Re: [cx-oracle-users] an old question

[Chris G. <chr...@to...>]

Hi -
I've been following the discussion in this thread and have to say I think
the objective of the whole exercise is flawed.  You seem to be trying to
write a set of wrappers to make Oracle appear to behave in the same way as
other RDBMS, and I think that isn't realistic.  To get the best out of any
database you need to exploit its features and avoid its "quirks".  Oracle's
features include returning a data-sets as a REFCURSOR where you don't know
in advance how many records are in the set - you and Oracle only find that
out when you fetch past the last row.  Similarly, if you're talking of
deleting a million rows and trying to return all the deleted data into a
collection of some sort that gets passed back to the invoker, what are you
going to do with that data?  It sounds like you'd be better off trying a
"SELECT .. FOR UPDATE" and then deleting each row - inefficient in Oracle,
but that seems to be of secondary importance.

I really think the further you go with this, the more problems you're
likely to encounter and you're never going to get any decent performance
out of what is a very expensive RDBMS.

C



On 30 April 2013 09:17, Glyph <gl...@tw...> wrote:

>
> On Apr 30, 2013, at 1:04 AM, Mark Harrison <mh...@pi...> wrote:
>
> On 4/30/13 12:21 AM, Glyph wrote:
>
> As I keep pointing out:
>
> Currently, I'm doing this with host variables, but that's a limited
> solution [...]
>
>
> You're trying but apparent failing to make your point...
>
> I tried helping you out earlier, but your response left me not
> caring a lot whether or not you solved your problem or not.
>
>
> I do appreciate you taking the time to help, even if your answer wasn't
> directly useful to me.  I realize what I'm looking for is a bit esoteric.
>  "What you are trying to do is impossible" is a totally acceptable answer -
> I am really just trying to understand *why* Oracle makes this impossible.
>
> The database-specificity of cursor.var() makes sense to me, idiomatically,
> but the restricted size of query results seems odd.  For example, if I want
> to 'delete...returning...into...' a million rows, rather than the default
> allowed ~1000 or so, how can I do that?  Would a pipelined procedure with
> PIPE ROW address this problem, and the general expectation is just that I'd
> be able to put such a procedure into my schema?
>
> -glyph
>
>
>
> ------------------------------------------------------------------------------
> Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET
> Get 100% visibility into your production application - at no cost.
> Code-level diagnostics for performance bottlenecks with <2% overhead
> Download for free and get started troubleshooting in minutes.
> http://p.sf.net/sfu/appdyn_d2d_ap1
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>
>