Well, it didn't take me that long to do, and if we aren't just going to remove the deprecated method, I wanted to be able to sleep at night. :)
If we don't want to include code like this then I think we should just remove the deprecated method, because even if we remain binary backwards-compatible for one of these folks, the caching module won't work for them properly (they'll get UnsupportedOperationExceptions) and they'll have to upgrade to the new API anyway. I'd rather have that discovery happen for them at compile time than at runtime, to be frank. So I guess I'm coming around to: I think we should remove this method which only appeared publicly in a beta release and for which I think there is a high probability there are exactly zero people who will actually have stuff break if we remove it. If folks are happy with that, I'm happy to rollback the bend-over-backwards-for-binary-compatibility atrocity I just committed. :) Jon ........ Jon Moore Comcast Interactive Media On 1/11/11 3:07 PM, "Oleg Kalnichevski" <[email protected]> wrote: >On Tue, 2011-01-11 at 17:28 +0000, [email protected] wrote: >> Author: jonm >> Date: Tue Jan 11 17:28:20 2011 >> New Revision: 1057745 >> >> URL: http://svn.apache.org/viewvc?rev=1057745&view=rev >> Log: >> On the off chance someone wrote a custom serializer for use with the >> 4.1-beta version of HttpCacheEntry, we want to make sure things still >> work properly. The current implementation of HttpCacheEntry already >> throws an UnsupportedOperationException if you try to get at the variant >> map but the old set-based data structure is there; now we handle this >> exception, log a warning, and then use the old structure. This is a bit >> less efficient but still demonstrably works (I added a unit test to the >> Ehcache suite, where you can submit a custom serializer, that shows that >> the implementation still conforms to all the MUST/MUST NOTs). >> > >Jon, > >Actually I think this is really an extreme case and the fix was >unnecessary. > >Oleg > > > >--------------------------------------------------------------------- >To unsubscribe, e-mail: [email protected] >For additional commands, e-mail: [email protected] > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
