The Pivot Table / DataPilot styles, which are automatically created upon DataPilot creation, are not localized / translated. This means if a DataPilot document is interchanged with a user at a different user interface language, custom DataPilot style settings are lost and DataPilot styles get duplicated. How to reproduce: 1. Create Calc document with DataPilot. 2. Choose a DataPilot styles and change one of the settings (e.g. change font color). 3. Save and close document. 4. Go to Tools -> Options - Language Settings and change the User Interface language. 5. Close LO and restart. 6. Open formerly saved document with the DataPilot. 7. Open Styles (F11) and notice the names of the styles didn't change and don't fit to your current user interface. 8. Right click somewhere in the DataPilot and choose "Refresh" from the context menu. 9. Notice: a) your custom style setting is lost b) in Styles is now a new ("duplicate") set of DataPilot styles which names fit to your user interface language.
See also bug 44762 (DataPilot naming of styles and request for more styles)
Thanks for bugreport Please, attach document that demonstrates this problem
Created attachment 61716 [details] Spreadsheet with a PivotTable, which was saved under english and german locale (In reply to comment #2) > Thanks for bugreport > Please, attach document that demonstrates this problem HTH
Thanks for attachment Reproduced in 3.3.4 and 3.5.3 on Fedora 64 bit Steps to reproduce: 0. Open attachment 1 [review]. Enable Stylist pane Expected: All styles are with current UI language Actually: One half is on German, other half on English 2. Modify some German named style, for example Datenpilot Titel, change cell background 3. Right click somewhere in the DataPilot and choose "Refresh" from the context menu. Expected: DataPilot formatting remains unchanged, no additional styles appear Actually: DataPilot formatting lost, some additional styles appear Changing version to 3.3.4 as most early reproducible
@ Andras What do You think about this bug?
This has one gotcha. First of all, I have these pivot table styles and I would LOVE to get rid of them. But besides that, if we ever decide to keep them for the time being and decide to translate them, then we'd have to make sure that the implementation would continue to use the non-translated versions. Or else it would break if the document is shared between different locales.
Regarding your new idea/proposal on pivottable design, I replied in the other bug 44762. Regarding this one: I don't quite get your comment: (In reply to comment #6) > But besides that, if we ever decide to keep them for the time being and decide > to translate them, then we'd have to make sure that the implementation would > continue to use the non-translated versions. AFAIK, the style-names are already translated, as they are available in every locale. Second, there is already an "engine" for style-translation which applies to all other styles (Default, Result, Heading, etc.). So my question is, why is this "engine" not apllied to the pivot-table stles? Wouldn't this be the simplest solution (for the time beeing)? > ... Or else it would break if the document is shared between > different locales The current implementation *IS ALREADY* broken, when shared between different locales, so no worry about breaking things ;-) ...
(In reply to comment #7) > Regarding your new idea/proposal on pivottable design, I replied in the > other bug 44762. > > Regarding this one: > I don't quite get your comment: > (In reply to comment #6) > > But besides that, if we ever decide to keep them for the time being and decide > > to translate them, then we'd have to make sure that the implementation would > > continue to use the non-translated versions. > AFAIK, the style-names are already translated, as they are available in > every locale. > Second, there is already an "engine" for style-translation which applies to > all other styles (Default, Result, Heading, etc.). > > So my question is, why is this "engine" not apllied to the pivot-table stles? > Wouldn't this be the simplest solution (for the time beeing)? You are asking the wrong guy. I didn't design it, so no idea "why". Maybe you can see if that's indeed the simplest solution. Do you need a code pointer for this? > > ... Or else it would break if the document is shared between > > different locales > The current implementation *IS ALREADY* broken, when shared between > different locales, so no worry about breaking things ;-) ... That's lame. It's already broken, so who cares if you break a little more doesn't give me confidence.
(In reply to comment #8) > > > ... Or else it would break if the document is shared between > > > different locales > > The current implementation *IS ALREADY* broken, when shared between > > different locales, so no worry about breaking things ;-) ... > > That's lame. It's already broken, so who cares if you break a little more > doesn't give me confidence. And that's even another reason why we should get rid of this. It's broken beyond repair.
(In reply to comment #8) > (In reply to comment #7) > > Second, there is already an "engine" for style-translation which applies to > > all other styles (Default, Result, Heading, etc.). > > > > So my question is, why is this "engine" not apllied to the pivot-table stles? > > Wouldn't this be the simplest solution (for the time beeing)? > > You are asking the wrong guy. I didn't design it, so no idea "why". You wrote: "I have these pivot table styles", so I thought you might know this. Also this was meant as a constructive question. I'm not trying to blame anyone for this history-grown code - sorry if it sounded otherwise. > Do you need a code pointer for this? I'm user doing some Q&A, I have no experience whatsoever with the LO-code. However if you think it is worth/promising to try - sure go ahead. Another option might be to name me the "right" guy (if you know) and I will try to contact & point him to this bug. Maybe he can give some useful insights. > > The current implementation *IS ALREADY* broken, when shared between > > different locales, so no worry about breaking things ;-) ... > > That's lame. It's already broken, so who cares if you break a little more > doesn't give me confidence. Judging from your comment #6 one could get the impression, that by trying to fix this, things might broke which worked before. I just wanted to clarify this, also notice the smiley.
(In reply to comment #10) > (In reply to comment #8) > > (In reply to comment #7) > > > Second, there is already an "engine" for style-translation which applies to > > > all other styles (Default, Result, Heading, etc.). > > > > > > So my question is, why is this "engine" not apllied to the pivot-table stles? > > > Wouldn't this be the simplest solution (for the time beeing)? > > > > You are asking the wrong guy. I didn't design it, so no idea "why". > > You wrote: "I have these pivot table styles", That's a typo. I HATE is what I meant to type. > Judging from your comment #6 one could get the impression, that by trying to > fix this, things might broke which worked before. Yes. You could say that. In short, 1) This mis-feature is broken by design, and as it stands today, it already causes issues in several areas which cannot be fixed without removing this mis-feature. 2) No point improving it because of 1).
If I can expand on this a little. In the future we should remove this and implement much more flexible custom cell properties inside pivot table, combined with some sort of predefined styles that can be applied per table. The current system allows uniform styles to ALL tables, which isn't extensible, conflicts with these new proposed features, not to mention it causes severe interop problems.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.