Re: PDF image import

Hi Otfried,

Thanks for the quick reply!

>> I have tested the epspdfipe.lua Ipelet[0], which uses pdftoipe
>> internally, but pdftoipe doesn't work reliably enough (though
>> surprisingly well) to convert my test PDFs into a truthful
>> representation in Ipe.
>
> What doesn't work well?  Some problems might be very easy to fix if you
> file a but report with example PDFs attached.
>
> However, anything related to text is a problem.  Ipe's text model is
> just too different from PDF text, as it relies entirely on Latex.

It is only partially related to text, I will file a bug report with
the example files.

> There are no plans to add PDF import to Ipe now.   While it would be
> relatively easy to support the equivalent of \includegraphics, where you
> would see the included image in the PDF output but nowhere else,  most
> people who ask for this feature really want to see the included image on
> the Ipe canvas.

I wonder whether a hybrid approach is possible, as I don't think one
needs to edit the PDF image in Ipe: For the display on canvas one
could quickly (upon import) calculate a rasterized "thumbnail" of the
PDF, which in the final PDF output is replaced by the actual scalable
image. This way one wouldn't need a PDF interpreter. But I can of
course not judge whether this needs large changes in Ipe.

> And that would imply adding a complete PDF interpreter
> to Ipe, which I have no intention to do.

Can't poppler-qt4 provide all the necessary bits to do that easily?

Cheers,
Frank