On Mon, 2011-11-28 at 13:43 +0100, Jean-Baptiste Faure wrote: > > Now, my question is what to do? Is it a good idea to modify > Date::IsValid in tdate.cxx without knowing why it works in LibO 3.3 ?
I can sense that you already know the answer to this question. :-) But I wouldn't modify IsValid() without first checking why the same thing worked in the 3.3 code base. Perhaps the 3.3 chart2 code didn't even call IsValid()? I'm just guessing that, but that's the sort of stuff we would need to find out in order to come up with how to fix this. Now, I know that building the 3.3 branch is not as easy as building master, due to the split repo that it still uses & it lacks many nice features that we've added to our build system over the years. So, if you are willing to build and investigate why it worked in 3.3, great, and you would be more than welcome to do it. If not, then just leave it for now until we find someone (could be me) willing to look into it. Anyway... While I was writing this, I just decided to check the 3.3 code and see what it did back then. In the 3.3 code base, chart2 doesn't even have DateHelper.cxx file. Looks like chart's date handling has changed quite a lot since 3.3... In fact, in 3.3 chart2 doesn't seem to do anything date-specific; must be something that Oracle worked on post-3.3 to handle dates better, which in turn came caused this problem as a result... My conclusion? No idea what advice I should give you to get it fixed just yet. Maybe someone else knows the work that Oracle has done to the chart2 module after 3.3 & give us some background? Kohei -- Kohei Yoshida, LibreOffice hacker, Calc _______________________________________________ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
