hedgewars/uFloat.pas
Wed, 03 Apr 2013 23:56:59 +0200 koda strip PAS2C, old WEB symbols and outdated pas2c sources from default branch, all c-related development is done on the webgl branch
Sun, 11 Nov 2012 17:15:19 +0100 koda merge xymeng's gsoc engine with a few updates (and further checks on symbol definitions) webgl
Fri, 28 Sep 2012 00:39:20 +0400 unc0rr Shutup pas2c
Fri, 28 Sep 2012 00:18:07 +0400 unc0rr vector2Angle function which converts vector to angle from -2048 to 2048
Sun, 16 Sep 2012 22:55:33 +0400 unc0rr New hwSqrt which is up to 8.7% more efficient than old one on synthetic tests
Tue, 04 Sep 2012 11:09:23 +0400 unc0rr More precision
Mon, 03 Sep 2012 12:35:51 -0400 nemo Let's try *4 instead. Not nearly as much of a win, but might be usable.
Fri, 31 Aug 2012 11:10:21 -0400 nemo oops
Wed, 29 Aug 2012 10:56:32 -0400 nemo A few more operators. Committing in case unc0rr wants to fiddle.
Wed, 29 Aug 2012 09:58:23 -0400 nemo isZero appears to be never used. Use it in a few obvious cases and add web variant.
Mon, 27 Aug 2012 16:52:50 -0400 nemo Starting to rewrite uFloat routines to avoid QWordValue - in the interests of JS which may get uint64/int64 in the spec in the future but does not have it now.
Fri, 24 Aug 2012 18:51:29 -0400 nemo fromdos + fix end of turn velocity
Fri, 24 Aug 2012 14:51:12 +0400 unc0rr An attempt to run main rope code only every 8th tick:
Tue, 07 Aug 2012 13:21:15 -0400 nemo oops
Fri, 11 May 2012 00:28:06 +0400 unc0rr Some work to make more units compile after conversion to c
Thu, 10 May 2012 23:51:05 +0400 unc0rr Support overloaded operators on (hwFloat op hwFloat) calls
Wed, 11 Apr 2012 22:46:59 +0400 unc0rr Simplify converter's life
Wed, 14 Mar 2012 21:12:04 -0400 nemo Allow RC plane to go through portals
Mon, 12 Mar 2012 01:00:17 -0400 nemo Change rest of AI tests to floats, reenable snowball
Fri, 17 Feb 2012 18:23:36 +0100 koda the most important commit of the year
Tue, 17 Jan 2012 09:01:31 -0500 lovelacer A partial reformatting of the pascal code to have consistent syntax. Things that are still inconsistent.
Fri, 02 Dec 2011 15:03:27 +0400 unc0rr Some tiny optimizations
Sat, 26 Nov 2011 21:12:23 +0300 unc0rr - Help parser by dropping that insane formatting syntax in str function
Thu, 24 Nov 2011 16:18:45 +0100 koda GSoC 2011: Android port - merged mainstream
Sat, 12 Nov 2011 17:06:49 +0100 Xeli merge...i think hedgeroid
Wed, 09 Nov 2011 18:40:12 +0300 unc0rr Less precision - less confusion
Sat, 05 Nov 2011 20:51:15 -0400 nemo Due to the small values and friction, halve the step in low gravity, instead of the value
Fri, 16 Sep 2011 18:17:16 +0200 Xeli merge hedgeroid
Sat, 10 Sep 2011 21:02:34 -0400 nemo For better or for worse. Add sheepluva's forcing of clean on engine subdirectory. This works around a bug in fpc 2.4-2.6 that appears to generate incorrect inlining of loops on 2nd build. -B avoids this, but without a clean, the -B never got called. Also restore options, which I had confused w/ config. my bad.
Fri, 09 Sep 2011 18:30:40 -0400 nemo Remove some unused variables and options.inc which uFloat doesn't use, probably should never use, and was getting in the way of my testcase - but most importantly, remove the inline on hwSqrt which was causing very bad math on my compiler/machine. We may have to remove more inlining. A pity.
less more (0) -50 -30 tip