Lee has asked me to pop into this thread and give all you anxious people a quick update. Hopefully you guys are aware of how GG numbered the bugs and the ones that he has passed on, so the following will have some meaning for you all:
First the bad news:
Compiler bugs, will not be fixed by me.
05 Duplicate label bug
18 Add pck to exe problem
Now the not so bad news:
Not a bug.
13 Turn camera problem
- the angles returned are not in error, which may not be obvious. Unfortunately, there are many years of code that may rely on the existing system, and I'm not keen on breaking that existing code. I may revisit this one at the end of the current bug hunt.
15 Delete cloned object bug
- it's the way it was designed to work. Any objects that rely on information within other object cannot survive that object's demise. Right now, if you delete the original, any instances or clones with shared data will also be deleted.
And finally the good news:
Bugs fixed
04 system dmem available bug
- Relatively major clear-out of unneeded code as there was a lot of stuff being done that didn't need to be done and that was causing the problems. As a result, this command should also be faster.
08 create light maps crash
- Various pieces of memory were being cleared incorrectly. In addition, the end of the program would also cause a crash when accessing a null pointer.
09 instanced animation bug
- Ensure that animation is carried out for animating objects that have been instanced, whether they are on-screen or not.
10 fix object pivot
- Apply pivot to object collision boxes.
17 texture backdrop fog bug
- Render the backdrop without fog, then reinstate the fog if necessary when done.
27 GameFX problem
- Caused by the same issues as in bug 08.
28 multiple camera bug
- This one was caused by an attempt to avoid preparing a camera for rendering when it wasn't needed (ie deactivated using the SYNC MASK). I removed all of the prediction code for this and simply moved the camera preparation code into its own function and made that the first step in the rendering pipeline.
That leaves another 11 bugs to go (that totals 22 - Lee snuck another 2 into the list, one of them today!)
I'm not on a fixed timetable, so no pestering me guys (or not too much, at least)