From: Kostya Vasilyev <kostik@verio.com>
  To  : ggi-develop@eskimo.com
  Date: Fri, 14 Aug 1998 08:40:32 -0700

RE: SVGAlib wrapper now actually works!

: -----Original Message-----
: From: wolfwings@arenda.lightspeed.net
: [mailto:wolfwings@arenda.lightspeed.net]On Behalf Of WolfWings
: ShadowFlight
: Sent: Friday, August 14, 1998 5:23 AM
: To: ggi-develop@eskimo.com
: Subject: RE: SVGAlib wrapper now actually works!
:
:
: On Thu, 13 Aug 1998, Kostya Vasilyev wrote:
:
: >15.8 fps == 63.29 ms per frame
: >14.4 fps == 69.44 ms per frame
: >diff.    ==  6.15 ms per frame
: >
: >or at 200 MHz it's 1.23 MILLION cycles.
: >
: >now let's re-do the math at higher fps:
: >
: >60 fps == 16.66 ms per frame
: >+ 6.15 == 22.81 ms per frame
: >which is  43.82 fps
: >
: >this is pretty significant loss of performance.
:
: That's a truly unfair measurement.
:
: If the computer is sped up to reach 60fps, the performance loss will be
: proportional, not fixed as you calculated it. Using bc:
: 14.4/15.8 (To calculate the percentage difference)
: 0.911329240506392113924
: Now, taking that figure (I'll keep using 14.4/15.8) and multiplying that
: times 60.0 fps, bc gives me:
: 54.68354430379746835440
: Which, when one looks at it, is a very reasonable performance loss indeed.
: Not quite a 9% loss, and at 60fps it only reduced things to around
: 54.68fps, not all the way down to 43.82fps, I'm sorry. :-)

The assumption in my math is that the game is sped up to run at 60 fps (on
same hardware), not the computer. Or, that another, faster game is run by
same user on same hardware. In that case, the SVGAlib wrapper overhead will
stay fixed.

However, as Marcus already pointed out, the SVGA wrapper is for
compatibility only, and native libGGI apps should work faster.

~kostik