>Well, I bitched about this before, but it doesn't necessarily depend on the
>game or the market: it may just be that the interpreter sucks on many
>platforms.
If the interpreter sucks, then the game usually isn't very fun to play.
Unfortunately.
>In the case of TADS, I seem to recall being limited by a
>less-than-wonderful parser as well.
You "seem to recall"? Please substantiate!
IMO, TADS has the best parser available right now. Inform's parser is
potentially better, since the user can extend it in any way he likes, but
the one included in the Inform Library is (IMO, of course) slightly less
good than that of TADS.
Of course, there are bad TADS games. What you perceived as a bad
parser could be a sloppily programmed game. If the author provides too
few synonyms, or chooses his vocabulary in a suboptimal way, or
doesn't think through things like ambiguities sufficiently, or (in the
case of TADS) fails to provide the needed verification methods, no
parser in the world can make up for the deficiencies in the game.
Both the TADS parser and the Inform parser are extremely good (better
than Infocom's) *provided* the programmer of the game does a good job. What is
a good job? Well, it depends on what you're trying to do. For a plain vanilla
game where every object has a unique name, and things like that, the
programmer merely has to provide the objets and the vocabulary. In other
cases, the programmer has to make sure his code will work together with
the parser. This is true for *all* parsers; at least until somebody
invents a DoWhatIMean command.
Magnus