Drew McCormack
"You should use Core Data" by @atcrawford
<<< Great posthttp://pocket.co/sjIZL
Marcel Weiher
Michael Tsai
Milen Dzhumerov
@drewmccormack It’s possibly convenient, but not high perf, to bring all objects into RAM before operating on them.
-
-
@mjtsai It's always much faster operating on in-memory objects than on disk objects. In many cases it's a big perf win. - View other replies
-
@drewmccormack And an SQLite index on disk can be way faster than Core Data’s in-memory predicate filtering. -
@mjtsai@drewmccormack Core Data predicates are translated to SQL queries. -
@atomicbird Right. You have to bring the objects into memory to modify them; then, until you save, you get slow in-memory filtering.
-
-
-
@mjtsai Squeak does it. 1.2M objects in 500ms. My XML parser also does 1M objects in ~800ms (Mac). Better with a *good* binary rep. -
@mpweiher Right. Core Data is probably at least an order of magnitude slower, and its objects are heavier. -
@mjtsai Yeah, leaning strongly towards 2 orders. 1KB/obj when writing, order of magnitude on read access. Dict fetches OTOH are FAST!
-
Tom Harrington
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.