Tri sur GUID SQL Server 2008

Question

Quelle est la meilleure façon de trier les enregistrements sur Guid. Ci-dessous ne donne jamais de bons résultats.

With UIDs As (--       
      Select ID = 1, UID = NEWID() 
    Union Select ID = 2, UID = NEWID() 
    Union Select ID = 3, UID = NEWID() 
    Union Select ID = 4, UID = NEWID() 
    Union Select ID = 5, UID = NEWID() 
    Union Select ID = 6, UID = NEWID() 
    Union Select ID = 7, UID = NEWID() 
    Union Select ID = 8, UID = NEWID() 
    Union Select ID = 9, UID = NEWID() 
    Union Select ID = 10, UID = NEWID() 
    Union Select ID = 11, UID = NEWID() 
    Union Select ID = 12, UID = NEWID() 
    Union Select ID = 13, UID = NEWID() 
    Union Select ID = 14, UID = NEWID() 
    Union Select ID = 15, UID = NEWID() 
    Union Select ID = 16, UID = NEWID() 
) 
Select * From UIDs Order BY UID, ID 

Au lieu d'utiliser Guid, comment puis-je générer des entiers séquentiels dans mon select.

Merci

Answer 1

Ne pas est la meilleure réponse.

Cela explique comment SQL Server: How are GUIDs sorted by SQL Server?

Vous avez NEWSEQUENTIALID mais cela est not guaranteed pour trier comme prévu. Et pourquoi utiliser un GUID lorsque vous avez des fonctions de classement?

Answer 2

SELECT * 
FROM myTable 
ORDER BY CAST(myGuid AS VARCHAR(36))