Navigate
A Closer Look At Tomb Raider III
From the moment that the demo for Tomb Raider III was released, a flurry of complaints have been raised about the latest installation of the Tomb Raider franchise. I just wanted to comment quickly on the most common complaints and problems that people experience. Again, I am not an employee of Core Design or Eidos Interactive. I am not here to defend or criticize anyone. These are just my opinions and comments. I have broken down these complaints into two categories: Bugs and Gripes.
The so called "corner bug" and clipping in general are problems that have plagued the Tomb Raider series since the very first game. Both these bugs are related. Allow me to quickly define them. Clipping results when solid objects pass through one another. Often times this means you can see through objects, parts of Lara or an enemy can actually pass part way through a solid wall.

If Lara is facing diagonally while standing in a corner and adjacent to the corner of a wall, when she starts jumping up and down in place, she will begin to creep forward toward the corner. This creeping forward is a feature of jumping in place, however, as she creeps forward, more and more of her body will become embedded into the corner of the wall. When enough of Lara is embedded in the wall, the collision detection of the game engine decides that it is impossible for Lara and the wall to occupy the same space and so it resolves the conflict by placing Lara in the correct coordinates in the horizontal plane but on the first available surface from the top. As a result, Lara appears to suddenly "pop" onto the top of the wall. This ability to get Lara up onto structures that she is not meant to get to is known as the "corner bug".

If you ask ten different people, you are likely to get as many complaints about the corner bug as you would praises for it. Many players find entertainment in "teleporting" Lara to places that she isn't meant to go. In contrast, however, ask ten people how they feel about clipping, most (if not all) will agree that clipping is bad.

It's interesting that the bug that results from clipping can be tolerated, while clipping itself cannot. Of course the reason is that clipping presents itself as a visual flaw in the game, whereas people who dislike the corner bug can simply refrain from taking advantage of it.

Pretty cut and dry, wouldn't you say? You're probably wondering why Core programmers haven't simply figured out what causes the bug and simply do away with it. Well, it's not so simple. One reason that this problem hasn't been fixed is that collision detection is a costly operation. We don't mean costly to fix, but rather costly for a computer processor to calculate. Many game programmers choose to leave some amount of clipping in their final program to reduce the game's minimum CPU requirement. To attest to this fact, I was able to successfully install and play Tomb Raider III on my 100 MHz non-MMX Pentium computer with a 3Dfx Voodoo2 accelerator card. Another player reports that he was able to play Tomb Raider III without any problems on his 90 MHz Pentium computer with a 3Dfx Voodoo Banshee video card. Granted that both these computers have 3D hardware acceleration, but the fact that these computers, which are rated at well below the recommended minimum CPU speed of 133 MHz, can run Tomb Raider III shows that the programmers were able find compromises to allow the game to run on slower machines.

For clipping to completely go away, the minimum requirements for Tomb Raider would likely increase beyond what it is currently. That may not seem like a big deal for people who own 450MHz Pentium II computers, but for people who do not want to have to upgrade their old PC simply to play the latest video games, that would be bad news.

Keep in mind, that although the Tomb Raider games have suffered from clipping, they have never suffered from "fogging". The term "fogging" refers to a technique in which programmers cause things in the distance to be grayed out, and as the player heads in a given direction, the game engine renders more of the viewable area. This is similar to walking around in the fog, where you are only able to see a short distance in front of you. Tomb Raider employs no fogging whatsoever, so the player can see to the limit of the map. Although in many levels of the Tomb Raider game, the map consist of a small room, there are many other levels in which the room is large and spacious. The programmers artfully use lighting effects to darken the distant objects to speed rendering. To the player, you don't feel as though you're walking through the fog, but rather it is just the natural lighting.

