Non-trivial component calculation times
 

I have noticed for a while that SE4 dosen't handle large numbers of component calculations well.

Today, I just retrofitted my battlemoon from the ~150 component "Stage Two" of a retroseries, to its final, hulking 453 component "Stage Three". http://www.shrapnelgames.com/ubb/images/icons/icon7.gif
However, it took about five seconds to complete the operation!
The fourth, and possibly final form of my BattleMoon will have another 300 components. 753 components total. (500 of them armor http://www.shrapnelgames.com/ubb/images/icons/icon7.gif)

I've also noticed in combat, that Heavy-mount torpedoes hitting small-tile armor take a noticable time between image impact, and the explosion image.

80 components per second seems reasonably fast, but P&N seems to overwhelm it at times...

I can forsee very long battle calculations in a PBW P&N game, especially if you get two races with heavy armor use. With 200 PPA components per DN (a ship would easily be 20% defenses), you're talking 2+ seconds just to destroy the ship, never mind the moving and firing phases. Moderate sized fleets (20 DN or equivalent) could take over a minute just to explode after they've been hit!
Don't even think about putting 2000 armor on a battlemoon and attacking it with fighters! http://www.shrapnelgames.com/ubb/ima...ons/icon12.gif

-----------
Running on a 1GHz Pentium, WinME

[This message has been edited by suicide_junkie (edited 30 October 2001).]

Re: Non-trivial component calculation times
 

That's an interesting observation. I wonder what takes so long? You'd think that components in a ship could be calculated in an unmeasurable fraction of a second. I can see some delays in combat, since there are animations and random selection of components to hit and so on... but what could it be doing at refit that takes so much calculation?

[This message has been edited by Baron Munchausen (edited 30 October 2001).]

Re: Non-trivial component calculation times
 

reading each instance of a component from the components file, instead of only once?

Re: Non-trivial component calculation times
 

Which would make sense if you were doing testing by constantly updating and modifying the components file, but definately slow down the calculation time...

Re: Non-trivial component calculation times
 

it would? im having a hard time seeing why it would be valueable to read the component file 400 times instead of just once and multiplying by 400. then again, it might not be doing that at all. it was just a guess.

------------------
"...the green, sticky spawn of the stars"
(with apologies to H.P.L.)