>>> On 3/21/2008 at 12:34 PM, in message
<64ideqF2bn43lU1@mid.individual.net>,
Serge Rielau<srielau@ca.ibm.com> wrote:
> Frank Swarbrick wrote:
>> Oh hell, I missunderstood. It looks like this feature is available for
>> PROCEDURES, FUNCTIONS and METHODS, but not for PACKAGES. Which I'm sure

> is
>> why you said it was an ideological issue and not a technological issue.
>>
>> Do you think that if a PMR was made to allow this for packages that IBM
>> would consider implementing it?
> It needs a carrot and whip. "Invonvenience" makes for a bad incentive.
>
> Now we have drifted a bit from the original request. The OP was asking
> for table-objects. You are asking for package objects.
> This is embedded SQL in C, Cobol, ... applications?
> Because if it's about external routines then the schema.* approach works
> on the routine level.

Yes, I think I have drifted. Seems to me that allowing GRANT of any type of
object for schema.* would be useful.
In my most recent message I was referring to packages created by embedded
SQL (Cobol) applications. But I would think it would be useful for tables,
views, and other things as well.

But if it's the general philosophy that these priviliges should be granted
on an explicit one-by-one basis, well, who am I to argue. (Rhetorical
question. I'll argue about anything. <g>)

Frank