Blogs

A bridge too far...

  • Comments 1
  • Likes

Anybody remember the brouhaha over “missing” bridges in Flight Simulator 2004?

(There’s a beautiful rant here: http://www.flightsim.com/cgi/kds?$=main/op-ed/ed343.htm)

 

What was up with that anyway? Well… lots of things really. A real cluster if you will.

 

We dramatically changed the way we produced scenery for the world, starting with the end of Flight Simulator 2002, and part of that change meant that we had more control of the whole scenery creation process. We also changed some key technologies used in the scenery systemaround the same time. The FS world was littered with content created for FS 95 and 98. Most of that carried on just fine to FS 2000, but by the time FS 2002 and then FS 2004 rolled around, that legacy content looked bad, and/or no longer worked in the product. And when I say it looked bad, I mean very bad. Like it was produced in 1995 or something.

 

Amongst that content were roughly 160 or so bridges. Some famous, some not so much.

 

Now, we’ve been in the process of replacing and adding (and occasionally deleting) landmarks to the world with each version, and bridges are part of that work process, but you can imagine that if you create a 1000 landmark objects, do you really want 16% of those objects to be bridges? Especially when you haven’t done objects that are arguably more important or interesting, either for aerial navigation or for cultural reasons? Heck no!

 

So the bridges had to go.

 

But wait, it's all okay ‘cause we had a much better plan.

 

Why not devise a system that’d put a bridge *everywhere* there’s supposed to be a bridge, or at least everywhere it’s *believable* there should be a bridge? Forget 160 bridges, we’ll do thousands… Sure, a few unique bridges would be replaced with something more generic, but it’d do until we would next be able to do them again.

 

Okey doke. Sounds great.

 

Meanwhile, back on planet reality, remember how I said that we changed the process by which we created scenery? Well, most stuff in the world came from data that had been hand massaged (or created), and road and bridge placement was no different.

 

Which is where we started to run into some problems.

 

A lot of road data was clipped to shorelines (who needs a road under water?), some road data ended inexplicably in the middle of a lake or river. Which meant that we got a lot of missing bridges (‘cause according to the data there was no bridge needed, two roads just happened to end across from each other over a river or lake…), and bridges that ended abruptly in the middle of the water. To say nothing of the tiny bridges generated between to points really really close to each other, or the multiple “fan” bridges spouted for some reason or other…

 

Right solution: imperfect implementation.

 

Most of the missing landmark bridges weren’t replaced with generic versions because of the clipped road problem mentioned above. By the time we realized the magnitude of the problem (that there were that many bridges missing), it was too late (and too risky for the ship schedule) to try and fix or replace with brand new content.

 

At the time, it didn’t seem like it was going to be that big of a deal. True, some content that had been around for awhile would no longer be there, but we *were* adding (even though there were some warts) a bunch of new stuff that should more than make up for the old *ugly* stuff that was now gone. We get requests for world landmark items all the time -- the Flight Sim world has missed having some pretty big landmarks in it’s day (and yes there’s many still missing…)-- and that’s where we tend to focus our efforts, making sure what’s in the product is reasonably acceptable, while adding as much new stuff as possible. Imagine how the product would be if all we did was the same old stuff over and over again. The same places’d look better to be sure, but they’d still be the same places. As far as requests go, occasionally a bridge’ll be requested, but for the most part not. Didn’t think people were really going to miss them much if at all. So we took the list of now gone bridges, and added them to the potential workload for next time.

 

At release of course, there was quite a great hue and cry. We quickly put that list of bridges into production to be ready if there was ever a patch, and if not, at least they’d be there in the next version.

 

Patch we did, so we scrambled to get as many replacements done as possible.

 

In the end, I think the biggest reason it became such a fiasco was there was no way to communicate to the customer not only *what* was changing, but also *why.* Yes, there’d be people who’d be upset that something they’d come to expect had changed or was gone, but the majority of people would know that we didn’t just leave something out of the product out of spite or negligence.

 

Wow. Long post for ole Jasers!

 

I’ll prop this post up, but don’t be surprised if I edit it over the next couple of days. I fear that in long posts like these I lose what I’m really trying to say, and usually after a couple of days I want to rewrite the whole thing.

 

 We’ll see. J

 

Cheers,

 

Jason

Comments
  • Thanks for filling in the blanks about what really happened :)

Your comment has been posted.   Close
Thank you, your comment requires moderation so it may take a while to appear.   Close
Leave a Comment