[
Lists Home |
Date Index |
Thread Index
]
Rick Jelliffe writes:
> Its clearly a completely bogus principle! In fact, for markup
> languages the reverse is true: having more forms makes data capture
> and modeling easier because you can choose the form that requires
> the least work. E.g. (<![CDATA[ ]]> or & ) and (element or
> attribute) and (<x></x> or <x/>) and ( y="z" or y='z').
Flexibility is a benefit for users but a cost for implementors;
simplicity is a benefit for implementors but a cost for users. Sadly,
there's no magic formula that works every time for programming
languages or markup languages.
> The other bogus principle is that there should only be one syntax
> for everything. Looking at the triumph of the C family over the
> LISP family, it is more likely that people prefer a variety of
> embedded syntaxes which serve to indicate semantics or role
> graphically.
Again, it's a trade-off: you have to consider both internal and
external factors. Using a syntax that's 10% wrong but 50% better
supported might make sense; using a syntax that's 50% wrong but only
10% better supported might not. The interesting part is that once
something gains a bit of momentum, support tends to snowball because
of economies of scale, so that it makes sense to use it places where
it really is suboptimal in and of itself. Think of the people who
have switched their development work from FORTRAN to C/C++ or Java.
All the best,
David
--
David Megginson, david@megginson.com, http://www.megginson.com/
|
