KNOWN REPEATABLE AND FIXED PROBLEMS Servants Strike/Plot Hang: In certain cases, it turns out to be possible to hang the plot (i.e., to get the conversations in a state where it is impossible to progress any further in the plot of the game). There are actually two problems here: 1. If you go to Lord British about the strike and, after he tells you he's going to think about it, you then bug him about it before he decides. (He will wait until you've gone off and made some significant progress in the game.) Bug LB too soon and he will get himself stuck thinking forevermore. This means Nystul will remain hors de combat for the rest of the game, and THIS means that the game has become unwinnable. 2. The other, rarer form of this bug is when the fact that Nystul is out of action prevents you from making the plot progress that would get Lord British to finish his decision-making. The patch fixes both, without backtracking or restarting the game. Sleeping with the moonstone on the cursor: Well, we never thought of that, and unfortunately the code doesn't support it. A patch was built, however, that works with savegames in which the problem has already occurred, as well as making sure that it will not break things in the future, and supporting it. We would like to point out the game plot works better doing that puzzle in one of the designed ways -- players who win in the unexpected way will miss important parts of the game plot and design. Once again, sorry. I've done everything, but Nystul wont go through with the endgame: There are two things we know of that can cause this problem, one of which can happen during the normal course of game-play. This is to get the horn from Praecor Loth by just killing him and taking it, as opposed to talking to him and having him give it to you. The endgame sequence checks to see if you have merged 8 key gems, and if you have accidentally gotten into a state where you have merged 9 or more, it will not run correctly. Local Bus Video: Some cards (NCR chipset as shipped with current ZEOS machines, some DELL on-board VGA's) seem to have problems getting the palette. We use a VGA port-based palette load technique to get around problems on some clones with BIOS techniques. It seems these on-board cards need port input slowed down to work correctly. On some this is due to cards missing outs due to the lack of wait states. We have put a patch in (replacing "rep outsb" with "slow: outsb, loop slow" for you assembly programmers) which works on the DELL we have here. Hopefully it will work on other cards as well, but we don't own one of every machine made, so cannot promise. As we hear of particular problems, we will attempt to find/buy the hardware and patch it as well. THE PATCH The patch that fixes the problems described above is available as a self-extracting compressed file, UW2PATCH.EXE. Place it in your main game directory (ie. \UW2) and run it, and it will create 3 files: UW2.EXE, DATA\CNV.ARK, and DATA\STRINGS.PAK. The patch should be available on CompuServe, Genie, and AOL, as well as other major boards, BBS's, and the usual internet sites. OTHER PROBLEMS, COMMENTS, INCOMPATIBILITIES AND REMARKS E001, E002 Check for hard drive space. If you don't have enough (1200K free) you will get these errors fairly often. A003, D014 A003 indicates that the map data is corrupted. Often this is because the file was closed incorrectly last time. We do NOT know why this is happening, and have not been able to reproduce it here. It seems like some cases involve messed up hard drives. Whether this is due to software which is working with the drive (caches and compressors, for instance), rebooting while in applications, we do not know. However, if your drive is corrupted, UW2 will, like any other program, have a chance of misbehaving. We have also seen save games where the files are ok on the drive, but filled with what seems to be random data. This too we suspect is related to some sort of drive manipulating TSR, but cannot be sure. Can't Crunch Bad Objlist Most cases of this we have seen have been ones where the memory on the machine got trashed, somehow, and this was the first visible symptom. Except for a few reports of this happening near a particular liche,which we are checking on, most cases look to have been manifestations of a general system failure on some machines. If you get one of these which is location/game flow reproducible, report it. If instead it seems coupled with screen garbage and other random problems, it is probably a symptom but not a cause. Things to try if you are having inexplicable problems. Chkdsk your drive to see if there are lost chains ("CHKDSK"). Make sure it is okay and there is enough remaining space (1200k). If you are using QEMM, try switching to EMM386. Try a boot disk and clean system. How to help us We tested the game thouroughly, completing many playthroughs and feature checklists, in and out of house. Machines were used running most memory managers, video cards, and a wide variety of playing styles. The final shipping version (F1.90S) ran the game successfully, for over a week, with no A003, objlist, or other significant problems. So, what we need to figure out is what is different about your machine from all of ours. If you have a friend with the game but a different machine setup, see if the savegame which breaks on your machine breaks on theirs. If you are having a problem which does seem repeatable, try and do a minimal number of things and still cause it to break (i.e., if the liche fight seems to be a problem, try saving right before, and then trying to fight, and reduce it to the minimal number of things which cause the problem). That makes it much easier for us to try and understand and reproduce your problem. SOUND Occasional sound slowdown: This seems to happen when running smartdrv w/write behind caching. The actual problem seems to be extended interrupt disabling while smartdrv is writing out the cache data or something, therefore the timer interupt doesn't happen, therefore music slows down. For now, one can either turn off write behind (smartdrv c, for instance, explicit drive reference no + or -) or put up with it. If we can think of a workaround, we will post it, but it isn't our top priority because we cannot think of anything to do about it. Game becomes jerky/digital sound goes away. This seems to happen on some SoundBlaster ROM/Memory manager setups. It once happened under a debugger here, and the SB was not responding to interrupts we were sending it. We do not know why it was not, or what it was doing. Thus, we have seen it happen, but we do not know how to make it happen. We will continue to investigate the code and SB documentation, but the checking of our internal state the time it did happen were fine, so currently we are investigating what the SB was doing. If we do solve the problem, we will make a patch available. So far we do not understand what the hardware is doing. For now, if this starts happening in your game, the easiest way to fix things is to exit and restart the game. Things should be fine from then. The reason the machine is jerky is because the game is sending a digital effect to the SB and waiting for a response. The game pauses while waiting, and then finally gives up on it and goes on. Note: It also seems that sometimes this can trash random memory, in terms of filling the screen with junk and making the mouse act weird and so on, if one plays for a while while this is happening. It is unclear, since we cannot reproduce the problem well, what exact sorts of behavior are possible. If you get strange graphics trashes and such exiting and reentering the game makes everything okay again. Final note: Anyone having particular problems like this is encouraged to describe it and include an autoexec.bat and config.sys and such, especially info on QEMM parameters and sound boards and so on. As always, the wonderous world of infinite variation among memory managers and other memory resident stuff makes it hard to tell exactly what is going on. We are attempting to fix all problems we hear of, but it is hard to diagnose problems on other people's machines, especially when many we do follow up on turn out to take lots of our time and reveal something we had nothing to do with. We want to help out, but we want to get to work on our new rendering engine and new game designs too. Thanks again, and enjoy -UW2 design team and such