Path: utzoo!utgpu!watserv1!watmath!att!att!emory!wuarchive!julius.cs.uiuc.edu!ux1.cso.uiuc.edu!ux1.cso.uiuc.edu!m.cs.uiuc.edu!johnson
From: johnson@m.cs.uiuc.edu
Newsgroups: comp.object
Subject: Re: Do we really need types in OOPL
Message-ID: <77500059@m.cs.uiuc.edu>
Date: 29 Oct 90 12:34:00 GMT
References: <2444@runxtsa.runx.oz.au>
Lines: 20
Nf-ID: #R:runxtsa.runx.oz.au:2444:m.cs.uiuc.edu:77500059:000:967
Nf-From: m.cs.uiuc.edu!johnson    Oct 29 06:34:00 1990


Andy Ormsby (aro@cs.aber.ac.uk) wrote:

>I've often seen complaints about the performance of Smalltalk. To what
>extent is performance really an obstacle to the adoption of Smalltalk
>these days?  My (very limited) experience is that Smalltalk seems to
>produce much more responsive applications than C++/Interviews. 

As with anything else, it depends on the application.  For user-interface
intensive applications, Smalltalk is plenty fast enough, and is probably
just as fast as C++/Interviews.  However, if you are doing a lot of
computation, whether floating point or symbolic, then Smalltalk can be
an order of magnitude slower than C/FORTRAN/etc.

Modern Smalltalks make it possible to call C routines from Smalltalk.
A common trick is to write the computation entensive part in C and
call it from Smalltalk.  This works well when the expensive part is
fairly well contained, which is quite common.

Ralph Johnson - University of Illinois at Urbana-Champaign