On Apr 14, 2008, at 11:30 AM, Chuck Hill wrote:
On Apr 11, 2008, at 8:50 PM, Jeff Schmitz wrote:
That seems like an exceedingly odd way to get a single entry on a
single Pool. Does the code that calls this not know the Pool?
How does it get the Pool and Entry names?
Yes, it will only have the Strings (you gotta start somewhere).
Even if you only have the names,
EOQualifier entryQual = ERXQ.and(Entry.NAME.eq(entryName),
ERXQ.equals(ERXQ.keyPath(Entry.POOLKEY, Pool.NAMEKEY), poolName));
Will the fact that entryName is not globally unique (it's only
unique within a pool) have any affect on the speed of the fetch?
There will be a LOT of entries, and not near as many pools. While
I have no idea how the above query will ultimately run, I would
think finding the pool first, and then just searching for the
entry name from within just that pool would be the most efficient
way.
No, that will be slower. You will be doing two fetches, where one
would do. Indexes (you will need to add them manually) will keep
the queries fast.
If you want to have any concerns about optimization and WO, you
really need to get familiar with SQL optimization for relational
databases.
<snip>
I missed the first part of this thread (through some lost e-mail). I
wanted to check about something though.
As I am looking at this API, I am noticing that the er.extensions
package is really, really, really big with lots of different kinds of
classes and that there is no package documentation at all.
Would anybody mind if I take a stab at adding javadoc as I try out
these classes? Are these classes going to be part of some upcoming
factorization?
thanx - ray
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com
This email sent to [EMAIL PROTECTED]
