** Version 1.1 Announcement **

[Kerry]

Hi there. I have a quick question for you. Where in Canada can you buy the boxed version of the aquarium? I really want the jewel case and cd. I have been looking for it and can't seem to find it anywhere. Or, alternatively, can I get it right from the website?

Thanks )

[ESHIREY]

Hey that makes a lot of sence , I'll get right to work on it .......hehehehehehe . I wish I knew what you were talking about but as long as you and Mr Sachs does , that's all that counts . Right .

[Jim Sachs]

That's how I do it now, except I allow the fish to pass around the coral on the left side, too.

There are a couple of problems with this. First, where "is" the fish? I use the 0,0 position of the mesh, but the fins stick out quite far from that at their 90-degree position.

Second, the fish is moving. When it wants to start its turn, the position may be legal, but still interfere with the coral or other fish during the turn.

Bear in mind that I obviously know how to "strong-arm" the fish and force it on a mechanical path which would keep it away from the coral at all times, but the whole difference between my aquarium and the rest is the lifelike nature of the movement. In giving the fish free will, all I can do is "suggest" that it stay away from the coral. Making it WANT to stay away from the coral is where the problem lies.

[ESHIREY]

Damn ........ There sure is a lot involved to make this all work . When you do get it fixed I think everyone around should buy you a drink . Because god knows you will deserve it . Not saying your a drinking man , just the thought .

[feldon34]

Kerry,

If the Aquarium is available in Canada in stores, Prolific hasn't told me where.

Prolific is in an interesting situation because they want people to buy the Aquarium online instead of in stores, so that they make a few more bucks and so does Jim. In-store sales only make Jim about $4.

[Kerry]

thanks Morgan. so there is no way that I can get the cd and jewel case at all?

[feldon34]

Sure there is.

Send an e-mail to Service@serenescreen.com asking for as many copies of the boxed version and jewel case version as you need. $19.95 for the boxed, $14.99 for the jewel case, $5 shipping. (Confirm details with them)

[Jim Sachs]

If you absolutely must, then go ahead, but Prolific would really like to avoid selling individual copies. The time it takes to process, package and ship one copy to an individual is more than it takes to ship a 16-unit carton to a store, because the carton is already boxed.

[peter piscis]

re:
----------------------------------------------------------------------------
"In giving the fish free will, all I can do is "suggest" that it stay away from the coral. Making it WANT to stay away from the coral is where the problem lies"
----------------------------------------------------------------------------
These complexities are fascinating to we mere mortals who just appreciate the way the fish move. Is a possible solution something like adding a "strong nuclear force" (well, a repulsive one, rather than attractive) that acts at close distances to suggest that the "particles" stay away from each other?

[Jim Sachs]

Yes, in fact I use a variation of that now. But it's still not a natural-looking effect, like someone running toward an object but prevented from reaching it by an unseen force.

[peter piscis]

re:
----------------------------------------------------------------
"running toward an object but prevented from reaching it by an unseen force"
----------------------------------------------------------------
Thanks Jim. Since you're in a conversational mood, what if the force acted on the fish's desired direction of motion and how hard it turned to get going that way, rather than acting directly to move it's position? Though I've got the feeling you've been down this path many times long ago...

[Jim Sachs]

That's exactly what I'm trying to do. The amount of "lead time" is the main problem. By the time a potential collision is noticed and inputs have been put in place to use the fish's natural movements to avoid it, the pass-through has already happened.

[David]

It seems to me that the fish needs to check its own decisions to see if a decision will take it on a collision course. For example if a fish is moving forward at a certain rate and “decides” to turn right a certain amount, then he needs to check the calculated path ahead for five or seven inches or so. If a collision is detected then the fish needs to cancel the decision and make a new decision.

When calculating a collision the fish needs to check not just it’s center coordinate but several points outlining it’s maximum dimensions. This would include the tips of it’s side fins and probably it’s top and bottom most points and it’s nose.

[Jim Sachs]

Most of that takes math. I have none.

[ESHIREY]

Sorry to hear you're having so much trouble with the little guy . If it turns out to be to much , Just put a shark in , and as soon as the lion fish get's to close to something , the shark eats him ...... hehehehehe No but seriously , I know you'll be able to fix it . I look at everything you have done so far an I just know you will beat it . I wish it wouldn't take so long , but like I said before , all good things take time .

[fubarhq]

Jim,

sounds like this is going to be revisited when you go ahead and make the 3D background for the aquarium down the road? I wouldn't want to attack that problem personally? In any event, good luck and we are all eagerly awaiting...

[Jim Sachs]

That's the most frustrating part. All this work goes out the window when the 3D background is done.

[Darrell L]

Jim, are you making collision detection calculations in devicespace, viewspace, or worldspace?

Worldspace calculations would work with both your current planar background as well as your future 3D background.

There are also a number of D3D api functions to help with volumetric calculations (such as mentioned above) as well as transformations which could be used to rotate the lionfish's dorsal fin triangles (treated as a whole) around an "articulated joint" vertex, providing you with a mechanism to allow the fish to proceed with its turn "between" the coral while "folding back" its dorsal fins against its body.

Just a thought from a monkey.

[feldon34]

Jim, are you making collision detection calculations in devicespace, viewspace, or worldspace?

None of the above. He is using an "area avoidance" routine which does not take into account the shape or geometry of the object but only it's maximum radius at any point.

[Jim Sachs]

Darrell - at the moment, I'm using Magnitude() to determine the distance between the two objects, then adjusting the speed, turn, and pitch of the fish depending on whether they are coming toward each other, overtaking, etc.

Regarding the coral avoidance, your idea of keeping the inboard pectoral fin on the Lionfish folded in (at least partially) during the turn is a pretty good one. It could easily be implemented by juggling the min/max angles for the fins. I'll look into that.