[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Seeds of the "movement" [was: peasant revolt]



>       You have done a wonderful job, I am very grateful. Of course you
>       teach useful things. But you alone won't save the world.
> 
> No I won't. We all need each others help. We need to partition the labor.
> Some of us are good at writing Schemes. Thanks Matthew. Others write
> Programming Environment. Thanks Robby. Some write books that 
> bridge the gap to Java. Thanks Max. A lot of you embed MzScheme into C 
> programs and do real engineering, which shows the world where the line
> is to be drawn. Thanks. 
> 
> It's probably impossible to make up a strategic plan but if a 
> "movement" can emerge, let's go for it. 
> 

Absolutely!

I have probably posted something like this before, but I think it bears
repeating.

I think we need (at least) the following two things to happen:

(1)  Agreement amongst the various Schemes as to a "standard library":

     I.e., all Scheme's must roughly meet the R[4/5]RS specifications.
     But that's really not quite enough to do real-world work.  We need
     efficient, ready-made libraries to handle XML, sockets, GUIs, etc.

     Of course, we all know that MzScheme posesses these tools already.
     But Guile is only partially there.  And I can't speak to the Chez
     or Stalins of the world.

     Perhaps we need a "Common Scheme"?

(2)  Easy distribution of modules:

     One of the really wonderful things about Perl is the presence of the
     CPAN (Perl Archive).  While MzScheme has a pretty good system for
adding
     new modules (via the *.plt files), there perhaps needs to be a public
     archive for storing user-contributed versions in a searchable fashion.

Once those two pieces are in-place network effects begin to take hold,
which will tend to amplify small contributions, as the whole becomes greater
than the sum of the parts.

-Brent