I just committed a SoftReferenceCache implementation in utils and it is used by InMemoryCache. So far testing it looks pretty good. The point of a SoftReferenceMap is probably obvious to prevent (as much as possible) OOME. This is fine IMHO cause it's more important to have the application running an not crash due to large potentially large caches. Of course if a cache template is gc-ed it will be re-cached on the subsequent attempt to use it.
I would appreciate any early feedback. Br's, Marius On Apr 6, 1:00 am, Derek Chen-Becker <[email protected]> wrote: > I agree. Thread safety would be nice, and should be easily achievable with > some existing code. The beauty of traits is that we can get these orthogonal > behaviors through composition. > > Derek > > On Sun, Apr 5, 2009 at 2:21 PM, Timothy Perrett > <[email protected]>wrote: > > > > > Just taken a look over the code - looks pretty cool! > > > I like your ideas for ConcurrentHashMap - all sounds pretty awesome... > > regarding the use of EHCache, I rekon as long as provide a hook > > mechinism into the cache system, then sure, we should let people worry > > about those issues in there specific implementation as 90% of users > > simply wont need that functionality IMO. > > > Cheers, Tim > > > On Apr 5, 8:53 pm, Timothy Perrett <[email protected]> wrote: > > > Wow, derek you must be watching Github like a hawk haha ;-) > > > > Just to bring an off list convo between myself and Marius onto the list, > > are > > > we looking at having some generic caching infrastructure in lift? This > > would > > > be great re the localization / translation stuff im working on which > > > currently uses KeyedCache in lift-util as a base. > > > > Im just doing a git pull for the cache branch... > > > > Cheers > > > > Tim > > > > On 05/04/2009 20:38, "marius d." <[email protected]> wrote: > > > > > Good :) ... I was also thinking on a flush-able caching mechanism. So > > > > far the InMemoryCache is more for exemplification as it is not yet > > > > thread safe. It is based on LRU cache but I'm also thinking to also > > > > combine the ConcurrentHashMap approach with LRU ... also I was > > > > thinking to a SoftReference map (as I implemented this once in Java) > > > > but we'll see. It's evolving. I'm not sure at all if we should use > > > > EHCahe or JBossCahe like solutions as we don't really need distributed > > > > caching. Basides if people will want this they can just plugin their > > > > own caching mechanisms. > > > > > Br's, > > > > Marius > > > > > On Apr 5, 10:24 pm, Derek Chen-Becker <[email protected]> wrote: > > > >> I just saw the commit from Marius on a preliminary template cache and > > it > > > >> looks good. I was thinking the same thing in terms of where and how to > > hook > > > >> it. I think that there are some possibilities for some more > > functionality on > > > >> the TemplateCache trait, including a programmatic flush (in case > > you're > > > >> pushing new template files in production and want to force > > re-fetching, for > > > >> instance), but I like the basic concept. > > > > >> Derek --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Lift" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/liftweb?hl=en -~----------~----~----~----~------~----~------~--~---