If my choices were between fixing the clipping problems and raising the minimum CPU requirements or leaving the clipping problems alone, I would choose to leave them because in the long run, the clipping is fairly tolerable and it would cost me less than upgrading my computer.
A related problem to clipping results when Lara kills an enemy very close to a step, a slope, or a ledge. Often when the dead enemy falls, it will land and its corpse will lie perfectly horizontally on what ever surface it was last on. As a result, if an enemy dies on a slope or near the edge of a step, often times the creature will fall and a significant portion of the enemy's body will float in midair, while a foot or a toe is on or in a ledge.

This is the result of two separate issues. The first is the clipping problem. The game engine apparently does not bother calculating collisions when an enemy is killed. The second problem is that enemies are rendered using a small fixed number of polygons. The enemy's appearance is texture mapped to the outside of the polygons and the enemy has a fixed set of animations. Therefore the movements and three dimensional appearance of enemies is fixed. Without more complicated animation routines or a skeletal based model, a fallen enemy is basically rendered as a rigid object, place at whatever horizontal level the engine last rendered it at.

Once again the problem we face is one that can easily be solved by throwing more computing power at it and investing more energy in developing more realistic computer models. The resulting compromise is the same.
One bug that was identified by various persons is that of a "memory leak". A "memory leak" occurs when a computer program allocates a certain amount of memory, but over the course of its execution does not deallocate or "free" that memory. As a result the operating system detects that the program is using more and more memory and none of it is being released back to the operating system for other programs, including the operating system itself, to use. There are two common causes for memory leaks.

The first has to do with memory that is simply lost or forgotten about. This happens when a program allocates some piece of memory for its own use, and keeps a handle or pointer to that piece of memory, and then later uses that handle or pointer to reference a new piece of memory, without first releasing the previous contents. The other common cause of memory leaks is when a program completes its execution and not all of the allocated memory was equally deallocated. The operating system thinks there is still an amount of memory that is in use by the program, even though the program has quit. Many modern operating systems keep track of programs and the memory they allocate, and free all the memory associated with a given program when the program completes execution.

Users report that the Tomb Raider III game would crash or quit suddenly if they save and reload their saved games repeatedly. After a fixed number of reloads, they can predictably crash the program. This is a classic symptom of memory leaks. Depending on the operating system, as you leak away more and more memory, you can start using up hard disk space which the system is using as "swap space" or "virtual memory" and eventually the whole computer can become unstable.

It is easy to criticize the programming team for this kind of bug. In general, memory leaks are the result of poor programming, poor testing, or poor debugging. The best way to avoid memory leaks in programming is to program in such a way that you avoid them in the first place. This may mean using programming techniques that automatically free memory when a program is done using it, or it may mean implementing a system in which the programmer keeps track of when memory is allocated and so is careful about releasing it when the program is done with it. However, this is easy enough to implement when someone is the original author of the program. Tomb Raider III represents the efforts of four years of programming and the work of two mutually exclusive programming teams.

When programmers undertake the task of modifying other people's programs, invariably new bugs will be introduced. Often times it can be memory leaks, if the new programmers do no fully understand how the code works. Speaking as a programmer, memory leaks are notoriously hard to track down and in turn fix. Using sophisticated profiling and debugging tools may make the task of finding memory leaks easier, but in general, its like looking for a small needle in a large haystack. It is a difficult enough task for a programmer to sets out to write "leak-free" code, but it is not impossible for that programmer to adhere to his or her own disciplined programming habits. When someone else gets a hold of that code, then all bets are off.

Core programmers deserve some credit for having released a patch that either fixes or works around the memory leak that was reported. Some users complain that the bug still exists, but they can now save and reload more times than before. I did not experience the bug when I first installed Tomb Raider III. I installed the first patch fairly soon after my initial installation, so I don't have any data with which to make any comparisons. I imagine that it is possible that the programmers merely extended the maximum amount of memory the game is now allowed to allocate. That's not an elegant fix, but it's a finger in the dike that will at least give players enough reloads to satisfy the average player. If you continue to experience this particular bug, I would recommend adding additional memory, freeing up additional hard disk space, or buying an additional hard drive. With 32MB RAM and approximately 700 MB of free space, I have not yet been able to crash my copy of Tomb Raider III. If you experience these crashes frequently, I recommend that you reboot your computer which may help clear up some of the system's instability.
A bug exists in the original executable program, where if you saved the game in certain rooms, invisible walls would appear in other locations in the game preventing you from ever being able to complete the game. I was first notified of this bug by an Eidos Interactive customer support representative.

