Jeff Duke here from 'Problem with SetFullState and Datastore residing in NVO objects in PB 11.2' issue -
I tried to reply to your post, my original post, direct communications and I keep getting error on page and cannot get in.
Sent an email reply to the noreply - just in case, but....
Not sure if it's settings on our proxy or something on your side as it only seems to affect replies - I can still create discussions....
the info you requested is below - if you can add it to the discussion for me that's great.
I am heading out in a few min and will not be able to respond again until Monday.
Have a great weekend.
VR,
jeff
Hi Chris,
Answers as follows:
1) Version of MS-Windows you are using? – Win 7
2) Build of PB you are using? Build 8669
3) Did this problem just start happening? Yes and no – I recently found the issue with the ‘out of sync’ between the DS and DW when paging backwards and added the setfullstate to ensure the DS had the correct data for the update. Since doing this I have not been able to do a retrieve and see data after an update without shutting down the app and restarting it. It is acting like it dropped the handle to the NVO or it destroyed the DS, but ???
4) Any recent changes to your application's DWO's? – see #3
5) Could this be a problem of duplicate DWO's from a temporary PBL in the library path? – no, we do sometimes us a ‘test’ pbl to try new things before saving them in source control, but there is currently nothing in that PBL – the app has separate PBL’s for the different object groups and each has its own distinct set of objects. The hierarchy is maintained to facilitate the inheritance.
6) Any other relevant information around this issue? Our app is an MDI environment using 1 main window with multiple user objects on it that can also have multiple user objects on them. Everything worked (except they were out of sync when paging backwards) before I did the setfullstate from the DW back to the DS. Normally, the DS shares data with the DW so to do the update I am using the DW to update the DS, which is backwards. I set up the paging this way initially because the retrieves can be prohibitively long as you get higher in the page numbers. The screen scraper has to retrieve each page by starting at page 1 and then paging forward in the mainframe to get to the correct page – the higher the page number longer the time to get the data.