Cross Apply And Newid

I've been trying to figure out why these two return a different amount of distinct ids...
Is that a bug in optimization?




Code Snippet


USE AdventureWorks
go
Declare @myXml as xml
set @myXml = '
<lol>omg</lol>
<lol>rofl</lol>
';

WITH locations as
(
select newid() as Id
from Production.ProductModel
where ProductModelID in (7, 8)
)
select locations.*, T.c.value('.','nvarchar(max)') from locations cross apply @myXml.nodes('(/lol)') T(c);

with mytable as
(
select newid() as Id
union
select newid()
)
select mytable.* , T.c.value('.','nvarchar(max)') from mytable cross apply @myXml.nodes('(/lol)') T(c);