Text is not correct after PDF export

Question

I have a strange mistake after PDF export. If I select “vectorized text” all text look (ursuprisingly) cornered. If I use  “normal text with embedded fonts” A PART of the PDF is correct, a part is not, though I use the same font (Dialog). Even more strange: I can not reproduce it properly. Sometimes the part in the purple box below is correct, sometimes it is not. I tested so many things I already feel like in a lab.

My workaround: Export to .svg and print to .pdf, but what is the problem?

(The file is huge. I cannot upload it right now. I can do that later, if there isn’t a simple answer.)

Answer 1

I take it the problem is the slightly cut off first numeral '5' in the group label.

Unfortunately, there is indeed no simple answer as to why that happens. And similar to the SVG problem in Why does .svg export change positions and fonts and text from bold to normal? this problem might be related to the PDF viewer application as well. Again, the first step would be comparing how different viewer applications display the file. (It is important to always include Adobe Reader in such comparisions, because that is the de facto reference implementation for PDF viewer applications.)

Similar to the SVG post mentioned above, I would need the GraphML file, the version of the yEd instance you used to produce the PDF, and the version of Java you are running on to be able to investigate this problem further.