I often wonder if the folks writing technical documentation, particularly the detailed stuff in the "Help" files, ever treid to use it.  Are they rewriting stuff the unlettered programmer told them at the coffee machine or did they really work thru the problem?  I am trying to set up a SQL Server linked server so a .NET program can update Oracle.  You set up the connection to Oracle by editing teh TNSNAMES.ORA file, and test it by opening the database with SQLPlus.  That is easy.  The TNSNAMES has a "description" and a bunch of parameters.  None of the SQL Server writers relate the TNSNAMES to the SQL parameters for the two stored procedures: sp_addlinkedserver and sp_addlinkedsrvlogin.  And understanding the jargon depends on understanding the jargon, so someone new to some small corner of this world is caught in a loop.  Does this make sense:

[ @locallogin = ] 'locallogin'  Is a login on the local server. locallogin is sysname, with a default of NULL. NULL specifies that this entry applies to all local logins that connect to rmtsrvname. If not NULL, locallogin can be a SQL Server login or a Windows login. The Windows login must have been granted access to SQL Server either directly, or through its membership in a Windows group granted access.

Some days I could take a bunch of these snippets, out of context, and do a humerous recitation at a comedy club, like a comedian did with the IRS code.