Transcript and Dot Notation

[Alex Tweedly]

Judy Perry wrote:

>1.  It being "optional" didn't stop it from destroying accessibility to
>verbose Lingo in Director.  Latecomers to Director didn't have any other
>learning "options" or "choices" than dot.speak.
>
>  
>
Yes, I think that's a danger, especially if a large part of the audience 
for books (and tutorials, etc.) is going to be converts from other 
languages, who might find the dot notation more comfortable and easier 
than "natural" language style.

>2.  Optional isn't the same as transparent.  I'd be less leery if I could
>look in a crystal ball and see that it was implemented in a transparent
>fashion.
>
>  
>
Well, we know that allowing a mix is going to be hard for RR. So why not 
make it harder :-)

It wold be nice if there were a perf switch in the IDE: click 
"dot-style" and your scripts appear in dot-notation style, click 
"natural language" and your scripts are converted to traditional 
Transcript style; so the same statement appears as either
    set the stopColor of TrafficLight to red
or
   set the TrafficLight.stopColour to red
(or even     put red into TrafficLight.stopColour)

[btw - note how it corrects your spelling at the same time :-) ]

>3.  It's a slippery slope.  Once it heads down that path, there will be
>little reason to implement new functionalities in Rev taking care to do it
>in a natural-language fashion.
>
>  
>
Again that's a danger, which the community would have some role in averting.

But remember there are benefits to dot-notations style. There are the 
intellectual advantages that make OO a good approach, but I'm the wrong 
one to argue for them. What *I* like about dot-notation is how it allows 
the IDE to make it easier to use for casual, or inexpert, users.

I like the fact that in my other-language IDE, I can do something like type
  status = smtplib.      and then wait for 2 seconds

The IDE looks up the list of all properties and functions available 
within the smtp library, and displays a pop-up box showing them. I can 
scroll through that to see what all ones are available, hit CR to select 
the current on (or type the first few chars and it auto-scrolls there), 
etc. If I select one, e.g. type 'rf<tab>' to select the rfc822 
sub-library, I get
   status = smtplib.rfc822  

and again when I type a dot and pause for 2 seconds, I get a list of all 
the possible completions.

Then if I type Add<tab> it selects the first matching function, so my 
display now shows
   status = smtplib.rfc822.AddressList

and if I type the open paren, I see parameters needed by that function, 
and the documentation describing them.


I don't know for sure that it would be *impossible* to do that in a 
Transcript IDE - but it seems like it would be much much harder than it 
is in a dot-notation language. And it makes using an unfamiliar library 
so much easier than hoping between a script editor and one (or more) 
documentation windows.

-- 
Alex Tweedly       http://www.tweedly.net



-- 
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.375 / Virus Database: 268.1.0/269 - Release Date: 24/02/2006