I do no know, for a fact, what causes this bug, but I speculate that it has something to do with the corruption of the saved game data. Upon reload, the corrupted information is loaded into the current map data, and as a result, invisible walls occur in certain locations of the level.

Core released the fix for known occurrences of this bug as part of their first patch. I have not experienced this bug, though I admit I haven't gotten very far into the game. I'm glad that Core released a fix for this bug because workarounds were originally suggested, such as not saving in the second Shiva room in Temple Ruins, but I think its wrong to expect the consumer to workaround an issue that was caused by the manufacturer. Fixing the bug and releasing the patch was the right thing to do. As for whether or not consumers should have to patch their purchased games, that remains the topic of another discussion.
A common horror story that was reported early after Tomb Raider III was released was the dreaded message that the Tomb Raider CDROM was not in the CDROM drive when people attempted to run the initial installation program.

This is the result of a CDROM trick that Core programmers attempted to use to discourage the reproduction of their CDROMS. This involves "spoofing" the directory information on the Tomb Raider III CDROM. What is done is a couple files are created on the CDROM. The files are actually quite small, but their directory entry is changed such that they appear to be files that are 640 MB in size. Well, any savvy computer user knows that a normal CDROM can only hold 640 MB of data. As a result any attempt to duplicate the CDROM will result in various errors, ranging from the apparent inability to fit all the contents of the Tomb Raider III CDROM onto the copy CDROM to the inability to read the small file, whose directory information is inaccurate.

An unforeseen side effect of this copy protection scheme was that many people own CDROM drives that are then unable to read any of the files off of the CDROM. Eidos Interactive did not want their copy protection scheme to prevent people from playing their game or requiring that these people should buy a new CDROM drive. Instead, they plan to release a second patch for registered owners of Tomb Raider III, but currently have a form set up on their web site to communicate with users that are experiencing this CDROM access problem (http://www.eidosinteractive.com/techsupp/tr3/tr3-patches-disclaimer.html).
Another common complaint is related to sound, video, and 3D accelerator card problems. Unfairly, people have labeled these problems as bugs with the Tomb Raider game engine. I think, the cause of most of these problems are improper installations of Windows 95/98 and/or DirectX 6.0, or the presence of outdated drivers for the hardware.

I managed to get DirectX 6.0 and Tomb Raider III installed on my computer without any problems. The game's setup program recognized the sound card that I had installed and allowed me to choose the appropriate DirectSound driver. The setup program also detected my Voodoo2 3D hardware accelerator, and I was able to choose a resolution that I liked.
Aside from bugs, there are a few common complaints about changes in Tomb Raider III that are not actual bugs. The most common complaint voiced about the Tomb Raider games is that the game engine is now four years old and each successive game is just "more of the same".

It's all very easy for critics to harp at the fact that the Tomb Raider engine hasn't been updated, because most of them have access to the fastest computers with the latest in 3D acceleration hardware. For the average computer user who can't afford to upgrade his computer every other year, it is quite a feat to produce three installments of the Tomb Raider that can run on older machines.

