[elephant-cvs] CVS elephant

ieslick ieslick at common-lisp.net
Sat Feb 17 12:13:19 UTC 2007


Update of /project/elephant/cvsroot/elephant
In directory clnet:/tmp/cvs-serv11213

Modified Files:
	TODO 
Log Message:
Final migration fixes for BDB and restructuring of BDB default transaction handling to allow for nested controllers and transactions; migration info

--- /project/elephant/cvsroot/elephant/TODO	2007/02/16 23:02:50	1.51
+++ /project/elephant/cvsroot/elephant/TODO	2007/02/17 12:13:18	1.52
@@ -9,33 +9,32 @@
 TASKS TO GET TO ALPHA:
 
 Migration:
-- Validate BDB migration 0.6.0->0.6.1 (Ian)
 - Validate SQL migration 0.6.0->0.6.1 (Robert)
-- Improve migration robustness
-  high - resolve store-controller conflicts
-  high - avoid transaction limits in map-btree by adaptive chunking
-  low - Nested persistent objects in standard objects, arrays, etc
 
 TASKS TO GET TO BETA:
 
 BDB Features/Cleanup:
 - Always support locks that timeout?  Tradeoffs?
 - Perform checkpoints (prep for DCM functionality)
-- Verify db_deadlock for other lisps (launch and kill background program I/F)
-- Derived indices fail to re-connect, verify this
+- Derived indices used to fail on re-connect, verify this
 
 Lisp Support:
-- 64-bit lisp verification
 - Win32 builds
   - Windows support for asdf-based library builds?  Include 32-bit dll in release?
+- 64-bit lisp verification
 - Validate OpenMCL 1.1 on Mac OS X
 - Validate Lispworks
+- Verify db_deadlock for other lisps (launch and kill background program I/F)
 
 Stability and Performance:
-- Validate that migrate can use either O(c) or O(n/c) where c << n memory for large DBs
 - Review and address all NOTE comments in the code
 - Review SBCL string serialization performance
 - Improve SQL base-64 serializer performance?
+- Migration: Validate that migrate can use either O(c) or O(n/c) where c << n memory for large DBs
+- Migration: Improve support for nested persistent objects inside standard objects, arrays, etc?
+- Migration: Improve scaling properties of migration so all objects do not need to be resident in memory?
+- Migration: Validate that graph structures with loop are copied properly
+- Migration: Improve printing and informative messages
 
 TASKS TO GET TO FINAL RELEASE:
 
@@ -64,6 +63,7 @@
 ----------------------------------
 
 February 13-17th, 2007:
+x Fix BDB migration 0.6.0->0.6.1, improve nested transaction/store interactions
 x Legacy conversions issue for SQL (and BDB?) due to package rename (both)
 x Figure out how to compact a specific btree and/or key-range using optimize-storage.
   Probably need to update keyword part of the API




More information about the Elephant-cvs mailing list