[ruby-talk:02320] Re: Question about attribute writers

On Sat, 1 Apr 2000, Clemens Hintze wrote:
> Dave Thomas writes:
> > Clemens Hintze <[e-mail]> writes:
> > 
> > > At least I would expect it! It is like answered in your FAQ! How does
> > > Ruby determine what a variable is? It is really stright-forward, IMHO.
> > > 
> > >    Look if there was an assignment before
> > >       yes => compile a variable access
> > >       no  => compile a method call.
> > 
> > True - I asked the wrong question. Rather than asking "is this the
> > expected behavior", I guess I meant "is this the behavior that someone 
> > would expect?" It seems to violate the principle of least surprise.
> 
> Hmm difficult thingy here, I think. I would never assume that 
> 
>    age = 12;
> 
> could invoke a method, because I *know* that this syntax normally
> means assignment to a local variable. It would be very strange to me

Agreed.

> > Perhaps a warning would a good idea in these circumstances:
> > 
> >   warning: local variable eclipses instance method
> 
> Erhm ... IMHO this warning is wrong! Because there is *no*syntax* that
> allow 'age = 12' meaning a method call to 'age='. So if this is

No. Warning is really good, I think! It should just not be issued for 
age = 12. Only for next case as you then "correct yourself" ;)

> I would propose the opposite, if we really want to issue an
> warning. The warning should be issued for the case
> 
>    age / 12

> But I admit, that this solution would probably issue too much warnings
> because the 'age / 12' syntax is often used!

And your suspects (is it suspicious'? That has ambiguity too, help :) are
probably right. But, what I would like to see is that I could ask
interpreter to give his opinion, which is of course always right, to check
if my assumptions are always in line with reality.

So optional if -w should give some warnings, this could be activated at
-w=3.

> Perhaps it would be better to not change anything, and describe this
> case very carefully in the books and docs. Every language has its
> gotchas. We cannot eliminate all of them. This is one of Ruby.

I fully agree.