Show More
Commit Description:
Optimize PreserveCounts and only recalculate when needed....
Commit Description:
Optimize PreserveCounts and only recalculate when needed. Previously would recalculate preservecounts every Update call (~1 per frame), which isn't necessary when there's no tick. Might be some room to tweak, like doing these updates only when preserves change. Some measurements: This takes about 30ms versus the .25 ms with no preserve (then like .0002ms). When the map is filled up with preserve, about 35ms and 9ms. With a handful of cells, it's more like 0.8ms (before JIT optimizes most of it away).
File last commit:
Show/Diff file:
Action:
FNA/licenses/lzxdecoder.LICENSE
29 lines | 1.1 KiB | text/plain | TextLexer
LzxDecoder.cs was derived from libmspack
Copyright 2003-2004 Stuart Caie
Copyright 2011 Ali Scissons
The LZX method was created by Jonathan Forbes and Tomi Poutanen, adapted
by Microsoft Corporation.
This source file is Dual licensed; meaning the end-user of this source file
may redistribute/modify it under the LGPL 2.1 or MS-PL licenses.
About
-----
This derived work is recognized by Stuart Caie and is authorized to adapt
any changes made to lzxd.c in his libmspack library and will still retain
this dual licensing scheme. Big thanks to Stuart Caie!
This file is a pure C# port of the lzxd.c file from libmspack, with minor
changes towards the decompression of XNB files. The original decompression
software of LZX encoded data was written by Suart Caie in his
libmspack/cabextract projects, which can be located at
http://http://www.cabextract.org.uk/
GNU Lesser General Public License, Version 2.1
----------------------------------------------
https://www.gnu.org/licenses/old-licenses/lgpl-2.1.html
Microsoft Public License
------------------------
http://www.opensource.org/licenses/ms-pl.html