[Checkins] SVN: zope.component/branches/wosc-test-stacking/NOTES.txt Update todo list
Wolfgang Schnerring
wosc at wosc.de
Thu May 19 03:00:11 EDT 2011
Log message for revision 121730:
Update todo list
Changed:
U zope.component/branches/wosc-test-stacking/NOTES.txt
-=-
Modified: zope.component/branches/wosc-test-stacking/NOTES.txt
===================================================================
--- zope.component/branches/wosc-test-stacking/NOTES.txt 2011-05-19 06:46:02 UTC (rev 121729)
+++ zope.component/branches/wosc-test-stacking/NOTES.txt 2011-05-19 07:00:10 UTC (rev 121730)
@@ -1,16 +1,12 @@
TODO:
-- Tests for zope.component-Stacking
- Registry-Object: new registration, delete, view old, bases chain
- getGlobalSiteManager()
- getSiteManager / setSite
+- persisting/pickling stackable registries
- - stackable.reset() needs to take into account "dying" stackables
+- Tests for zope.component-Stacking: can we run the existing tests on a push
+ and then pop "level"? (Maybe use another existing application as a
+ cross-check, too).
- - Do we have to make _v_subregistries stackable?
-
-
- zope.interface uses C-code for some parts. Is it a performance problem that
'stackable' is Python-code? (This should only apply to tests, since we're
going to want to create a way to bypass the stackable stuff)
@@ -19,8 +15,16 @@
- Needs a better name
- Create separate egg?
+- stackable.reset() needs to take into account "dying" stackables
+
- don't create a new class object for each stackable() call
- prettier class name, repr, etc.
- have a name or "stack context", so you can say push('zope.component')
-- do we leak memory since we never unregister stackables?
+- do we leak memory regarding unregistering of stackables?
+
+
+
+* later
+- Do we have to make _v_subregistries stackable, i.e. do we want/need to
+ support changing __bases__ and restoring that after pop?
More information about the checkins
mailing list