David Mitchell escreveu: 
Most of the things that make Smalltalk great (what makes Smalltalk
Smalltalk) are the things that hold it back for a lot of people.

If you want a more Unixy, scripty, Smalltalkish thing with syntax
blended C and Perl that you can hack with a text editor, try Ruby.

If you want objects all the time with a crazy amount of integration in
the tools and little attempt at conforming to outside ideas, Smalltalk
is your game.
Some things that hold back smalltalk:
  1. It was not supported by a media boosted corporation the way Java was in the 90ties. Today everybody knows that Java sucks. Everybody knows that with current technology (JVM) it is not possible to build large intensive Java applications (just study how the  garbage collection works... what happens when you have to handle large data...). But Java got a good marketing while smalltalk and Self and other really interesting things were related as "geek things".
  2. Smalltalk community was not able to create a really functional organization capable of delivering important things as standards and documentation.
  3. Not taking into account the commercial versions of smalltalk that are supposed to satisfy small closed markets, the open versions like squeak never succeeded in releasing and maintaining "distributions" (like Linux) where all bundled parts are more or less assured to work properly with each other.
Much of the heat recently generated around smalltalk results from people perceptions about the dead-ends imposed by other programming models (like Java). With a little lucky it will attract investment and supporting infrastructure (money to pay for people doing the hard works of documenting things, establishing certain standards, testing & debugging, etc).