Tony Gravett wrote:

>> ps -- comparing Rev with REALbasic is apples and oranges. They are
>> superficially similar, but in detail are far too different to compare
>> reasonably. RB has some strong suits, but Rev has more, and their strong
>> areas
>> are again very different.
> 
> I am sorry, Geoff, but this "PS" begs the question: why not create a
> comparison chart that helps potential new customers understand the "strong
> suits" of Rev in comparision with, say, five immediate apparent rivals?  Or
> is Rev "too different" to take this approach?

About RB and Rev (I'll say MC for engine features), here are some areas that
I personally can compare:

Text fields: MC is way, way on top. Try changing the color of 20 words in a
text field in RB. No, problem, each time you just choose the selection start
and length, then set the color; and just use another field offscreen to hide
it from the user, or better yet, make your own style resource! Good research
experience for you. ;-)

MC text fields can also be scrolled to any pixel, vertical or horizontal,
compared to line by line for RB. RB does have some useful features now to
get the character based on row of a field or vice versa, which I think is
the only thing it currently does easier than MC for fields.

Text: xTalk is ahead of its time. Referring to words and items and lines to
get, set, and delete chunks of text is revolutionary in dealing with human
language text. There's still nothing to compare in other languages that I
know of.

Scripting: RB's visual representation of code elements and properties looks
"cool", but for many tasks it's very tedious to have to define each global
or method with a dialog--and very difficult to manipulate large numbers of
them--this is a weakness of RB compared to other BASICs too, not just to MC.
I find the xTalk way better for almost every aspect. RB does allow control
arrays and classes, though, which share scripts and features.

Developing: There's just no serious comparison between two systems, where
one lets you design and play on both platforms, and the other lets you
"design and guess how it will look and work" on one platform and "play and
guess what's wrong" on the other.

Also, the card metaphor and permanent automatic storage of information in
controls puts you way ahead in MC and Rev; you make it and forget about it,
whereas with other tools, that's just the first step, you still need to
provide all the save and open and setup stuff.

Low-level access: RB excels here with a simple format for declaring system
calls on both Mac and Windows from DLLs and System Libraries. MC needs this
to round out its otherwise powerful features. RB currently simply rules in
this area, no comparison. It also takes pressure off the company and allows
developers to make their own solutions for simple additional features while
only having to use one development tool for everything.

Graphics: RB has more raw ability hidden in the canvas controls, but the way
to access it is much more primitive. I like scripting graphic objects in the
high-level way that MC and Rev allow. The "move" commands and support for
animating GIFs in MC simply blew me away. For most practical animation
tasks, using Rev and MC is much more enjoyable to get final results in a
fraction of the time it would take with RB, at least in my experience. I
think the new features in the new MC--accessing image data directly--are
smart and will help to close the gap about lower-level drawing access, with
help from some third-party routines.

IDE: RB has to stick close to Microsoft's lead here, with an old-fashioned
list of object properties which are pretty bad in most respects compared to
the clean, intuitive Rev equivalents. RB does have a neat feature about
auto-aligning objects on the window. RB's IDE can't be customized and is
very difficult to add on to. As for Rev, that's the name of the game, since
the whole thing is made in MC. That right there could save a big developer
loads of trouble by making a simple customization as an add-in to the IDE.

Information access: MC needs to get the clipboard access and full drag and
drop support taken care of. These are areas where any user of software hopes
for both to work for just about any kind of information, and expects at
least one to! I'm not sure, maybe the next version is adding a lot of it,
but currently I think RB is still ahead here.

I personally think that working with MC through Revolution is about 10 times
as fast for the kind of things I need to make as doing the same thing with
RB. But, disclaimer, that's just my experience for my own personal needs.
Additional note: also 10 times as enjoyable during the development process!
I do think RB is a great product in its area, but I prefer the xCard genre
in all respects. I think MC with Revolution is the final answer to practical
programming, and if the system calls ability is added, and a few more
various features here and there, it will be the final answer to practical
programming, period.

Curry Kenworthy
[EMAIL PROTECTED]


Reply via email to