News from 9 November 1999

[randi]

Perhaps I need to be more specific about that last update. Some people seem to be reading too much between the lines.

I never said ZDoom was dead. I said I had begun work on a new engine that shares features with Build, and I offered a glimpse of it after one week of work. This engine is like Build in the same sense that Build is like the Doom engine, yet Build is not the Doom engine. That is, I plan to implement features found in Build that are difficult to do with the Doom engine (slopes, wall sprites, moving sectors, stacked sectors, etc).

I will also be rewriting other aspects of the engine besides the renderer from scratch. This new engine will be written with scripting in mind from the very beginning. Implementing DoomScript into the existing Doom engine would have amounted to an even more monumental task, since I would have had to gut parts of the engine and then figure out how to glue it back together. The engine will also be developed with client/server in mind. In many ways, changing my focus from trying to retain the Doom engine, but glue DoomScript and other features on top of it, was a matter of convenience.

Doom support is not disappearing, and I am not making a Duke Nukem 3D clone or a clone of any other Build-based game. If anything, I am making a Doom clone. Removing Doom support would have defeated the purpose of creating the new engine, as I doubt very many people would be interested in another 2.5D engine if it was necessary to create entirely new content to use it.

I will, in fact, keep the old Doom renderer around for compatibility with maps that use special effects that take advantage of the way the Doom engine does things. When I said that "all active development will occur on the new [engine]," what I meant was that I would not add some feature to the new engine and then try to splice that back into the Doom engine.

This new engine will require a new editor, because the internal data structures are different, and Doom doesn't leave enough room in its map format to support much in the way of new features. (This is why many of ZDoom's features have to be activated using ACS.) The major advantage of a new editor will be support for new actors added using DoomScript that have their own custom properties. With Doom, thing attributes are limited to a few bits that can be set with the editor. Hexen extended this to five one-byte values, but at the cost of no death triggers for things that used them. With DoomScript, an author will be able to mark any attribute of an actor as customizable with the editor.

A conversion utility will be available, so people can still build all their geometry in a separate editor and then import it later. This is no different from what zwadconv already does.

The new editor will not be ZETH or based on ZETH. DETH was apparantly not written with expandability in mind, and trying to change it would be more trouble than writing a completely new editor. ZETH also doesn't work under NT, support for which is one of ZDoom's strong points.