LibreOffice Working On A New Cairo Graphics Back-End

[phoronix]

Phoronix: LibreOffice Working On A New Cairo Graphics Back-End

Merged yesterday into the LibreOffice code-base was introducing yet another graphics drawing back-end for this open-source office suite...

https://www.phoronix.com/scan.php?pa...raphicsBackend

[Slartifartblast]

Sir Rowley Birkin QC explains Cairo..........

[rockiron]

Is this how they are spending donations, instead of improving compatibility with Microsoft Office?

[slacka]

Is this how they are spending donations, instead of improving compatibility with Microsoft Office?

From the article:

Tomaž Vajngerl of Collabora

Looks like Collabora is paying for this. I don't see it as a 2021 TDF tender. No idea if anyone is contracting them for this work.

Donations to TDF mostly go to infrastructure. There was a 2020 budget tender for improving compatibility with Microsoft Office proposed by this same developer. Tomaž has done a lot of great work improving compatibility. But I do agree that TDF ought to use more of it's donations to pay for development. More tenders for improving MSO compatibility would have a big impact on the usefulness of their project.

The question I have is what is this for? Especially considering all the work put into Skia and that remains to be done.

[ⲣⲂaggins]

Of all 2d graphics libraries, why Cairo? Yes, the current maintainers are putting in a herculean effort to keep it afloat, but... have you seen the state of the test suite? Half of it's disabled because nobody can remember how it works. And skia has much better performance.

[quikee]

Oh geez... this is not a "new" cairo backend. The idea is to refactor the existing structure to be more in-line with other VCL plugins, where you can exchange the graphic rendering part with another one (on Windows VCL plugin for example you can change GDI to skia and previously OpenGL backend). The headless VCL plugin was the only one that doesn't have such a structure with the separation of graphic backend (a SalGraphicsImpl class) as a separate part, so this is the work on that, but all it is is just moving the code around. In the future this will actually allow a skia graphic backend to be used instead of the cairo one.

[quikee]

Collabora is paying for this. I don't see it as a 2021 TDF tender. No idea if anyone is contracting them for this work.

Nobody is paying for this. Developers also contribute some of the free time to work on LibreOffice.

Donations to TDF mostly go to infrastructure. There was a 2020 budget tender for improving compatibility with Microsoft Office proposed by this same developer. I can't find any info as to whether is was accepted by TDF and completed. But I do agree that TDF ought to use more of it's donations to pay for development that would have a big impact on the usefulness of their project.

There will be TDF funded work starting soon. For MS compatibility I think we might see funded implementation of SmartArt this year. If you want to see what has been accepted by the TDF members, you have to look at tenders [1].

[1] https://blog.documentfoundation.org/tenders/

The question I have is what is this for? Especially considering all the work put into Skia and that remains to be done.

Refactoring the structure of the headless VCL plugin to get it inline with the structure of other VCL plugins. This will make the code simpler and allow other graphic backends to be used in headless (with more work).

[quikee]

Is this how they are spending donations, instead of improving compatibility with Microsoft Office?

No, just free time.

[smitty3268]

Oh geez... this is not a "new" cairo backend. The idea is to refactor the existing structure to be more in-line with other VCL plugins, where you can exchange the graphic rendering part with another one (on Windows VCL plugin for example you can change GDI to skia and previously OpenGL backend). The headless VCL plugin was the only one that doesn't have such a structure with the separation of graphic backend (a SalGraphicsImpl class) as a separate part, so this is the work on that, but all it is is just moving the code around. In the future this will actually allow a skia graphic backend to be used instead of the cairo one.

Thanks for the explanation, that makes a lot more sense than what Michael's article seemed to be saying.