Author Topic: Games with bugs  (Read 1991 times)

0 Members and 1 Guest are viewing this topic.

Offline Petari

  • Administrator
  • *****
  • Thank You
  • -Given: 2
  • -Receive: 56
  • Join Date: Aug 2010
  • Posts: 609
  • Country: 00
Games with bugs
« on: 10-09-2012, 13:56:25 »
Games with bugs

This should be list of known bugs in game originals.
Bugs like:
Not possible to finish game - probably the worst case.
Diverse errors ocasionaly - like graphic ones - but it may happen only on
certain configs, TOS versions.
Not working on STE - but not because TOS version.
Not working on specific configs or emulators - because silly bug.
Not working with more RAM in machine.
Bad files, corrupted data on original floppies - some releases of game involved.
Other.


Iron Lord:  not possible to finish maze - no stairs down from 6th level.
Additionally, single floppy English release (Action 16)  has corrupted file, and game will
crash after winning battle.

Crazy Cars 2 - after doing all routes, driving thru final location makes nothing
- instead some message, congrat. you can only drive further.

Dyter-07 - game crashes at final level (descent to end boss). Code/data bug.
After some hacking I reached and defeated end boss, and may say that it is
poorest of all them - just to mention.

Trantor - one version of game tend to crash after some playing.
Recogniseable by 45283 bytes long executable.




Bad release:

Tetra Quest - US release has corrupted last 2 files on floppy.

Iron Lord, action 16 release - corrupted file, crashes after winning battle.





Not working on STE - because bugs:

Grand Monster Slam: palette writing overshot - on STE will see lines instead graph.

Hyperforce - same as above.

Nightdawn - writing into video pointer crashes on STE. (On ST does nothing).




Not working in Steem with current Pasti.dll :

Rubicon - there is silly bug in code (forgotten #), what causes both floppy
drives beeing selected at once. On real HW, with solo floppy drive it harms
not, and likely not on 2 drive configs, when second drive is empty. If Pasti is
inactive, it works in Steem.


Not working on higher TOS versions - but not from usual reasons like more
low RAM occupied or Timer-C problem - caused by bugs in code.

Deflector -  stoopid bug with address register: copy (w instead l) in game.

Heimdall - graphic bug appears by running on higher TOS versions. Because
some silly code part.



Not working on specific configs or emulators - because (silly) bug:

Pro Tennis Tour - will crash on 2MB TT, if going straight in tournament - because
uninitialised var. - then try to access RAM area between 2-4MB, what on ST(E) does
nothing, but on TT with 2MB causes bus error.
Well, this is really very specific error, and likely only few people, or even
only me noticed it - but is interesting, and needed some time to track down.

Technocop - crashes on TT, Falcon because silly bug with sr setting.

Rick Dangerous II - '0FFF' bug - and it is present in some others. Makes trouble on
TT only - surely, the reason is mistype by coding, so we have something like: $FFF8800 instead $FFFF8800 or  $FF8800 .  Problems only with 32-bit address bus.

Ivanhoe - '0FFF' bug - see above.

EPIC - If play with mouse on Falcon, will loose controls after little time. The reason is some
strange code (in traps) activated periodically if 68030 is detected - but it does nothing useful.
I guess that because game is released in 1992, they planned Falcon support, but since it delayed game is published without testing on Falcon.


Not working with more RAM in machine:
There is many. For instance Sidewinder.





Other:

Wolfenstein 3D - this is not commercial game, and is not really finished. I found diverse bugs and
partially fixed it. But some levels still make problems. According to coder, he placed crucial parts of code in low 32K RAM, for extra speed. But that area is TOS workspace, so what works under one TOS version, may crash under other. Bad idea for TOS calling game.

Crossland - in audio code IR level is lowered, what may cause activating IR again, before current is finished - and it causes crash. On ST(E) it happens very rare, but on TT for instance after few seconds.

Shiftrix - I don't know on what they tested it, because there is extremely silly bug at start: enabling interrupt before setting it's vector !

Nicky 2 - even it game self needs only 512KB, it crashes with 512K RAM only. Because bug in code.



Special cases, affecting mostly hard disk adaptations (with exit to Desktop opt.) :

Sometimes RAM area above 512KB is intentionally cleared, but there are cases that it is because bugs:

Prehistoric Tale - mistype d7 instead d0 in code.

In many games, mostly graphic code has overshots, writing in higher RAM - Vroom for instance.

IKBD coding seems as big problem - there is a lot of games with bad IKBD code. Usually, playing on regular HW, from floppy you will not notice it. But if doing exit to Desktop from game strange things may happen, if IKBD chip is driven nut. Bad IKBD code may cause problems in Steem too - which has not perfect IKBD emulation. I guess that lack of good documentation is the reason for many bad coding.


If someone of game coders, publishers accidentaly reads this, and has some objections, corrections, infos about updated versions, feel free to write here.

And of course, I expect users that write here own experiences.