The "Which Authoring System is Better?" FAQ; TADS Update
N. K. Guy, October 15, 1998
[email protected]
_______________
In 1996 Bob Newell released the last version of his comprehensive
"Which Authoring System is Better" FAQ, which compares the most
popular development systems available for writing IF. It's a great
piece of work, but hasn't been updated in the past two years and thus
doesn't incorporate some of the most recent changes.
I thought I'd write this little addendum concerning TADS, as that's
the development system with which I'm most familiar. I'm not as
knowledgeable about Inform or Hugo, so I haven't included much
commentary on those development systems. Hopefully devotees of those
systems can also offer updates on the latest news.
5.1
- TADS is now up to 2.2.6.
- Mike Roberts can be reached at mjr_ @ hotmail.com. (remove spaces,
of course)
- TADS is no longer published as shareware. It is freeware software -
ie, it continues to be a copyrighted work by Michael J. Roberts, but
you do not have to pay a fee to use it. It is not in the public
domain. Mike continues to release and update new versions of TADS
himself, and he has dissolved High Energy Software, the company
he used to publish the shareware version.
- The most important recent change on the TADS scene was the April
1998 release of HTML TADS. This is a new and completely revised
version of the TADS runtime system. It understands HTML, the
HyperText Markup Language used to create Web pages. With HTML TADS
it's possible to write TADS games that include graphics, sound,
complex text formatting, multiple status lines and so on. In addition,
an HTML TADS game can be written such that it's fully compatible
with text-only runtimes as well. (so you aren't losing out a big
share of the audience if you choose to develop a graphical TADS game)
This is quite a step and there's currently nothing quite like it
on any other IF development platform. Hugo supports graphics and
sound and output formatting, but does not use HTML, which is a
very popular and easy to use standard. It is theoretically
possible to develop Z-machine z6 games, but the tools are not yet
complete, nor are z6-compatible interpreters widely available.
Unfortunately, graphical TADS interpreters are only currently
available for Windows 95/98/NT and Macintosh, though source code
is available for people interested in porting the code over to
other platforms.
There's more information at:
http://www.tela.bc.ca/tela/html-tads/
1:
- The manual is now available online in HTML format. The URL is:
http://www.tela.bc.ca/tela/tads-manual/
The various release notes have recently been consolidated and are
no longer separate pieces, thus addressing Bob's very valid
concern over "disjointed and scattered" documentation. Also,
since it's readily viewable with any Web browser, it's possible
to view both code and manual on-screen simultaneously. And
naturally you can still print out a hard copy if you like.
3:
- The source code to TADS is now available. You can download the C
source and examine how it works. However, Mike asks that you don't go
and release modified versions without his permission - he made the
source available in order to ease porting, not so that people can
create derivative works. Thus the TADS parser is not, at present,
as readily modifiable as the Inform parser.
4:
- As noted above, High Energy Software is no longer a going concern.
Mike Roberts is maintaining TADS as a person, not a company. TADS is
no longer shareware and is available as a free download in its
entirety from ftp.gmd.de. Also, High Energy's bulletin board system is
no longer in existence.
6:
- Portability is still an issue with TADS. TADS games are playable on
the majority of desktop systems out there, but Inform definitely
has an edge, particularly in the palmtop category. Inform games
are playable on many small portable palmtop systems, but TADS
games aren't, because the interpreter is bigger. In addition,
graphical HTML TADS interpreters are only available for Windows
32 and Macintosh.
7:
- Regarding speed, I think Bob may have been misled somewhat by the
WorldClass problem. Large games compiled using WorldClass libraries are
indeed very slow on older machines. This is related to the way
WorldClass handles its scope issues. However, very large games
compiled with the standard adv.t libraries do not suffer from this
problem. (ie: it's a WorldClass problem, not a TADS one) The TADS
runtime may be slightly slower for some things than the Z-machine
used by Inform, but I don't think it's a serious issue.
9:
- The debugger is now freely available off ftp.gmd.de and thus
distribution is no longer a problem.
- HTML TADS ships with a new debugger. Currently it's Win32 only, but
is a big advance from the earlier TADS debugger for MS-DOS. It's
similar to the Macintosh windowed debugger, but with a number of extra
handy features. Hugo ships with a similarly powerful debugger, but at
present there are no Inform debuggers available.
10:
- Mike Roberts did indeed go on hiatus for a couple years following
the release of TADS 2.2. However he's been very busy the past
year developing HTML TADS. TADS is far from dead. Of course, as
TADS is entirely his baby the future of the system is entirely
dependent on his own interests. The Z-machine, by contrast, is
not owned by one person. (though it's also true that Mike can
make rapid improvements to TADS whenever he wants to, whereas
changes to the Z-machine require the cooperation of an informal
cabal. Cuts both ways.)
12:
- Very large games are indeed unplayable on very old MS-DOS
machines, because of the 640K limit those machines are stuck
with. However, TADS games can be considerably larger than
Z-machine games - the limit is much higher. Also, although it's
true that huge games can't be played on old 286 computers, I
think that's becoming less and less of a problem as time goes by.
- Note that games compiled with Inform tend to be much more
compact than those compiled with TADS. It's obviously hard to
compare the two, but various experiments by raif regulars seems
to suggest that this is the case. The fact that TADS is 8-bit
clean is probably a significant factor, as 8-bit bytes take up
more room than the Z-machine's 5-bit Z-characters. Also, TADS
games do not use text compression whereas Z-machine games do.
14:
- Just as another data point, TADS was my first experience with a real
programming language beyond BASIC back in the 1980s, and I didn't find
it that difficult to get used to. I personally don't think it's really
necessary to come to TADS with a sound knowledge of Pascal or C,
though I'm sure it helps. But that's my take; your kilometreage may
vary.
15:
- As noted above, TADS is now completely free in monetary terms. You
don't have to pay a shareware fee to use it.
16:
- As noted above, the TADS source is now available for download.
However, it's not public-domain and remains Mike Roberts' copyrighted
property.
