[comp.lang.smalltalk] Product'izing Smalltalk Applications

san@ut-sally.UUCP (Steve Sanderson) (11/20/86)

(As I open up Pandoras box...)
What are the opinions on Smalltalk as an environment for commercial
applications (e.g. the mundane spreadsheet)?  The reason I ask is two-fold:
As a software-type, I really appreciate all that Smalltalk supplies as
a programming environment, and can see that it could make development of
applications (or at least prototyping) easier than many other environments.
The other reason could (perhaps) be better re-phrased as "Can any interpretive
environment succeed as a base for commercial applications?"  When thinking
interpreted, the first thing that springs to mind is BASIC.  I seem to recall
applications built entirely in BASIC, though it appears they may have all
together disappeared.  Obviously a prime consideration is speed, which brings
up the important question:

	"How possible is it to compile Smalltalk to native code?"

Another reason Smalltalk may not be suitable for commercial applications is
due (ironically) to one of its (many) advantages, the ability to make a copy
of the running code and in addition modify the code.  The modification aspect
could be a customer support nightmare!  Yes, this article has a definite
non-theoretical slant, but frankly if I were to develop applications I could
do it better and more enjoyably in Smalltalk, causing the word of Smalltalk to
be spread far-and-wide, and as a nice benefit we may end up with that elusive
goal... Portability!  (or course compiling it to native code may hinder that
a little... 8-)  As another tangent , integration with future products would
seem to be easier, as long as they provided objects that spoke the proper
protocol and reasonable import/export procedures (aka Inter Process Comm.)
were provided ahead of time at some level.

	Thanks,

	Steve Sanderson
-- 
Steve Sanderson,
Hosehead at Large,
{ihnp4,seismo,ctvax}!ut-sally!san, san@ut-sally.UUCP