Tim Ellison and I were discussing about the component model to use in harmony in the thread "[arch] VMCore / Component Model". The component model should be some mechanism to load components, resolve dependencies between the components and provide some versioning mechanisms for the components and their interfaces to ensure forward/backward compatibility. I have posted a proof-of-concept implementation to the wiki which uses shared libraries and tables of function pointers to provide this functionality. [http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/200509.mbox/[EMAIL PROTECTED]
Rana Dasgupta clarified some points about VM accessors in "[arch] VM/Classlibrary Interface ( VM Accessors )". He defines them as a VM access toolkit which could be used by the class library or JVMTI agents. He started to define an initial set of candidates for standardization. He also mentioned that restricting access to these classes might result in better performance of these accessors. [http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/200509.mbox/[EMAIL PROTECTED] In the thread "[Arch] Class unloading and VM objects reclaim", Robin Garner explained how this is done in JikesRVM with MMTk: "where the classloader allocates its objects is essentially configurable". He then explains more details about how this is done. Peter Edworthy replied to this explaining where he thought OS could be used to assign seperate memory areas to the parts of the VM and then do acces checks at the OS level. In the same thread, Tim Ellison replied to Archie Cobbs that he agrees that "All identical string literals must refer to the same instance" and that a pure Java implementation of String.intern() would be acceptable. [http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/200509.mbox/[EMAIL PROTECTED] [http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/200509.mbox/[EMAIL PROTECTED] Chris Elford started the thread "[Arch] Designing in support for JVMTI in Harmony" where he explains what he expects from harmony's implementation of the JVMTI specification. Weldon Washburn has posted two interfaces between the VM and the GC in the thread "[arch] Modular JVM component diagram" and the wiki. [http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/200509.mbox/[EMAIL PROTECTED] [http://mail-archives.apache.org/mod_mbox/incubator-harmony-dev/200509.mbox/[EMAIL PROTECTED] Last week Mark Wielaard suggested to exdend "This week an harmony-dev" to also cover some of the other mailinglists of the sister projects, so I started reading some other mailing lists. It's really interesting what goes on in these projects, but it is quite a lot of traffic and I don't think I can give a valualbe summary in this series. These discussions can't be summarized in a single paragraph per sister project, and everything more would be too much for this series. Maybe a solution would be that these projects try to find somebody who summarizes their discussions (if they want it), and all these summaries are then aggregated at http://planet.classpath.org so we have a single point where we can find them. I hope you understand that I don't do this, I really enjoy writing these summaries for Harmony but it would be too much work for me to write about several other projects too. Regards, David. -- Read the archive of this series at http://deltalabs.at/ -- RSS feed: http://deltalabs.at/?q=taxonomy/term/8/0/feed -- Also aggregated at: http://planet.classpath.org/ -- David Tanzer, Haghofstr. 29, A-3352 St. Peter/Au, Austria/Europe http://deltalabs.at -- http://dev.guglhupf.net -- http://guglhupf.net My PGP Public Key: http://guglhupf.net/david/david.asc -- Pinky, Are You Pondering What I'm Pondering? I think so Brain, but how do we get a pair of Abe Vegoda's pants?