Newsgroups: comp.lang.java,comp.lang.c++,comp.lang.smalltalk
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!scramble.lm.com!news.math.psu.edu!chi-news.cic.net!newsfeed.internetmci.com!howland.reston.ans.net!newsxfer.itd.umich.edu!qiclab.scn.rain.com!gemstone.com!servio!servio!aland
From: aland@servio.slc.com (Alan Darlington)
Subject: Re: Advice to Java proponents (was Re: Will Java kill C++?)
Message-ID: <1996Apr11.184145.17550@slc.com>
Sender: news@slc.com (USENET News)
Nntp-Posting-Host: servio
Organization: GemStone Systems, Inc., Beaverton OR, USA
References: <31684F33.2528@ibm.net> <denatale-0804960926250001@grail1213.nando.net> <316D09A4.7A92@possibility.com>
Date: Thu, 11 Apr 1996 18:41:45 GMT
Lines: 81
Xref: glinda.oz.cs.cmu.edu comp.lang.java:39339 comp.lang.c++:184093 comp.lang.smalltalk:37128

Todd Hoff <tmh@possibility.com> writes:
> Rick DeNatale wrote:
> 
> > Smalltalk also had the benefit of being used for many years before it was
> > widely available and therefore not having to 'grow up' in the intense
> > arena generated by the frenzy accorded C++ and now to an even larger
> > extent Java.
> 
> The why did not smalltalk take off?

Some think it is...  :-)  Historically, lack of availability, high
cost, high learning curve, the reluctance of managers to risk their
necks on unproven (i.e. IBM is not using it  :-) technology, etc.
Being better (IMHO) does not guarantee success.  Just ask Apple...
As a former co-worker liked to say, you can tell the pioneers by
the arrows in their backs.

> One question i asked a while ago, is if smalltalk is so easy to
> use and develop in then why are the tools so damn expensive?

A major software company can afford to sell 10,000,000 copies
for $100 each.  If they only have 1,000 customers, they go broke
at this price...  These guys are not totally dumb.  Also, it is
_expensive_ and _hard_ to make things easy-to-use - ask any
application developer!  (I have been one for 30 years.  Sheese...)

> Why are there expensive runtimes?

A way to keep the initial development costs down, and let you
pay the rest when _you_ have revenue coming in.  (This has always
been a controversial point - not all Smalltalk vendors do this -
but would you rather pay _more_ initially?  Smalltalk is not a
cheap product to maintain, enhance, and market.  You should see
the list of new features that people want!  :-)

> Why is there no portable standard?

The Smalltalk language is completely portable among all major
vendors and platforms. (Alright, there are - or were - a few minor
gotchas like declaring local variables inside blocks, but compared
to C and C++, this is nothing.  Also, the new standards will
eventually take care of this.)  What really kills portability in
Smalltalk is the differences in class libraries, and C++ certainly
shares this problem.  :-(

(I currently use VisualWorks 2.5, Visual Smalltalk 3.1, and
VisualAge 3.0.  Our product runs on everything from mainframes to
Unix workstations to PCs, so I have had experience with portability
problems!).

> Java is free as air compared to smalltalk. Java is in addition
> a good language with near universal support. That's a lot to 
> overcome with tepid arguments of productivity gains.

While productivity may not seem very important to you, companies
live or die on this issue.  In big New York financial institutions,
security traders need new kinds of securities in their trading
programs in a matter of days (at least if they want to keep their
jobs  :-).  These Fortune 100 (or 200) companies have tried many
solutions (including C++), and a lot of them are now using Small-
talk because major changes _can_ be made in a matter of days.  I
have previously worked in this field and have done it myself...

A lot of companies are finding that if you don't have an application
done it time, it doesn't make any difference how fast (or how good)
it is.  Your competitors have done you in, and you are in Chapter 11.
And since most applications are needed yesterday, the pressure
is on us programmers...   As if this isn't bad enough, many companies
now want enterprise-wide applications, which means that they will be
orders of magnitude more complicated.

Smalltalk is certainly not the ultimate in fast development (after
all, the first version of Smalltalk was done in 1972!).  But I have
to believe that future application development systems are going to
be a lot closer to Smalltalk than to C++ or Java.

  Just a bunch of my opinions,
  Alan
    (standard disclaimer)
    (plus my crystal ball isn't working today.  :-)

