azure db drop pk and add a new one - sql

Question

I have MS SQL database on Azure , and I want to drop my Primary key constraint and add another column as a primary key , every time I try to run the script to drop the primary key I get the error:
" Tables without a clustered index are not supported in this version of SQL Server. Please create a clustered index and try again."
by running this script:
IF EXISTS (SELECT * FROM sys.key_constraints WHERE type = 'PK' AND parent_object_id = OBJECT_ID('dbo.Paypaltransaction') AND Name = 'PK_PaypalTransaction')
ALTER TABLE dbo.PaypalTransaction
DROP CONSTRAINT PK_PaypalTransaction
GO
Then I go and try to create another primary key :
-- add new primary key constraint on new column
ALTER TABLE dbo.PaypalTransaction
ADD CONSTRAINT PK_PaypalTransactionId
PRIMARY KEY CLUSTERED (PaypalTransactionId)
GO
and I get this error:
"Table 'PaypalTransaction' already has a primary key defined on it."
I understand the error's , but I can't delete the primary key because it appearently has to have one , and I can't add a new one because I already have one . Am I just stuck forever and ever and ever with the wrong column as my primary key :'(
I see This Question here that is the same and they ended up dropping the table and creating a new one - no way that is the only way to do this , that is just silly.

Answer 1

Copy over from the DBA.Se answer:
Q5) Can the primary key of a table be modified when is enforced via the clustered index if the table is populated?
A: No. Any operation that converts a populated clustered index into a heap will be blocked in SQL Azure, even if the table is empty:
create table Friend (
UserId int not null,
Id int not null identity(1,1),
constraint pk_Friend primary key clustered (UserId, Id));
go
insert into Friend (UserId) values (1);
delete from Friend;
go
alter table Friend drop constraint pk_Friend;
As a side note: the constraint can be modified if the table is truncated.
The workaround to change the PK constraint of a populated table is to do the good old sp_rename trick:
create table Friend (
UserId int not null,
Id int not null identity(1,1),
constraint pk_Friend primary key clustered (UserId, Id));
go
insert into Friend (UserId) values (1);
go
create table FriendNew (
UserId int not null,
Id int not null identity(1,1),
constraint pk_Friend_New primary key clustered (Id, UserId));
go
set identity_insert FriendNew on;
insert into FriendNew (UserId, Id)
select UserId, Id
from Friend;
set identity_insert FriendNew off;
go
begin transaction
exec sp_rename 'Friend', 'FriendOld';
exec sp_rename 'FriendNew', 'Friend';
commit;
go
sp_help 'Friend';
The sp_rename approach has some issues, most importantly being that permissions on the table do not carry over during the rename, as well as foreign key constraints.

Answer 2

I have just tested and it seems that on SQL AZURE V12 it is possible!