.. by running a packet frame if the current renderframe is closer to
the time the packet frame *should* be run than the next renderframe
(presumably) will be.
This should also help when using cl_maxfps -1 on a slow system that
can't reach the desired rfps (vid_maxfps or display refresh rate).
Without this change, we might run int the problem described in the
following example:
Imagine having cl_async 1, cl_maxfps 100, no vsync and a system that
mostly only reaches about 60fps. So rfps is 100 and pfps is 50.
But then (without this change) running at 60fps means that only every
second renderframe is a packetframe, so the packetframerate
*effectively* is 30, which can cause movement/clipping/physics bugs
(for example when hugging some non-perpendicular walls, like the right
wall with the window that leads to the last room in base1).
With this change the packet framerate would effectively be 60 (every
renderframe is also a packetframe), which is less buggy and also closer
to 50 than 30 is.
(this is the default value of that cvar now)
In Qcommon_Frame(), if cl_async is 1, the packet framerate should
ideally be a fraction of the render framerate, and to avoid movement
glitches and such it should be between 45 and 90, ideally around 60.
With cl_maxfps set to -1, pfps now is set to such a value automatically.
When setting rfps based on GLimp_GetRefreshRate() and vid_maxfps,
take into account that GLimp_GetRefreshRate() might return a value
that's slightly too low (like 58 or 59 when it's 59.95 and should be 60)
The 20% tolerance that, in case of vsync enabled, used to be handled
with `packetdelta < 0.8 * 1000000/pfps` (and the same for rfps) is now
instead added to rfps (and thus implicitly pfps, if it's >= refreshrate)
It could make things stutter, especially if cl_maxfps was deliberately
set to a fraction of the display refreshrate (as it'd target a few
frames less).
The 0.95 factor was supposed to ensure that we don't have more packet
frames than renderframes (or two packet frames in a row without
rendering in between), as that apparently breaks the movement prediction
code. We now ensure the same thing my not running a packet frame if no
render frame is run at the same time.
because on some systems (like RPi4 with my experimental GLES3 branch)
the overhead of a FBO is really noticeable, so no reason to pay for it
when it's not needed
Can be disabled with gl3_usefbo 0.
Mostly this adds an underwater warping effect, like the soft-renderer,
and also like the vulkan renderer (the shader is based on the vk one).
When this is enabled, the v_blend effect (for fullscreen overlay with
one color, like when hit or to make the screen white-ish when under
water) is now applied in the shader used for rendering the FBO instead
of rendering a fullscreen quad in blendmode.
This fixes the framecounter in the soft renderer when `viewsize` is set
to something smaller than 100. This requires the renerer to rerender
bigger parts of the borders which has a measurable speed impact. About
5 frames less with the framecounter enabled on my system. No impact with
`viewsize 100`.
The GL renderers require that the borders are redrawn after every
glClear() call, the damage tracking doesn't take that into account.
Since the speedup by the damage tracking is neglibiable in the GL
renderers, don't use it. Just redraw everything when we're running
with everthing which isn't the soft renderer.
I think this looks ugly, I always called it nightmare and it was one of
the first changes I've made to Quake II. But for the sake of peace and
quiet change it to be standard conformant. Closes#809.