Apache OpenOffice (AOO) Bugzilla – Issue 74895
Forms not redrawn properly when switching betweyen Calc sheets
Last modified: 2013-02-07 22:01:01 UTC
See the screenshots attached to this report. I have an application that uses Forms on mulitple sheets in a Calc document. I switch between Calc sheets to pull up a different Form. The application was developed and works in OOo versions from 2.0.0 to 2.1.x. Sequence -------- 1) When I first start Calc with the document in edit mode the forms are correctly drawn. The first two forms sheets are shown as images "Screenshot-eWater_RBA_Application-split - OpenOffice.org Calc.png" "Screenshot-eWater_RBA_Application-split - OpenOffice.org Calc-1.png" 2) When I click Edit Page to turn edit mode off, I get the first form drawn as shown in "Screenshot-eWater_RBA_Application-split (read-only) - OpenOffice.org Calc.png" 3) Switching to the second form sheet I get a mix of both forms visible as shown in "Screenshot-eWater_RBA_Application-split (read-only) - OpenOffice.org Calc-1.png" 4) When I return to edit mode, the first form (for example) is now drawn incorrectly as shown in "Screenshot-eWater_RBA_Application-split - OpenOffice.org Calc-2.png" 4) If I turn edit mode back on and off again I usually get the same result, but if I reload using File -> Reload (while in edit mode), if it doesn't crash OOo, it appears to improve things considerably when I re-enter run mode. I get the form sheets in the running application drawn correctly, as shown in "Screenshot-eWater_RBA_Application-split (read-only) - OpenOffice.org Calc-2.png" and "Screenshot-eWater_RBA_Application-split (read-only) - OpenOffice.org Calc-3.png" I've also attached the Calc document containing the application although you won't be able to run it of course without the database. Macros do get called when switching from sheet to sheet, but IIRC these are pretty much no-ops when switching between the two sheets shown. Thanks.
Created attachment 43392 [details] Sheet 1 - Edit Mode
Created attachment 43393 [details] Sheet 2 - Edit Mode
Created attachment 43394 [details] Sheet 1 - re-entered edit mode
Created attachment 43395 [details] Sheet 1 - run mode
Created attachment 43396 [details] Sheet 2 - run mode
Created attachment 43397 [details] Sheet 1 - run mode after reload
Created attachment 43398 [details] Sheet 2 - run mode after reload
Created attachment 43399 [details] Application document
Hi, I can't reproduce this. For me everything is drawn correctly. Can you please also attach the database or a sql script to build the database. Thanks and Bye Marc
Additional information and sample documents:- This is not just a redraw problem, because (using the full application with the PostgreSQL backend) the second sheet actually contains fully working form controls from both sheet 1 and sheet 2. I'm running these tests on the following system: HW: 2400MHz, 1GB RAM, i686, Tower PC OS: FC6 Linux with current updates installed. JRE: 1.5.0_09 PostgreSQL backend running locally on this machine. The attached ZIP file with included ods plus odb documents are cut-down versions of the application and database used originally, but they still show some of the effects although nowhere near as severe. E.g. switching to the second sheet appears ok. Note: I have also removed all events from the document and forms to avoid calling any macros. I use Zoom = Entire Page (in case that matters for redrawing). 1) Register the included odb as a data source. 2) Open the ods document (should open in edit mode and appear ok). 3) Turn edit mode off to enter run mode. 4) I see the first sheet as shown in the screenshot "Screenshot-eWater_RBA_Application-OOo2.2-Bug (read-only) - OpenOffice.org Calc.png" 5) Switching to the second sheet appears ok. 6) Switching back to the first sheet and re-entering edit mode appears as shown in:- "Screenshot-eWater_RBA_Application-OOo2.2-Bug - OpenOffice.org Calc.png" Additional information re the original issue description -------------------------------------------------------- The original full application connects to a PostgreSQL database via the current version of the PostgreSQL-SDBC driver (0.7.4). If I change the document included in this zip file to use that datasource, I get the original behaviour as described. Possible causes --------------- Because I see a much reduced but still residual effect using the Base HSQL database, the effect doesn't appear to be backend or driver related directly, except that I'm beginning to suspect it may be a timing issue. Because, in the full application, the second sheet contains controls from both sheets 1 and 2 I suspect that, in the cut-down version, the doubled up table control shown in point 4 above may actually be a second instance of the same table control. If you have a POstgreSQL backend available I can probably send you the reduced schema which still shows the full effect. Hope this all helps. Thanks.
Created attachment 43476 [details] Sample document with additional information
Oh, and in case it's not clear to anyone reading this for the first time, the images referred to in my description earlier today are those contained in the attached Zip file, which may have the same names as some of the originally attached images. Regards.
Hi, I can reproduce this. In a easy way. 1. create a calc sheet 2. insert a table control ( cancel the wizard ) 3. move the control around ==>> repaint problems with the table control. msc -> fs: I think this is duplicate to a global repaint problem in calc. However can you have a look. Bye Marc
fixed in CWS dba23a find more information about this CWS, like when it is available in the master builds, in EIS, the Environment Information System: http://eis.services.openoffice.org/EIS2/cws.ShowCWS?Path=SRC680%2Fdba23a
fs-> msc: please verify in CWS dba23a
verified in CWS dba23a find more information about this CWS, like when it is available in the master builds, in EIS, the Environment Information System: http://eis.services.openoffice.org/EIS2/cws.ShowCWS?Path=SRC680%2Fdba23a
Hi, this is fixed in the current master. The current master is available at http://download.openoffice.org/680/index.html I close this issue now. Bye Marc
This problem is still present in OOo 2.3. The attached zip from March 1 2007 shows the problem as described previously and in the included screenshots. The attachment is a cut down version of the real application that is in productive use but restricted to OOo 2.1. The real application running in OOo 2.3 continues to display the more severe problems seen in the screenshot attachments of Feb 27 2007. Note: msc's comment on Mar 2 describes a simplified method to reproduce the issue, however I note that this no longer shows the problem in OOo 2.3. It therefore has been fixed and would appear not to have related to this issue after all. I'm currently running OOo 2.3, JRE 1.6.0_2 and Fedora Core 7 Linux if that is relevant. Regards. Ross
resetting target then
putting back into pool, to have better chances of being looked at