Re: Import of curved connectors from OOXML

2023-11-11 Thread Regina Henschel
Hi Miklos, I have thought about it again. I think, we need nothing new in ODF. We can decide whether to use our routing or the OOXML routing from the svg:d attribute. If the svg:d attribute is missing, we use OOXML routing. MS Office does not write the svg:d attribute, when it exports to odp.

Re: Import of curved connectors from OOXML

2023-11-09 Thread Miklos Vajna
Hi Regina, On Wed, Nov 08, 2023 at 02:38:03AM +0100, Regina Henschel wrote: > If we are in running LibreOffice we could distinguish two ways by a new > constant in enum class SdrCompatibilityFlag, for example. But how to save > such information to ODF? > Keep path definition as vague as it is no

Re: Import of curved connectors from OOXML

2023-11-07 Thread Regina Henschel
Hi Thorsten, hi Tibor, hi all, Thorsten Behrens schrieb am 04.11.2023 um 13:52: Hi Regina, Regina Henschel wrote: It is not an ODF problem. Oh ok - then what I implied, was instead of trying to emulate the curvedConnector3 connector by mapping it to our existing ones, why not implement the m

Re: Import of curved connectors from OOXML

2023-11-04 Thread Thorsten Behrens
Hi Regina, Regina Henschel wrote: > It is not an ODF problem. > Oh ok - then what I implied, was instead of trying to emulate the curvedConnector3 connector by mapping it to our existing ones, why not implement the missing drawingml ones natively? Cheers, -- Thorsten signature.asc Description:

Re: Import of curved connectors from OOXML

2023-11-03 Thread Regina Henschel
Hi Thorsten, Thorsten Behrens schrieb am 03.11.2023 um 20:07: Hi Regina, Regina Henschel wrote: I could try to approximate the OOXML curve by defining ersatz adjustment values. What criteria should be used for such an approximation? Anyway, it is mathematically complex, especially for cases wi

Re: Import of curved connectors from OOXML

2023-11-03 Thread Thorsten Behrens
Hi Regina, Regina Henschel wrote: > I could try to approximate the OOXML curve by defining ersatz adjustment > values. What criteria should be used for such an approximation? Anyway, it > is mathematically complex, especially for cases with two or three handles. > Sounds like a good reason then t