Regardless of what people think about the game engine, each installment has given us new moves and new effects, and the subsequent games are not merely expansion packs to the original game. I believe one major reason the Tomb Raider engine has not been rewritten in the last couple of years is due to a contract between Eidos and Sony in which Eidos agrees to deliver a certain number of Tomb Raider games for the Sony Playstation for some number of years. The Sony Playstation's CPU and memory limitations restricts what can be done with the Tomb Raider game engine. It is certainly easier to continue to tweak the previous engine than attempt to rewrite the game engine with all new features that is still capable of being run on the same old hardware.
Perhaps the single most controversial and vocal complaint came from players that wanted Core Design to release their level editor for public use so that gamers can create their own levels. Core maintains that their in-house level editor is not suitable for public use and would require much work to deliver a product that would be suitable. Core has also flexed their legal muscle to reiterate that they will not tolerate reverse engineering of the Tomb Raider games. A few users suggest that Core simply release the specifications for the game's level's and leave it to enterprising gamers to write a level editor.

In this area, I have to stand with Core for their decision not to release a level editor. The first and most obvious problem that Core is faced with in trying to develop a level editor is that it takes valuable time and effort to build a good product for release. If Core were to invest any new effort, I believe that they should invest it in working on the new games and not trying to release a level editor.

The second problem is more subtle. For Tomb Raider to continue to be a success, the game's image and mystique must be maintained. Tomb Raider levels are not merely a collection of rooms and passageways in which the player runs around and blows away enemies. The game play hinges on unique and challenging puzzles that make use of the moves available to Lara Croft. This task is daunting for professional level designers, what makes an average gamer think that he or she is capable of designing an equally interesting level? User released levels would have no cut-scenes and no video sequences, and hence have no plot. If the market was flooded with mediocre or even bad user-generated levels in which Lara Croft's main object is to get from point 'A' to point 'B' while killing everything in sight, then the Tomb Raider image would be destroyed and that which first attracted players to the game would be lost.
One of the more mind boggling of the changes in Tomb Raider III are the disappearance of certain features that were availabe in the previous games.

The most noticeable missing feature is the removal of the side step keys from the PC version. It has been told that Tomb Raider III was written for the Sony Playstation and was then ported to the PC platform. This being the case, two of the Playstation controller's buttons had to be allocated to the new of sprint and crouch moves, and apparently it was the side step buttons that were replaced. Ingeniously, Core programmers integrated side stepping into part of the walk button's features, but now the arrow keys don't behave the same for walk as they do for running. Although we understand the console game controller's limitations, I don't think it's right to change the PC user's interface to a program that people are familiar with. Some players praise the change, noting that now you don't have to take your fingers off the arrow keys to side step, but I still find myself reaching up to the Delete and Page Down keys when I want to get Lara to side step.

Another notable feature of the previous Tomb Raider games and even the Tomb Raider 3 demo that is now missing from the full game is the preview of the items that Lara has just picked up. It was a fantastic move to replace the cheesy bitmapped pickups with real 3D objects, but not to be able to see a listing of what Lara has picked up can be quite annoying, particularly for walkthrough authors like myself. Having the pick up preview was great for documenting each of the items. I could pause the game when the items appear on the screen and carefully note down everything that Lara picked up. Now I have to stand in front of objects and carefully examine the pile of stuff to see if I can discern what items are there before picking them up. Sometimes I have to check the inventory before and after picking up an object to make sure that there was only one of that particular object there. Gladly, most pickups so far are stacked up in such a way that it is fairly obvious what's there. However, I can't shake this feeling that I've misdocumented one or more of the pickups.

Another minor feature that is missing in action is the "oof" and "aha" exclamations that Lara makes when she runs into a wall at high speed or picks up an item, respectively. Though these things are minor, they were noticeably absent from the latest game. Although at one point I complained that the "oof" sound was annoying when Lara bumped into things, but now I really miss those touches. I guess it goes to show, be careful what you wish for because you might get what you ask.
Many people have complained that Core did not live up to all their promises. Long lists of other features were mentioned but never came to fruition in the final release. Features such as A3D sound and footprints in the snow did not make the Christmas release. Many moves were mentioned in interviews, but did not find there way into the demo or the final release version.

