working method?

[Dan Shafer]

Charles...

As you can already see, this was far from a dumb question. It is one  
that comes up every once in a while on the list. I am always  
interested to see the various opinions about the best way to factor  
code and organize applications.

 From my early days in HyperCard, I have followed, more or less  
religiously, Sarah's recommendation to keep all handlers as low in  
the hierarchy as makes sense and no lower. I like the self-contained  
feel this gives UI components. If I copy a button from Stack A to  
Stack B, I don't have to go get its script separately from the group  
or card or stack and copy that over to Stack B separately.

To keep track of where in the hierarchy a particular handler or  
function is defined, I use the simple expedient of comments on the  
lines where these operations are invoked. I tried naming conventions  
but for me, they either make the name less readable and  memorable or  
are too hard to type or both. So I just add a comment. It helps that  
I'm a very fast typist, so this kind of thing, once the habit is  
ingrained, takes little extra time. It does have the downside that if  
I relocate the handler, which I do fairly often particularly early in  
the development cycle, I have to remember to change the comment. But  
then if I used a naming convention, I'd have to change the handler  
call itself; at least my way the script doesn't break.

I have not yet learned to make extensive use of libraries. As an old  
HyperTalker, I can be heard to mutter such things as "Libraries? We  
don't need no stinki' libraries, man!" But I do see the value and  
wisdom of using them, particularly for functionality you want to  
reuse in multiple applications or stacks. So recently I've started  
factoring out some of my code into small libraries for this purpose.

Clearly this whole issue is a matter of style and what works for you.  
The idea of using some sort of coded characters as comments to  
indicate handlers that are in the card, group or stack is a shorthand  
way of doing what I do and probably works particularly well for  
slower typists. It also makes the comments easier to find.


On Oct 26, 2005, at 7:25 PM, Charles Hartman wrote:

> I know this is going to sound like a *really* dumb question, if  
> only because it's so vague. But I'm wondering how people adjust  
> their workflow to the way Transcript's code is dispersed among many  
> separate scripts.
>
> I keep getting lost. I keep forgetting where my code is that does  
> such-and-such. (Which script was that in?) So I keep losing track  
> of what I was about to do next, and my concentration falls apart.  
> It's making Rev *much* slower for me to program in than supposedly  
> more complicated languages like Python and C++.
>
> Anybody think this makes any sense? Any hints how to think about it  
> differently?
>
> Charles Hartman
>
>
>
> _______________________________________________
> use-revolution mailing list
> use-revolution at lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your  
> subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-revolution
>



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dan Shafer, Information Product Consultant and Author
http://www.shafermedia.com
Get my book, "Revolution: Software at the Speed of Thought"
 From http://www.shafermediastore.com/tech_main.html