Re: Time Question..


30 Jun 1995 20:46:46 GMT

In article <3t1jhb$64t@agate.berkeley.edu>,
Gerry Kevin Wilson <whizzard@uclink.berkeley.edu> wrote:
>In article <4jx1Ase00WB94cR_kG@andrew.cmu.edu>,
>Andrew C. Plotkin <erkyrath+@CMU.EDU> wrote:
>>gt5048a@prism.gatech.edu (Jon A. Preston) writes:
>>> In the "old days", how long did it take to produce an if-game?
>>>
>>> And today? How long using inform or the likes?

>Heheh, couldn't resist. I remember that when I started writing Avalon,
>way back in Oct. '93, I figured it would take me two months to write.
>Looks to be closer to two years, at this point.

"Dunjin" took me six years... (of course, there were long stretches during
that time when I didn't do anything to the game for several months).

>TADS and Inform
>definitely save a lot of time

Indeed. Just the task of writing a parser/disambiguator (you can't
really separate the two, since without good disambiguation a good
parser is almost worthless) as powerful as the ones of TADS and Inform
is huge; I'd venture a guess of about a man-month for an experienced
engineer to design, write and debug it using C or Pascal. And then
there's the task of providing all the functionality present in, say,
the TADS library (all the standard classes for containers, actors, and
so on). Again, a matter of man-months.

And bear in mind that most IF authors are not experts in these areas of
programming (although they probably will be after finishing such a
project :-)).

> but I'll bet that most folks offset that
>gain by writing more complex games.

It's certainly true for me.

With TADS, you can write a "simple" game fairly quickly. Then you
spend twice as much time again adding detail, polishin your prose,
adding funny responses to unlikely actions, etc.

But, to give you some idea of how much time is needed:

Suppose you're working at about the level of detail of a typical Infocom
game. Then a game for Whizzard's competition, i.e. a game that will
be possible to finish in under two hours, will probably take you
several weeks to write (supposing you do it in your spare time). Then
you should add _at least_ as much time for playtesting, debugging, and
adding all the features suggested by the play testers.

For a game of the size of, say, "Wishbringer", you shouldn't count
on being finished with it in less than half a year, even if you
devote most of your spare time to it.

But, of course, your mileage may vary. They say that the difference
in productivity between an average programmer and a brilliant programmer
is on the order of 1000..

Magnus Olsson (mol@df.lth.se) / yacc computer club, Lund, Sweden
Work: Innovativ Vision AB, Linkoping (magnus.olsson@ivab.se)
Old adresses (may still work): magnus@thep.lu.se, thepmo@selund.bitnet
PGP key available via finger (to df.lth.se) or on request.