On Fri, 30 Sep 2005 09:03:09 +0200
Matthijs Willemstein <[EMAIL PROTECTED]> wrote:

> On Thu, 2005-09-29 at 20:21 +0200, Mattias Gaertner wrote:
> > > Pressing <ctrl><shift>C (thus invoking code completion) a new skeleton
> > > for procedure TFoo.FooProc is made. I think this is not the correct
> > > behaviour. Either it should do nothing at all, or it should adapt the
> > > previously created version.
> > > Any opinions on that?
> > 
> > I would prefer automatic adapting too.
> > But some Delphians would expect something different:
> > [cut]
> > So, the we should add automatic adapting as a checkbox to the class
> > completion options.
> > 
> > [cut]
> > 
> > Now, you don't know, which procedures corresponds.
> > What should happen?
> Oeps! I did not see all this. :) 
> 
> But what happened now was that in the end there was one declaration and
> two implementations. Usually the compiler starts complaining about such
> cases. :)
> 
> Could something be implemented like with the navigation from declaration
> to implementation? If you press <ctrl><shift><up> the cursor jumps from
> declaration to implementation (or v.v.). If either is changed it looks
> for the best fit and stops on the position of the first change.

Yes. Although Method jumping can guess wrong. If it is not sure, it sets a
jump point.
We can't do that for class completion. Nevertheless, if it is sure, it
should adapt.


> So this would mean automatic adaption and in cases where this is not
> possible a message or so.
> Changing both the implementation and declaration does not sound like a
> good idea to me. This might lead to not wanted overloaded methods. 

I totally agree. But many Delphians like that feature. That's why we need
some options.

 
> But it is clear there is more to this, then I thought of when posing the
> question! :)

What about procedure completion?
It would be nice, if here auto adapting would work too. But here you don't
have a 1:1 relationship.


Mattias

_________________________________________________________________
     To unsubscribe: mail [EMAIL PROTECTED] with
                "unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Reply via email to