[e-lang] early useful features considered harmful ? (was: Proposed grammar simplifications)

zooko at zooko.com zooko at zooko.com
Wed Jun 1 09:19:54 EDT 2005

 Chip wrote:

> Just to put in my 2 cents, I think the "verb=" syntax should just die.
> It's ugly, it's confusing, and it adds to the language's whole aura of
> weirdness without giving us much payoff in return for the trouble.  I
> don't think we particularly need to replace it with any other special
> syntactic support.

Python achieved popularity without having "x += y".  It later added it in 
Python 2.0.  Having used Python extensively before and after this change, 
I personally find the addition of "+=" and "-=" to be mildly satisfying, but 
hardly important, and I've never missed the absence of generic "verb=".

One lesson that E might consider to take from this is that all of the "Big
Success" languages gained their first widespread success *before* they had 
syntax as featureful as they have today.

There is a hypothesis that languages need to start with fewer features in order
to get early adoption, and only then to add more features.  This is opposite to
the idea that "If a feature X is proven to be useful in all current Big Success
languages, then feature X ought to be included in the current version of E.".

Instead, consider the perhaps surprising hypothesis that if Java 1.0 had come
with all of the features of Java 1.5, that it would not have been as widely
adopted, and that if Python 1.4 had started with all of the features of Python
2.4, it would not have been as widely adopted.

Personally I consider E 0.8 to be rather complicated and difficult for me to
get into.



More information about the e-lang mailing list