https://issues.apache.org/bugzilla/show_bug.cgi?id=46828





--- Comment #3 from Jeremias Maerki <jerem...@apache.org>  2009-03-11 00:19:35 
PST ---
Thanks for the patch, Dario. I agree with Andreas that the code in
AbstractLayoutManager should really be in BlockStackingLayoutManager. I can
faintly remember that I've once investigated the right time to release memory
in a layout manager. I think that was for page-number-citation-last. I'm not
sure that we already always know exactly when a layout manager is really
finished. Anyway, we've got a good test suite so any bugs in that area should
quickly be found.

@Andreas: what does "beneficial" mean? Beneficial in terms of memory
consumption or of speed? I think that can be mutually exclusive. Before we
decide to enable this by default I would want to have some numbers (performance
& memory usage) to base the decision upon.

Another comment to the patch: I'm not particularly happy about calling the
temporary area tree unloading to disk simply "the cache". We have a font cache,
an image cache and others. I'm not even sure if this is really a classic
"cache" as such. We need a better name for this IMO. How about
FOUserAgent.setTemporaryAreaTreeSerializationEnabled(boolean)? Or
FOUserAgent.setDiskBasedRenderPagesModel(boolean)? A bit lengthy but at least
more speaking. Or FOUserAgent.setConserveMemoryPolicy(boolean) which would be a
less specific approach that other code parts also could use as hint?

-- 
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

Reply via email to