. /Noctis/Known Noctis IV Bugs, and.../ MESSAGE #70734
absent more than alex
written by Shadowlord on Feb 10, 2006 02:29

You can get it <a href=http://anywherebb.com/postline/result.php?m=70732>HERE


  • Fixed that last annoying crash/freeze bug from R9. (I hit a crash bug in the windows version when another drifter shows up to refuel your drifter. The bug was in drawpv, which is also used on planet surfaces to draw critters. I fixed it in both the DOS and windows version, and it turned out to have been the crash/freeze bug in the DOS version. As for why the Windows version never crashed on planet surfaces for the same reason... Who knows.)
  • And then McWgogs reported another crash bug, which is now fixed too. It was caused by spinning around really fast: If you turned far enough that your beta angle was <-180 or >180, Noctis tried to fix it by adding or subtracting 360. Well, if you turned far enough that your beta+360 was still below -180 or beta-360 was still above 180, then Noctis crashed. (Now it keeps adding/subtracting 180 until your beta angle is in range, so it works properly even if you're spinning around like a spinning top.)
  • Determined the exact amount of memory needed for the PV data buffer, and shrunk it to that amount (DOS version only). This saves 4704 bytes, which just might make the difference in letting some people run NICE who couldn't previously run r9+.
  • SyncStarmap now writes out an up-to-date starmap2.bin after checking for updates and applying them (if there are any). This means for people who use NICE and SyncStarmap, NoctisMapper's map will always be as up-to-date as your copy of the starmap - and it means once you run SyncStarmap, NoctisMapper will include your own findings. (Compare to vanilla Noctis, where someone has to run a program to generate a starmap file for NoctisMapper, and to do the entire Noctis galaxy with the fastest available program of that sort takes several days on a decently fast computer. Needless to say, this was never done often and the last time was when I myself (Shadowlord) did it, after writing Triceratops (the program to do it). Technically, the entire galaxy was never done before then, since evolvesm was orders of magnitude slower than Triceratops)
  • The weird cupola bug in the windows version is fixed. Apparently that was caused by the same thing that caused the DOS crash/freeze bug. (The bug was that the colors of things viewed through the cupola/lander were different in the Windows and DOS versions.)
  • A buffer overflow related to ice has been fixed. This one only happened in areas with ice (it's related to the flicks/flakes/whatever you see in the ice). It was caused by code which sometimes tried to write to n_globes_map with an overly high index (n_globes_map only had about 22 thousand entries, while the code picked an index up to around 32 thousand or so). This bug was causing a "crash" in the Windows version, but it was really the buffer overflow checks detecting an overflow and executing a debugging break interrupt (which looks like a crash if you don't have a debugger attached to the program. If you have a debugger attached, then it immediately pauses the program and takes you to the line of code which called the interrupt). The overflow also existed in the DOS version, and I've fixed it in both the windows and DOS versions now. (However, the same bug was in the original Noctis IV, and I never heard of any weird glitches that only happened in icy areas, so I don't know if it even had any noticeable effects in the DOS version)
  • Also, I noticed a small amount of code for icy areas commented out in the DOS version (which wasn't commented out in the windows version, and wasn't commented out in the original NIV source code). It's no longer commented out, since I don't see a reason for it to be commented out (or any comments stating why it was commented out).
reading this thread
(not applicable to single message display)
. /Noctis/Known Noctis IV Bugs, and.../ MESSAGE #70734
14767, 9 queries, 0.065 s.this frame is part of the AnyNowhere network