Well, we use JCS as an in-memory database. Many of the objects we store don't implement Serializable, because they will only ever exist in memory, never anywhere else. Requiring cached objects to implement Serializable would be a pain for us.
Just some feedback.... Les > -----Original Message----- > From: Aaron Smuts [mailto:[EMAIL PROTECTED] > Sent: Thursday, May 29, 2003 8:46 AM > To: 'Turbine JCS Users List' > Subject: RE: Why a cached object must be Serializable > > > > I assume the problem Hanson encountered is something like this: A > program was using an object retrieved from the cache. > Another instance > of same program or another program modified the object at the > same time. > This probably caused some bugs that were very difficult to > duplicate and > track down. > > Since the cache passes its contents by reference, such errors are very > likely to happen unless they are prevented programmatically. > ----The question is whether this should be done inside or outside the > cache. > > I don't see why this should be the responsibility of the > cache. Though > it should, at least, be highlighted in the documentation. We can come > back to this later. > > > A second problem is that the cache is setup for distribution and > spooling to disk. Both of these features require > serializable objects. > Sometimes people might want to just store non-serializable objects in > memory though. ----The question is whether the cache access methods > should require serializable objects or not. > > > > (I should point out that Hanson authored significant portions of an > early version of JCS and came up with many of the features.) > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: > [EMAIL PROTECTED] > For additional commands, e-mail: > [EMAIL PROTECTED] >
