Warzone-dev] Re: [Warzone-commits] r686 - /trunk/src/projectile.c
On Mon, 29 Jan 2007 13:00:43 +0100, Dennis Schridde [EMAIL PROTECTED] wrote: Am Montag, 29. Januar 2007 schrieb Giel van Schijndel: On Sun, 28 Jan 2007 20:59:17 -0800, The Watermelon [EMAIL PROTECTED] wrote: On 1/28/07, Giel van Schijndel [EMAIL PROTECTED] wrote: Author: muggenhor Date: Sun Jan 28 19:19:32 2007 New Revision: 686 URL: http://svn.gna.org/viewcvs/warzone?rev=686view=rev Log: Modification 8 by Watermelon's bugfix patch: * Changed establishTargetHeight in projectile.c to use pIMD rather than displayImd for structures to fix some weird height problems(hopefully). * Giel: Slightly changed the patch's modifications to increase code readability Refer: Message-ID: [EMAIL PROTECTED] Modified: trunk/src/projectile.c some of your change swapped the correct order and causes crash: * height = asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymax - asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymin;* * psDroid = (DROID*)psTarget;* it tried to access psDroid pointer before it gets initialized as (DROID*)psTarget,the original/correct order is: * psDroid = (DROID*)psTarget; height = asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymax - asBodyStats[psDroid-asBits[COMP_BODY].nStat].pIMD-ymin;* It seems Ari fixed this in r688. Would have been nice if the compiler would just kick my ass on this sorts of cases (dereferencing uninitialized pointers). I thought GCC does with -Wall... I believe it does, yes. The only problem is that I need to distinguish those warnings from the ones already present. Hmm, maybe that's a nice idea for some spare time: writing a diff-utility that generates diffs on compiler warnings/errors. Although maybe a bit far fetched. -- Giel ___ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev
Re: [Warzone-dev] Re: [Warzone-commits] r686 - /trunk/src/projectile.c
On 1/29/07, Giel van Schijndel [EMAIL PROTECTED] wrote: Hmm, maybe that's a nice idea for some spare time: writing a diff-utility that generates diffs on compiler warnings/errors. Or just fixing those already present? I have to say, though: adding a ton of new warning flags to gcc for the debug builds may not have been a very bright idea. I think it would be better with the general rule that you do not add a new flag before you have removed the warnings it would generate *first* (ie the flag is there to avoid regressions, not to nag people about your TODO list). - Per ___ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev
Re: [Warzone-dev] Re: [Warzone-commits] r686 - /trunk/src/projectile.c
On Mon, 29 Jan 2007 15:20:54 +0100, Per Inge Mathisen [EMAIL PROTECTED] wrote: On 1/29/07, Giel van Schijndel [EMAIL PROTECTED] wrote: Hmm, maybe that's a nice idea for some spare time: writing a diff-utility that generates diffs on compiler warnings/errors. Or just fixing those already present? Yes, and not commiting code that adds new ones. I have to say, though: adding a ton of new warning flags to gcc for the debug builds may not have been a very bright idea. I think it would be better with the general rule that you do not add a new flag before you have removed the warnings it would generate *first* (ie the flag is there to avoid regressions, not to nag people about your TODO list). Honestly I don't know what you're talking about here. Since only -Wall is currently passed to gcc for generation of compiler warnings. Unless you count -fpermissive along, which actually turns certain (resolvable) compiler errors into warnings. -- Giel ___ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev
Re: [Warzone-dev] Re: [Warzone-commits] r686 - /trunk/src/projectile.c
Am Montag, 29. Januar 2007 schrieb Christian Ohm: On Monday, 29 January 2007 at 14:55, Giel van Schijndel wrote: Hmm, maybe that's a nice idea for some spare time: writing a diff-utility that generates diffs on compiler warnings/errors. Although maybe a bit far fetched. Tried that some time ago. Not perfect, but works. Looks nice! Why did you use tee and not output-redirection to create make-out.tmp? What I wonder about is that with the sort you use lines like ../lib/sound/track.c: In function ‘sound_PauseTrack’: ../lib/sound/track.c:504: warning: comparison between signed and unsigned might (will?) loose their connection... Maybe the script could be enhanced so it can set a reference point (eg before work, so you don't have compare small incremental steps, but can also look a little further back and see not only what you changed in the last minutes, but also what you changed since your last commit.) --Dennis pgpB9UHYxybMt.pgp Description: PGP signature ___ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev
Re: [Warzone-dev] Re: [Warzone-commits] r686 - /trunk/src/projectile.c
On Monday, 29 January 2007 at 16:39, Dennis Schridde wrote: Looks nice! Why did you use tee and not output-redirection to create make-out.tmp? What I wonder about is that with the sort you use lines like ../lib/sound/track.c: In function ‘sound_PauseTrack’: ../lib/sound/track.c:504: warning: comparison between signed and unsigned might (will?) loose their connection... Yeah, and I don't get error messages there as well. That's what the tee is for, you get the usual output for context, and the diff gives the changes in the warning messages. Maybe the script could be enhanced so it can set a reference point (eg before work, so you don't have compare small incremental steps, but can also look a little further back and see not only what you changed in the last minutes, but also what you changed since your last commit.) Well, my script is of the quick and dirty variant, it just works good when compiling the same set of files. -- I can remember when a good politician had to be 75 percent ability and 25 percent actor, but I can well see the day when the reverse could be true. -- Harry Truman ___ Warzone-dev mailing list Warzone-dev@gna.org https://mail.gna.org/listinfo/warzone-dev