ayanova/devdocs/todo.txt

@@ -12,14 +12,24 @@
 
 Aug
 
+	Migrate problems to alleviate
+		pm's still in the midst of generating when migrate starts so job is not stopped and it keeps churning
+			need way to stop active jobs before erase db and then migrate happens
+	lastentrydate issue
+		one issue is that the error message from the inventory adjustment is returned in teh workorderitem part or po operation which is confusing as it doesn't 
+			clearly state inventory adjustment anywhere so it seems to be nonsensical
+			maybe it can be put in a group of errors with a collection name regarding inventory adjustment??
+		Not sure how this is happening unless something is happening simultaneously elsewhere (pm generation still going??)
+			except it happened with a po and at that point no pm's have yet been migrated, so it seems like it's self intrinsic, maybe a really slow op of some kind
+			or something happening in parallel during migrate??
+		Retry sb in effect during migrate, if fails, retries at least once again before bailing
+			could do at both client end and at server for things like inventory if it fails, i.e. retry inventory once before bailing??
 	
 	Meter reading		
 		todo / outstanding
 			must migrate as well
 			NOTE: need to make a switch during migration v7 has workorderitem id but v8 links to workorderitemunit id so need to fixup that as well
-					
-			notification of multiple however the fuck that works
-			test notification
+		
 		
 	Cases from Joyce
 	prevent login during migration except for superuser
@@ -859,3 +869,10 @@ MID CENTURY MODERN TUNES - https://www.allmusic.com/album/ultra-lounge-vol-14-bo
 BUILD 125 CHANGES OF NOTE
 
 - case 3918 (no change but have appended explanation please close if you agree or append more info to be considered)
+- Migrate issues
+	Updated workorderbiz at server to provide more detailed error messages when it calls inventory processor and fails
+	Updated PM generator loop to recognize if server state is changed to migrate and will bail out and stop processing
+		this is in addition to it already being coded to not call it in the first place during migrate, just to cover a scenario where it's alerady started doing them when migrate starts
+	Updated v8migrate plugin to route all api calls through new methods that will try 3 times before giving up on any api call to work around transient communications problems
+	improved error messaging to show more details of error and cleaned up output
+	Fixed bug where workorderitempartrequestid was not being correctly set in purchaseorderitem that serviced it