The paradigm of containers and self-referenced names

[J. Landman Gay]

On 3/23/05 3:03 PM, Mark Wieder wrote:

> Ken-
> 
> Wednesday, March 23, 2005, 8:58:13 AM, you wrote:
> 
> KR> I think that if Rev were modified so that it would not accept unquoted
> KR> strings to be assigned to variables we'd be in a much better place, and it
> KR> wouldn't give aid to "sloppy" (IMHO) programming.
> 
> Menu:
> Edit | Preferences | Script Editor | Variable checking by default
> 
> Why this isn't on by default is beyond me except that it would force
> many programmers to clean up their code. If this had been left to me,
> I wouldn't even have made it an option.

And you would have reaped the ire of those involved in 20 years of xtalk 
legacy. ;)

For those of us who have used all the various forms of xtalk over lo 
these many years, the lack of variable typing and the ability to use 
variables without first declaring them is a major breakthrough feature 
of a modern programming language.

I still do not declare variables except under some rare specific 
instances. I consider it a freedom. When I read scripts written by those 
who do declare lists of locals inside their handlers, it looks messy to 
me and I wonder why they don't just go off and use Basic where it is a 
requirement. All that "dimming" really puts me off.

Is it sloppy programming to use a language as it was intended?

-- 
Jacqueline Landman Gay         |     jacque at hyperactivesw.com
HyperActive Software           |     http://www.hyperactivesw.com