Re: [1/4] isis git commit: ISIS-1137: Now cache a ProxyFactory instance by class (allowing reuse of the enhanced class for all instances of that type) rather than creating a newly enhanced class as cu

On Fri, Apr 24, 2015 at 1:15 PM, danhayw...@apache.org wrote: Repository: isis Updated Branches: refs/heads/master 69d559486 - ee834d9a3 ISIS-1137: Now cache a ProxyFactory instance by class (allowing reuse of the enhanced class for all instances of that type) rather than creating a

Re: [1/4] isis git commit: ISIS-1137: Now cache a ProxyFactory instance by class (allowing reuse of the enhanced class for all instances of that type) rather than creating a newly enhanced class as cu

Hi Martin, thx for the input. I did consider this, but realized that I don't know about the interaction of class(un)loading vs GC. My understanding (correct me if I'm wrong) is that GC daemon will quite eagerly clean up any objects if there are only weak references to that object. So I was

Re: [1/4] isis git commit: ISIS-1137: Now cache a ProxyFactory instance by class (allowing reuse of the enhanced class for all instances of that type) rather than creating a newly enhanced class as cu

Hi Dan, On Fri, Apr 24, 2015 at 3:13 PM, Dan Haywood d...@haywood-associates.co.uk wrote: Hi Martin, thx for the input. I did consider this, but realized that I don't know about the interaction of class(un)loading vs GC. My understanding (correct me if I'm wrong) is that GC daemon will

Re: [1/4] isis git commit: ISIS-1137: Now cache a ProxyFactory instance by class (allowing reuse of the enhanced class for all instances of that type) rather than creating a newly enhanced class as cu

OK, cool. As you saw, I've committed that change to use weak keys. And thanks for these hints on the JVM args; I'll add them to our new documentation. Cheers Dan On 24 April 2015 at 13:42, Martin Grigorov mgrigo...@apache.org wrote: Hi Dan, On Fri, Apr 24, 2015 at 3:13 PM, Dan Haywood