Well, with a pre-Christmas deadline, less than 11 months to work on the game, and a long list of new features, it's obvious cuts must be made. Perhaps to quell disappointment, Eidos and Core should take care not to over-hype their product by promising the moon and delivering a dud, but rather work quietly on their product deliver a solid product. Without false expectations, far fewer players will complain.
One thing that is becoming more and more apparent in the Tomb Raider sequels is the fact that Lara is raiding fewer and fewer tombs. The settings or locations for the various levels of the Tomb Raider games are wandering more and more away from those that gave the first game personality and ambiance.

There are yet many locations left to explore on this green Earth. Granted, things will get right boring right quickly if all Lara ever saw were the insides of damp caves. Perhaps Tomb Raider has reached the end of its life span. Would people continue to purchase and play Tomb Raider sequels if they consisted of only ruins to explore? I can't say for certain, but I believe it's certainly possible.
Plenty of senseless violence may appeal to some, but by and large, the votes have been tallied and I believe that the majority believe that Tomb Raider III includes too much killing. You may argue self defense, but if you trespass on a U.S. Government installation, you can expect to be met by angry military police.

What's so wrong with reducing Lara's character to that of a murderous thief? It must be remembered that we, the players, are controlling Lara, the character. When the player sympathizes and bonds to the in-game character, their interest is captured and their continued game play is guaranteed. I believe this was what happened with the original Tomb Raider game. In spite of Lara Croft's over-inflated ballasts, her character was endearing to people.

In one interview with the creators of the original Tomb Raider, the interviewer asked whether making the main character female was a marketing decision. The answer was that a female character was chosen because of its appeal to both men and women. They thought that the moves were simply more graceful and appropriate for the female form, but also the creators commented that perhaps, male players would be more inclined to play better and work harder so that, in some sick macho way, they were protecting Lara. The interviewer, a man, snorted, but admitted that he wasn't sure if the snort was out of contempt for the statement or the guilt of identifying with it.

In the original Tomb Raider game, the Lara Croft character turned her guns on enemies only when necessary, and most of those enemies were animals or mutant creatures. That game portrayed her as a resourceful adenturer, and not the cold blooded killer she seems to have become in the latest installment. That's the Lara that people have come to like and it's my hope that Core will work hard to restore that image in future games in this series.
Another less common complaint that has been voiced is that Tomb Raider III is too difficult a game. Some would argue that Tomb Raider III is not difficult enough, but most of those would be people who have played the previous games. It is true that after a third sequel, people will have developed many of the skills required to succeed in Tomb Raider games.

For novices, however, the game may seem overly difficult. People have suggested the possibility of varying difficulty settings. Though I agree that difficulty settings would be a nice feature, I believe that ultimately, new players can manage to learn everything they need to know if they spent more time in the training level. If nothing else, new players are less likely to stumble over the new side step keys as veteran players. For people who feel that the puzzles are too difficult, they should realize that there are plenty of resources, available for free, on the world wide web.
The final gripe that I've seen, is one that I personally crusade for. I believe that a lack of a strong plot to stretch throughout the latest game has really hurt this game's enjoyability. People don't come out and say that the game lacks plot. They're actually not sure of what they're missing, but a long series of questions are raised that clearly indicate that more work should have been invested building up the story for Tomb Raider III.

Questions asking why Lara is in Nevada, or why does she suddenly start a level on a rooftop in London are common. What was Lara doing in India in the first place? Our information from members of the Tomb Raider development team suggests that the locations are first chosen and then a plot is written to tie the levels together. With what the script writer is working with, I think she did a good job.

I think this lack of story is excusable in that the developers were working hard to create a good game, and not a good story in response to the complaints and suggestions after Tomb Raider II's release. I just hope that Core discovers that a good story makes for a good game.
Conclusions
That's a pretty long list of gripes. I understand that Core and Eidos actually listen and consider the complaints and suggestions of consumers. I think that's a sign of a "great" company and it's quite admirable. I don't know if we as consumers are really justified to gripe. There are certainly forums for us to do so, but ultimately our vote will come from our wallets and not our voices. ichou@trinity.westhost.com