It vs. the result (was Repeatloop in mouseDown...)

[Richard Gaskin]

Jan Schenkel and I have been enjoying a discussion of this in private email,
and this background on the issue from an email to him may be of interest to
the readers here:

--------------------------------------------------------------------------

Using "the result" over "it" in the context given appears to be an exception
to the conventions of the language, at least to this scripter only halfway
into his morning coffee (this appearance may change as I get closer to the
bottom of the cup).

While "the result" is not always reserved for error info, in most cases it
is.  But even more relevant here is the convention elsewhere in the language
of using "it" to return values from commands, esp. commands that create
things.

The integer returned by "send <message>" is effectively an ID.  While the
thing being created is not an object per se, commands that create objects
return the ID of the newly created object in "it".

Accordingly, many of us are in the habit of checking "the result" for error
info, and have come to expect "it" to contain non-error information for
commands that create things, esp. IDs.

IMHO, the original poster's expectations appear reasonable given these
conventions, unless there is some other linguistic rule at play that I'm not
taking into consideration.  Understanding the logic of this distinction
would seem useful for both experienced scripters and newcomers alike.

Being one of those "usability nuts", in cases of cognitive dissonance I tend
to favor user expectations over the designer's intentions, unless there's a
good argument for the design.  To Raney's and Miller's credit, most previous
cases in which a language issue appeared inconsistent at first glance have
turned out to have very solid reasoning behind them (about 99.9% of the
time).  

But nothing is perfect, and nearly everything can be refined just a bit
closer to perfection: once upon a time the MC engine used a non-standard
order for the items that described the printMargins (they were not L,T,R,B
as other rects are defined).  This was reported as a bug, and even though it
required changes to scripts that had relied on the old order it was promptly
changed in the next release, and today the language has one less exception
to remember.

In this case, we'll either see a change for greater consistency, or gain a
deeper understanding of the rules to help us understand why this instance is
truly consistent. 
 
-- 
 Richard Gaskin 
 Fourth World Media Corporation
 Developer of WebMerge 2.0: Publish any database on any site
 ___________________________________________________________
 Ambassador at FourthWorld.com       http://www.FourthWorld.com
 Tel: 323-225-3717                       AIM: FourthWorldInc