REALbasic vs. Revolution

[Lorin Rivers]

On 10/10/02 8:26 PM, curry <curry at kagi.com> wrote:

> The question of RB vs. Rev is interesting and is getting more so, not
> less. The main reason I'm still interested in RB is its potential for
> 3D. That's the only place where Rev doesn't offer anything.
> 
> The main difference for me is that I *enjoy* working in Rev. I'm not
> sure if that point has been brought up as much! If I want to make
> something creative, I feel that Rev helps me do that because it's
> enjoyable to use. I like the scripting syntax most of all, and there
> are also some things about the developing environment that I like to
> use better.

I think this is a personal thing, and not an absolute truth. I think it has
more to do with an individual's learning style than anything else.

> I feel that scripting syntax is not only convenient to learn, but
> it's *superior* to other programming syntax. Readability is not just
> nice, it's also practical. (And if more of the world's code was
> written with syntax more similar to the English-like scripting syntax
> we find in HyperTalk and Transcript, we'd have a better world.) I
> like BASIC too, but the way MS took BASIC in adapting to GUI just
> didn't make it very readable. A more readable syntax would be
> possible for supporting those elements, and a few BASICs out there
> have been a bit more creative, but as it is, scripting is much better
> if you want something English-like;  which of course was what BASIC
> set out to do originally, right?.

More English like than COBOL or Pascal, certainly. That said readability can
significantly and negatively impact write-ability, at least in my
experience. I have to say, I'm one of those people who has a really hard
time with xTalk type languages (including AppleScript) but not with
REALbasic (my first success in learning programming after YEARS of trying
all the others).

> Here's another thing for me: I don't like having to declare every
> single variable. I was actually flamed by many users on the RB list
> when I suggested that RB allow writing without explicit
> variables--which I think VB has an option for, and some people had
> let that trip them up with a hard-to-find problem until they finally
> thought of turning on explicit variables. Sounds like not everything
> is right for everyone and it's their problem, but these users were
> scared out of their pants to allow this capability--even though they
> could just not use the option themselves, they couldn't stand the
> thought of someone else doing it. That kind of mindset just didn't
> impress me, especially since RB seemed to accept it readily.

This is a religious issue. It's interesting to compare PHP and REALbasic
(the only two languages I have passing familiarity with) in this regard. In
PHP variables are weakly typed but operators cannot be overloaded. In other
words you can use the "+" operator with a string "123" and double "2.34" and
like so (note not real PHP):
myresult = "123" + "2.34"

print myresult

and get: "125.34"


In REALbasic, you might have the same two things:
dim myString as String
dim myDouble as Double
dim myResult as Double
dim myOtherResult as String

myString = "123"
myDouble = 2.34

myResult = val(myString) + myDouble

In this case, myResult = 125.34

If you want to display that, you would convert it to a string:

myOtherResult = str(myResult)

Here's where it becomes more interesting.

There's no concatenation operator in REALbasic (I think it's "&" in
AppleScript).

So, if what you actually want is "1232.34" you do this:
myOtherResult = myString + str(myDouble)

This is sort of a wacky example, but overloading is something that makes
Object Oriented Programming powerful.


> Then there's the runtime development environment. You always have to
> run and quit with RB, and I like the way you can "tinker" in Rev, and
> really run  it outside the environment when you want.

I, of course, don't find this an advantage in Rev or a disadvantage in
REALbasic. I like it fine, and don't mind the run, debug, quit routine. For
one thing, it's quite fast in REALbasic, though not in all IDEs that follow
this model.

> <Rev Feature Request Snipped>
> So I have many reasons to like Rev and will do everything in Rev that
> is supported. But RB is also a very good tool and with a Windows dev
> environment coming out, that's going to make it a lot more realistic
> for cross-platform development.

Well, I'm glad you think so. One other thing that will occur in the new
version is the debug process being separate from the IDE process, so should
you make an error in your code that causes your app to crash, you won't
bring REALbasic down with it.

> <multimedia related requests for Rev snipped>
> 
> <expression of a desire for a third-party community for Rev snipped>
> 
> Both Rev and RB are good products. But the ultimate factors for me
> are what I can do and how I can do it. I enjoy working with Rev a lot
> more, and I'm trusting that the few remaining areas that need more
> support will be supported.

Horses for courses, yes! I appreciate the thought-provoking email.

-- 
Lorin Rivers                      mailto:lrivers at realsoftware.com
Vice President of Marketing       512.328.REAL (7325) x712 v
REAL Software                     512.328.7372 f
1705 South Capital of Texas Hwy.  http://www.realsoftware.com
Suite 310                         REALbasic: the powerful, easy-to-use
Austin, Texas 78746               tool for creating your own software
                                  for Macintosh, Mac OS X, and Windows.