DelphiDoom 2.0

[Jimmy Valavanis]

Cool.

I planned to do the same with Duke Nukem 3D. I even have code that loads maps and renders in 2D. May be some day . If I re-start the project, can I ask for your help?

Yes, of cource. Are you planning to recreate the game in a new basis (using only the grp data file) or converting the build engine source from C to Pascal?
By the way, I had made a 3d map viewer for Duke3D a long ago (http://www.geocities.ws/jimmyvalavan...ukeviewer.html) .

[Ñuño Martínez]

Yes, of cource. Are you planning to recreate the game in a new basis (using only the grp data file) or converting the build engine source from C to Pascal?
By the way, I had made a 3d map viewer for Duke3D a long ago (http://www.geocities.ws/jimmyvalavan...ukeviewer.html) .

Just recreating the game in a new basis. Initially I tried to translate from C to Pascal but it is impossible, an I really want to know how does it work (actually Ken forgot it! I asked him!).

Do you still have the sources of the Duke Viewer?

[Jimmy Valavanis]

Just recreating the game in a new basis. Initially I tried to translate from C to Pascal but it is impossible, an I really want to know how does it work (actually Ken forgot it! I asked him!).

Do you still have the sources of the Duke Viewer?

Same here. Some time ago tried to play around with build engine source but it was difficult to understand, so never tried to convert it to Pascal. I think I can resurrect some portions of Duke Viewer source code, at least the main loader functions, I'll let you known.

[Ñuño Martínez]

I have no problems with loading. I have a program that loads maps and renders in a "build editor". Problem is rendering. I found a paper somebody wrote that analysed how the algorithm worked and I planed to use that information to re-build the algorithm. Problem is that the algorithm works with 2.5D but not with actual 3D (or not 6DOF).

Anyway I have no time now. Too much projects.

[Jimmy Valavanis]

I have no problems with loading. I have a program that loads maps and renders in a "build editor". Problem is rendering. I found a paper somebody wrote that analysed how the algorithm worked and I planed to use that information to re-build the algorithm. Problem is that the algorithm works with 2.5D but not with actual 3D (or not 6DOF).

Anyway I have no time now. Too much projects.

This is the source code of DukeViewer: https://www.dropbox.com/s/lsfexp3gus...r_src.zip?dl=0

The source code is propably incomplete, but the main Loader functions, as well as the base of the rendering engine are there.
As far as I can remember, it uses some custom components (not critical). The point of interest is function GetDuke3DMapData() at Duke3DUtils.pas. This function reads a map and feeds the 3d Engine with data (polygons). The texture reading is at TD3DScene.CacheDuke3DTextures() function inside surfaces.pas.

[Ñuño Martínez]

Thanks. I'll take a look.

[Jimmy Valavanis]

A new meintainance release (ver 2.0.3.705) is available.

Most notable new features/fixes since the base 2.0 release are:

• PascalScript can now access DLLs from inside pk3(zip) files. (or WAD files if using PK3ENTRY lump). Download an example here.
• Smooth freelooking. Now the engine does not ignore slight variations in mouse movement
• Added gl_no_glfinish_hack console variable to avoid with OpenGL slowdown on systems with Intel HD4000 and Windows 10 (default is True). More information can be found at intel communities forum.
• Support for PCX format external textures.
• Fixed alpha channel in TGA screenshots.

Download link at sourceforge