Various things...

Raphael.Quinet@eed.ericsson.se
Tue, 5 Mar 1996 09:41:28 +0100 (MET)

Date: Tue, 5 Mar 1996 09:41:28 +0100 (MET)
Message-Id: <199603050841.JAA24477@chapelle>
To: quake-editing@nvg.unit.no
Subject: Various things...
From: Raphael.Quinet@eed.ericsson.se

OK, I think I have to clarify a few things, and I will try to reply to
all questions in this message... Please bear with me, this message is
going to be a bit long.

Chris Cason <Chris.Cason@oaks.com.au> wrote:
> I was under the impression this time that id would be releasing some specs for
> the data structures. Can anyone confirm/deny this ? We (SciTech) have been
> working with id a little on a technical level (regarding video) but haven't
> really had a opportunity to raise this (and probably won't, either, since
> they're flat out.)

Apparently, even the guys at id Software don't know if they will
release their specs or not. :-) Dave Taylor wrote that they would
probably release some information, but he wasn't sure. Anyway, it
doesn't really matter, because most of the internal structures are
known by now. I'm sure that the Unofficial Quake Specs will be as
good as any official document distributed by id.

Also (warning: personal ramblings ahaead), I'm afraid that id Software
could try to do the same thing as they (unsuccessfully) did for Doom:
releasing a simple editor, BSP compiler or QuakeC compiler, but with a
license that forbids certain things and prevents people from
distributing their levels as they wish. Those who were around when
John Romero released the source code for IDBSP at the same time as Jay
Wilbur sent the Data Utility License to all authors of Doom editors
probably know what I am talking about. I think they did that in good
faith and were simply trying to protect themselves, but the legal
restrictions for the Doom editors were a bit too strong, IMHO. The
information in the unofficial specs will be free for everyone.

Brian K. Martin <brian@phyast.pitt.edu> wrote:
> Sorry, but I was under the impression that this list was formed
> to discuss file formats and to help get editors made. If there are
> some uncertainties in the quake specs, why not ask this list for
> help?

Most of the uncertainties are not in the data formats, but in the text
that explains them. I understand Olivier who doesn't want to release
the specs with poor explanations, because this list would quickly be
flooded with questions. You are right about the purpose of this list,
and all the information that was gathered in the unofficial specs will
soon be released so that we can all benefit from it and concentrate on
writing our editors or understanding the last unknown bits in the
structures. But the document is not ready yet and the bits and pieces
have to be put together in order to have a coherent structure. A
document which contains all the information but not the explanations
that go along with the data formats is likely to generate a lot of
noise. Please be patient, the specs will be released sooner than you
think (well, I hope so).

Christopher Michael Holland <chris@progsoc.uts.edu.au> wrote:
> Does anyone know if the engine itself does the anti-aliasing, or if there
> are actually three versions of each texture in the pak?

This technique is called mip-mapping and was described in a paper by
Michael Abrash. Now, don't you wonder why he was hired by id
Software? The images of each tecture at various scaling factors are
stored in the BSP files (which are in turn included in the PAK files),
and the anti-aliasing was pre-computed by a separate program.

And now a personal disclaimer... I wrote:
> [...] A new version
> of the soon-to-be-world-famous Unofficial Quake Specs will be out soon
> [...]
... and Olivier wrote:
>
> The Unofficial Quake Specs version 3.0 (by Raphael Quinet and me)
> are currently being finalised. [...]

Despite what Olivier wrote, he did 99% of the job and I only helped
him for some details. The comment about the "soon-to-be-world-famous"
specs applies Olivier's work. :-)

-Raphael