Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2.fluke 9/24/84; site vax2.fluke.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!ihnp4!houxm!vax135!cornell!uw-beaver!fluke!dan
From: dan@fluke.UUCP (Dan Everhart)
Newsgroups: net.lang.st80
Subject: Re: Stupid Questions about Berkeley SmallTalk
Message-ID: <432@vax2.fluke.UUCP>
Date: Thu, 7-Mar-85 15:37:16 EST
Article-I.D.: vax2.432
Posted: Thu Mar  7 15:37:16 1985
Date-Received: Sat, 9-Mar-85 06:21:21 EST
References: <419@harvard.ARPA>
Distribution: net
Organization: John Fluke Mfg. Co., Inc., Everett, WA
Lines: 30

I have not (yet) used BS for a real application, but I will make
observations based on the few toys I have built with it.

> 
> 	1) How much slower is Berkely SmallTalk than C on a Sun 2 workstation? 
Lots.  On the other hand you will probably spend a lot more time in
development using C.  Smalltalk might provide a lot of stuff that you can
use already built in.  With 1 meg of memory it is so slow that it's
unusable.  With 2 megs it usable, but still slow enough to be annoying.  I
suspect more memory would make it faster but have not tried it.

> 	2) Is it stable and fairly bug free? 

As far as I can tell, no.  Although the latest release fixed some of the
worst bugs, there are still references to unimplemented things floating
around, and occasional blow-ups.  The ability to maintain versions and
recover from crashes that is provided by the Smalltalk environment helps a
lot though.

This is printed in bold face in one of the documents that comes with BS:

"Beware: Although we rely on BS, it is a research vehicle and not production
software.  Expect bugs (we certainly do)."

In spite of all this I would recommend you consider BS carefully; there are
a lot of advantages to developing software with Smalltalk.

Dan Everhart
John Fluke Mfg. Co.
{ decvax!microsof, uw-beaver, allegra, lbl-csam, ssc-vax } !fluke!dan