Thanks, Marcel and Graham. It answers my question. I will now look for ways to optimize data transmission between processes, and I think I have already found one.
On Thu, Apr 16, 2009 at 3:10 AM, Marcel Weiher <marcel.wei...@gmail.com> wrote: > > On Apr 12, 2009, at 23:26 , Oleg Krupnov wrote: > >> I haven't tried either of the methods I mentioned so far (because I'm >> lazy, sorry:), but what I have tried is I created NSData from the >> large object by using NSKeyedArchiver. It has taken forever, so that I >> had to force-quit the process. > > Yes, NSKeyedArchiver is not really suitable for large data-sets. > >> That's why I am asking if the same >> thing is going to happen with piples and DO? Or maybe I'm doing >> something wrong? > > Well, the two questions are really orthogonal: both DO and pipes will > require serialization if you actually want to move the object across. > >> >> By "large" I mean about 10-100 MB. I don't think I am going to hit the >> boundary of VM anyway. > > Can you tell us a bit more about the object in question? Is it really one > object with 10-100MB of associated data? What's the data and how is it > encoded? Or is it an object graph with lots of objects involved? Why do > you want to move the object/data across to another task? > > Marcel > > _______________________________________________ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
