Doomsday Engine

2010-08-11

23:36 Revision 2eb88ab6 (github): Continued work on the new stencil-based skymask algorithm.

danij

20:59 Revision b5cd9aa9 (github): Fixed DOOM: Black water(?) in Plutonia2::MAP25 would not animate.

danij

14:21 Revision f4a37b5d (github): Missing new-line in unknown texture warning logged by GL_CheckTextureNumForName.

danij

14:13 Revision 69e43cc7 (github): Further enhancements to Rend_SSectSkyFixes to fix a couple of constructs seen in Plutonia2. The hunt continues...

danij

11:35 Revision 099d4719 (github): Missed a couple of necessary exceptions to the special cases added in Rend_SSectSkyFixes in the previous commit.

danij

02:05 Revision ef84138f (github): Added another special case to Rend_SSectSkyFixes to deal with the glitches seen on the midtexture columns in Plutonia2::MAP29

This routine is now a complete mess but I'm tired and this works; so I'm commiting. I'll refactor it later. danij

2010-08-10

22:28 Revision 0e009bb6 (github): Fixed handful of uninitialized assignments shown up by static analysis. Thanks for the report yagisan.

All but two of these were innoxious however as they are not read after the fact. The rest of the issues are false pos... danij

13:44 Revision bcc4bfb2 (github): Fixed: Animated blood waterfall textures in Plutonia 2.

Addressed the assumption that Materials in the MN_TEXTURES namespace were created in the same order as their primary ... danij

09:04 Revision 9fa2e389 (github): Fix inversed skyfix height checks in Rend_SSectSkyFixes.

danij

2010-08-09

20:14 Revision 3e30436c (github): The master is now open for Beta 7 development.

skyjake

16:34 Revision fc3c84de (github): I finally got fed up with the fps counter being drawn behind the console.

danij

16:33 Revision 811b3a27 (github): Optimize: If both left and right lightlevels are the same when drawing a wall, use the vector version of lightVertex so as to potentially benefit from SSE.

danij

10:50 Revision 6fc2d00a (github): Do render planes for unclosed sectors. Now our skymask strategy has changed this is precisely what we don't want to do (we need complete "sky volumes" without holes).

danij

10:46 Revision 5ac4fc33 (github): Addressed performance problem with MaterialBind for Material lookup.

danij

03:18 Revision 5fb2cd58 (github): Fix fi_obtype_e type checking in ui2_main.c due to silly copy paste oversights.

danij

00:27 Revision a67c09c6 (github): Fix typo in previous commit.

danij

2010-08-08

17:23 Revision c3031140 (github): Updated to compile with gcc 4.2 and the Mac OS 10.6 SDK.

skyjake

17:18 Revision 1eefd18d (github): Yet further finessing the handling of skymasked surfaces when the viewPlayer is in the void. It now appears to be working perfectly, although its now looking a bit of a mess. This needs rewriting.

danij

12:52 Revision 6441b83b (github): Further finessed handling of skymasked surfaces when the viewPlayer is in the void, so that the world is now rendered more pleasingly.

danij

06:17 Revision 5041eec7 (github): Fix line endings.

danij

06:11 Revision eddc01f3 (github): Changed: Reimplemented the skymask using a stencil.

Changed: Removed cvar "rend-sky-full" as it is no longer useful.
As yet untested on platforms which use SDL for wind... danij

2010-08-07

16:29 Revision d1580f35 (github): Updated style.

danij

02:48 Revision abbe9f09 (github): Updated style.

danij

02:19 Revision 60c185b0 (github): Updated style.

danij

01:06 Revision 3d9962d4 (github): Continued work on merging the FIPage stuff with the existing UI subsystem.

danij

2010-08-06

22:43 Revision a8d9e3d0 (github): Completed seperation of the FIPage stuff out of the InFine subsystem. I'm going to now begin transforming the object model, to parallel the features of the existing UI subsystem.

danij

20:41 Revision 462d916f (github): Added a drawer function ptr to FIPage.

danij

19:37 Revision e8f2b71f (github): Initial steps in converging the game menu implementation into the engine-side FIPage stuff.

danij

10:14 Revision 032af5b6 (github): Fixed compile issue with gcc becoming confused about the two FI_Register functions due to the shared symbol namespace (can't they address this??).

danij

09:46 Revision 7df633ff (github): Fix compile error with gcc due to mismatched function declaration/definition.

danij

09:28 Revision bae06b4a (github): Do not give up loading a flat graphic if the lump its from is smaller than 4096 bytes. Instead always create an image of at least this size and fill the gaps with black. Fixes the thrashing problem with the sky flat in Heretic.

danij

2010-08-05

18:56 Revision 12f2b628 (github): Split the new FIPage/FIObject stuff out of engine/api/dd_infine.h and into engine/api/dd_ui.h in preparation for the next phase. Now the fun starts :)

danij

17:36 Revision a9ee10f8 (github): Moved the InFine script stack functionality into common/fi_lib as a wrapper interface. Clean up.

Todo: Reimplement the handling the GPT_FINALE2 packets to pass the condition values needed client-side.
Todo: Refacto... danij

2010-08-04

00:23 Revision 8a950b72 (github): Fix off-by-one in M_InitControlsMenu resulting in zone corruption.

danij

2010-08-03

13:45 Revision 2c4ade0d (github): Fixed non-working detail textures and shiny surface reflections.

danij

2010-08-01

15:35 Revision bac0b49b (github): While a FinaleInterpreter is suspended, pause any FIPages created by it.

Todo: Further refactor this so that objects on multiple (simultaneously visible) pages do not tic more than once in a... danij

13:32 Revision 8974cdf6 (github): Removed extraneous semicolon in XSTrav_MimicSector.

danij

13:23 Revision 9a693687 (github): Fixed numerous warnings output by gcc. Nothing of interest here; all inconsequential pointer pedantics or unused return values.

danij

2010-07-31

21:28 Bug #928: [Hexen] Crash on autosave returning to seven portals

First of all we need to rule out the resource pack. Please disable it and re-run the experiment. Does it work? danij

21:28 Bug #928 (Closed): [Hexen] Crash on autosave returning to seven portals

In the second guardian of steel, after getting the message \"stairs have risen on the seven portals\", when I try to ... kanraku

2010-07-30

23:37 Revision e9b9c29b (github): Fix line endings.

danij

23:18 Revision 582a896a (github): Addressed draw order restrictions on objects in FIPages. They are now drawn in the order in which they are created, regardless of type. FinaleInterpreter will now allocate two pages:

* PAGE_PICS - used for Pics and its background.
* PAGE_TEXT - used for Pics and its filter. danij

18:42 Revision 58a5e133 (github): Began seperation of the finale stack behavior from the InFine system itself.

danij

17:55 Revision 46985849 (github): Added the Animator entry points to doomsday.def

danij

2010-07-29

17:44 Revision bd9f9ac4 (github): Further style and naming convention updates to the InFine system.

danij

12:11 Revision 2400e2f5 (github): Update style and general clean up of the InFine system.

danij

06:12 Revision 3b06917b (github): Fixed FinaleInterpreter stack popping problem which resulted in the game not returning to the expected state if a "local" script is above an "overlay" at the time.

Fixed FIPage background color and alpha not set when applying a Material for the first time by a FinaleInterpreter.
C... danij

02:19 Revision 52905251 (github): Maintain a global vector of allocated FIPage objects so we can be sure they are released when no longer used.

FIPages are now always drawn unless hidden. Previously only the page attributed to the top-most FinaleInterpreter in ... danij