Subject RE: [IB-Architect] Re: Is it _really_ necessary to expose the
Author Claudio Valderrama C.
> -----Original Message-----
> From: Mark Shapiro [mailto:Mark.Shapiro@...]
> Sent: Lunes 14 de Agosto de 2000 9:49
>
> My opinion on aliases and such:
>
> 1) Storage in isc4.gdb (as an option?) sounds very good, though if the
> security on the alias table is the same as on the users table, it limits
> creating and modifying of aliases to sysdba. Is this a bad
> thing, or a good
> thing?

Jim Starkey wanted isc4.gdb to be removed from the engine. This database
doesn't enforce some automatic permission settings. The users table was
secured on IB6, previous releases weren't so restrictive. You could have an
alias table with permissions granted to PUBLIC and hence to any user able to
authenticate against IB. However, letting any user to modify any alias is a
nightmare, at least for me.


> 2) Why have $ to indicate an alias? When a connection is made to the
> server, the server can look at the remainder of the connect string and
> decide if it's an alias or a file or a redirect based on this order of
> preference:
> Alias - check configured alias storage location(s) to see if an alias
> exists with this name
> Redirect - Check to see if the connect string refers to another server
> Local file - If none of the above, try to open a local file

You should change the code in the Y-valve, probably. This is where the
subsystem to be used is decided, given the connection string, the ODS and
other useful information.

C.