Subject: RE: sgml-parse and GC From: Peter Nilsson <pnidv96@xxxxxxxxxxxxxx> Date: Thu, 22 Jul 1999 21:56:33 +0200 (CEST) |
On Thu, 22 Jul 1999, Didier PH Martin wrote: > Didier says: > This is an interesting approach. If I understand you well, the grove itself > is marked for caching or not. All users of that grove object do a addref on > it, if ref is zero we can release it. > Every node has a refcounter. If you don't want to cache the grove, just skip putting it into the has table of DssslApp. Then it will get released when no one use it. The interresting question is *when* it will get released... > I'll work on this idea on paper. The simplest way to test it is to comment out the line: groveTable_.insert(params.sysid, rootNode) in DssslApp::load. Too simple to make a patch, but don't do it in CVS, since the groveIndex wouldn't be correct... Regards, /Peter Nilsson -- '(?P . (?e . (?t . (?e . (?r))))) DSSSList info and archive: http://www.mulberrytech.com/dsssl/dssslist
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
RE: sgml-parse and GC, Didier PH Martin | Thread | RE: About Constructions rules, Didier PH Martin |
RE: sgml-parse and GC, Peter Nilsson | Date | Re: sgml-parse and GC, Peter Nilsson |
Month |