da-zero (8/26/2010)

---

I do not really understand how the scope of the type variable works. It seems you can define it over and over again, without an error being thrown?

martin.whitton (8/26/2010)

---

It's interesting that a variable can be declared multiple times if it's within a "while" loop (provided that it's only declared once in each iteration).

When I first read the question I thought that declaring @a within the loop would cause an error on the second iteration.

IF 1 = 2
BEGIN;
  PRINT 'Skip this';
  DECLARE @a int;
END;
ELSE
BEGIN;
  SET @a = 1;
END;
SELECT @a;

  GOTO Label2;
Label1:
  DECLARE @b int;
  SELECT @b;
  GOTO Label3;
Label2:
  SET @b = 3;
  GOTO Label1;
Label3:
go

stewartc-708166 (8/25/2010)

---

The only possible use for the float datatype that I have found is in the calculations used in astro-physics, where the distance between stars and galaxies is an approximation, at best

Open Minded (8/26/2010)

---

isn't it possible to use approximation if one is not specific of the exact value of the approximation, like "less than" and "greater than" ?

Pete Cox (8/26/2010)

---

BTW, I think ROWCOUNT stops having this effect in versions of SQL server later than 2005