Rev & OOP

[Dan Shafer]

Oh, darn. I was sort of hoping this thread would fizzle out. Then you 
had to throw down a gauntlet, Rob, and you know how I am about 
gauntlets (and, for that matter, things littering the landscape like 
thrown gauntlets)....

I spent MANY years training and conditioning myself to think about the 
world around me in terms of objects. From a programming perspective, I 
find myself always more comfortable dealing with objects in the sense 
in which Smalltalk and Java (and decidedly NOT C++) think about them. 
OOP languages and OODBMS tools have *always* been more productive for 
me than procedural and relational models because once I trained myself 
to "think in objects," those approaches felt -- and were -- unnatural 
to me.

Far from being the "buzz concept du jour," OOP has been around, viable 
and in many places on the globe an all-but-inviolate standard for more 
than 30 years.

Now, that is not to say or suggest that every programming language that 
isn't strict OOP isn't usable or useful. Far from it. I use Revolution 
and Transcript because, even though it's not an OO environment, it is 
what I refer to in my books as "object-LIKE." That is, it represents 
enough of an accommodation of the key ideas of object orientation to be 
usable and useful on medium-sized, single-programmer projects involving 
non-object data. But I must say that if I had a choice of using an 
equivalent development environment that was syntactically as clean as 
Transcript or Smalltalk or Python and gave me the advantages of 
Revolution (cross-platform delivery while developing on my platform of 
choice, true stand-alone creation, great widget library, transparent 
database access), I'd switch in a New York nanosecond. The truth is, no 
such tool exists yet.

So I would agree that the programmer who rejected Revolution out of 
hand without digging more deeply into the advantages it offers and 
shares with OO environments was hasty and ill-advised (and probably, as 
you say, more interested in eliminating alternatives than in finding 
the correct one). But to dismiss OO out of hand is, IMNSHO, equally 
short-sighted. As you so rightly say, the two big concerns are 
programmer productivity and code maintainability. And in those 
respects, Transcript is awfully hard to beat.

FWIW.


On Apr 30, 2005, at 7:21 AM, Rob Cozens wrote:

> IMF(oole's)O, the programmer who ruled out RunRev as a development 
> platform on the basis of it not being a true OOP language was simply 
> looking for a reason to pan it rather than do the kind of in depth 
> analysis required to properly evaluate its potential.
>
> The bottom lines for software development are real-world productivity 
> and code maintainability, not compliance with the "buzz concept de 
> jour".
>
> Rob Cozens CCW
> Serendipity Software Company
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dan Shafer, Co-Chair
RevConWest '05
June 17-18, 2005, Monterey, California
http://www.altuit.com/webs/altuit/RevConWest