[jboss-dev-forums] [Design of JBossCache] - Re: Possible optimization of CachedSetImpl

[ben.wang@jboss.com]

I have looked at your patch. They look good. Just two comments here:

1. In putNoMask and getNoUnMask, you chosen to call putObject and getObject, respectively (instead of attach and detach). Any reason for it?

2. On remove(), shifting the POJO has always been expensive. I have been pondering on a way to get rid of that. Actually, it is OK to gap in between now since you have used the hashcode+. So actually, for remove(), we can get the Node.getChildren() and iterate on that child Fqn to the whole set if needed. This way, we don't reply on the specific consecutive odering. What do you think?

-Ben

View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3983615#3983615

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3983615