Comment 1 Deepak Azad 2011-05-03 13:20:31 EDT

The fix for Bug 287123 was to offer 'Join variable declaration' quick assist when the initializer is *null*. If the initializer is not null then the quick assist is not offered, because the initializer may be important to the user and we do not want to delete it silently/without warning.

So what needs to be done for primitive types ? Offer the quick assist when the variable is initialized to the default value of the primitive type ? e.g. int i= 0; or boolean b= false;

In my opinion we should not do this, as this would be confusing.

Comment 2 Dani Megert 2011-05-04 03:24:19 EDT

> So what needs to be done for primitive types ? Offer the quick assist when the
> variable is initialized to the default value of the primitive type ? e.g. int
> i= 0; or boolean b= false;
Yes.

> In my opinion we should not do this, as this would be confusing.
Why should we treat the default value for objects different from the primitive ones?

Comment 3 Deepak Azad 2011-05-04 04:21:14 EDT

Double d= 0d;
Double d= 0.0d;
Double d= 0.00d;

- In which case is the quick assist offered? (rhetorical question)
- And from a user perspective there is no scope of confusion here?

Comment 4 Dani Megert 2011-05-04 04:28:33 EDT

(In reply to comment #3)
> Double d= 0d;
> Double d= 0.0d;
> Double d= 0.00d;
> - In which case is the quick assist offered? (rhetorical question)
In none, as those assign real objects which are different from the object default ('null').

> - And from a user perspective there is no scope of confusion here?
Why?

Comment 5 Deepak Azad 2011-05-04 04:45:53 EDT

oops.. the D is double was supposed to be lower case

double d= 0d;
double d= 0.0d;
double d= 0.00d;

Comment 6 Dani Megert 2011-05-04 04:59:14 EDT

> double d= 0d;
> double d= 0.0d;
> double d= 0.00d;

Those are all identical to the default (0.0d) and hence the quick assist should be there.