WE NAME THIS ONE "EV0003"

The new release adds some bugfixes, improves the compatibility in general and adds new features.

Original post by Ghabry

Compared to the current development state the half a year ago released 0.1.1 version is already quite dated and has bad compatibility. We continue our naming scheme for the 0.1 releases and name this one "EV0003".

The new release adds some bugfixes, improves the compatibility in general and adds new features. Furthermore the build system was improved and should compile better on Unix systems. Bad news first: No, battle and save/load are still missing. They are planned for the 0.2 release we expect to get released this year.

Summary of changes (issue number is linked and contains further information):

  • Fixes for the games Yume Nikki (#115) and Ib (#151)
  • Status scene (#127) (see picture below)
  • Debug scene (#133) (see picture below)
  • More move commands implemented (#124)
  • Problems with picture rotation fixed (#134)
  • Writing non-ascii characters (basicly all non-English characters like äöü or あえい) to files caused corruption. This broke savegame writing and our tool LCF2XML (#135)
  • Incorrect charset rendering when charset has non-default size fixed (#137)
  • Mistakes in the RTP translation table corrected (that one is used to run japanese RPG Maker games with an English RTP and vice versa) (#149)
  • Add screen effects (Tone/Tint and flash). Tone change was already supported for normal pictures but was very slow, the algorithm was replaced with a faster one. (#118, #153)
  • Weather rendering was refactored and is faster now, too. (#118, #153) (see performance notes below)


Only for developers:

  • Build system improvements for Unix based systems (#128, #130, #132, #136 and many others)
  • OpenDingux Makefile updated (#144)


Performance notes:

Tone change and weather rendering are quite demanding and can cause lag on slow systems. I made some tests to approximate the cpu usage. Results:

  • Weather increases CPU usage by 25% (1.25 times more)
  • Tone change by 100% (2 times)
  • Reducing saturation by 300% (4 times)
  • Increasing saturation by 400% (5 times)


Calculating the CPU usage for combined effects is left open as an exercise for the reader.

Gallery:


Debug scene


Status scene


Darkened picture with weather effect in background


Tint Screen Effect

Posts

Pages: 1
TheRexion
N
1424
10/18/2013 03:32 PM
And... how powerful/weak is the CPU you tested with? I think that ought to have been important enough to include specs about if you're including information about CPU usage.
fdelapena
180
10/18/2013 06:15 PM
Current graphics implementation is pure software rendering (pixman), so it's using pure CPU for pixel graphic operations (blending, opacity, scaling, etc.). As far as I know, it was tested on a Core i7 Sandy Bridge CPU. Frames are limited to 60 fps by default, I don't know how it was tested (profiling maybe). Player features automatic frameskipping to prevent lag where possible.

Player is pretty usable on a Nintendo Wii and it should work fine with low end handhelds, however, it would work without problem when going hardware accelerated (using OpenGL and Direct3D with SDL2), but this has lesser priority for now.

Currently, it should be faster than original RPG Maker in modern computers. It's because RPG Maker uses old DirectDraw operations and most of them are deprecated and might be unsupported in modern graphic cards and drivers.

So, about hardware requirements will be very variant in the future, depending on hardware features, not only CPU.
Pages: 1