posted 08-04-99 07:31 PM ET        

1. Since you mention it, can you please provide at least *one* example, not matter how minor, of a bug being invented and then popularised on the SMAC forums. I can't recall a single one that wasn't either able to be duplicated or else shot down in flames (RTFM et al).

2. How bugfree is the average AAA game (not the average game, but the average AAA one. Using Battlecruiser 3000 to skew the stats is not fair.)? How bugfree is SMAC? Given that on opening the game I found within the first hour that the farm display information was wrong, am I wrong in taking this as being a sign that things were a little lax as far as attention to detail goes (in a game where details are very important, too)?

3. Since all bugs are made by programmers, where does that leave them in this analysis of QA stuff? You make absolutely no mention of them in your article, which is rather like talking social welfare while completely ignoring the economy, to give an example. For instance, anyone who tries to modify the weapon values to give a smaller attack value actually CHANGES the weapon graphic used for that weapon. In effect, the programmer messed this up by using one variable instead of two, like for all the other components. And SMAC seems to have some major memory problems, or at any rate moves those voxhals around the screen quickly for the first ten minutes and then agonisingly slowly after that (I'm talking about the low res caviar, too. Don't get me started on the stuff in the main game...) The A.I is another area that seems to have suffered from poor proceedures, whereby a 10-1-10 needlejet or 4-1-2 rover is roped in by the Opp Eng to serve as a defensive unit for an under siege city, instead of being used as an attack (give the conqueror bias in the game, active defense is extremely important, and the A.I's inability to understand this is a serious flaw).

And it of course widely suspected that a significant amount of SMAC is just cut and pasted straight from CivII (with good reason: the seven A.I slots, with specific behaviours for each, the infinite range missiles being two *examples*).

The point being that I suspect that the best QA in the world (your midas company example) can't survive sloppy programming. As distinct from buggy programming, where a good proceedure fails because of an incorrect value - a BUG.

4. I feel I must ask about your laughable (and oddly paranoid) psych report about on-line communities. What motivates this account, and how do you explain highly successful groups such as the online fans of WarcraftII, DescentII, Starcraft, Diablo (see next question), halflife, and, of course, QUAKE. Many of these groups number in the millions, and while they didn't initially enjoy 100% bug free games (though they were fairly close), they did not need the excuse of whinging about bugs to form thriving communities.

What is more likely - that the fans are desperate to keep the community together, and will invent any excuse, such as multiple fake bug reports, OR - that the fans are upset by a good game being wrecked by a few bugs.

[Developers (especially marketing staff!!) should be very wary of the fan base that builds up before a game is released, too. In my personal experience, both Firaxis and Cavedog have suffered from the release of underdone and badly balanced games (once you figure out the required tactics - another job of QA?).]

5. The myth of the bugless game? Almost certainly on release (and here I agree with you, although, like DescentIII's analysis of 3D accelerator cards, some attempt should be made to inform the customer of EXACTLY what rigs and setups the game has been tested on). Not even Starcraft and Diablo were entirely bug free, although the first bug lists did read like a list of extremely bizzarre things to do with Mutalisks. But the real acid should come on a company AFTER the game is released, the actual bugs are documented, and how they respond to that. You make no mention of this either, despite the fact that patching a game after release is probably the most important issue facing QA types these days. A bug that gets solved in two weeks after release (or with a simultaineous patch) is hardly a bug at all. One that lasts 4 patches into the game is serious. In my extremely humble and ignorant opinion, any patchs after the 2nd-3rd should really be to fix the bugs introduced by the earlier patchs - the game itself, so to speak, should be bugfree.

[From the marketing point of view, this is INCREDIBLY important. Otherwise you WILL get people taking the game straight back to the store, the publisher, or the developers. On the other hand, people will hold out for a solution in a patch long after they buy the game - even if it gets uninstalled and left on the shelf.]

6. Since I keep referring to their games, it's worth asking what Blizzard does that makes them different from Firaxis and Cavedog? All three are small companies trying to make at most a couple of AAA quality games at a time (unlike the merciless production engines of Sierra and Activision, who try to make AAA games purely by accident). All have large and vociferous fan bases. Why does one continually succeed when the others seem to falter?