Yes strange, probably my test image.

Or you could just be on a different platform, the Clipboard VM source for Linux/Windows is� something else entirely.

On OSX, it asks for the text in UTF8 specifically; 

ExternalClipboard >> clipboardText

...

bytes := self primReadClipboardData: clipboard format: 'public.utf8-plain-text'. (what�s the valid formats other than that you ask? Darned if I know, probably need to read the VM sources)

On Unix/Windows, it works differently, you don�t use a primitive where you specifically set the encoding you want for clipboard text entering the image.

My memory is a bit hazy, but IIRC, the Windows primitive code will also return utf8 consistently, so the code in Clipboard >> clipboardText will work correctly.

While on *nix, the bytes returned will be converted either to the current locale�s default encoding (which is why I asked for it), or the one specified by the vm�s -locale parameter when it was launched, I can�t quite remember the details.

So if you�re on Linux it *might* help if you start the VM with an -encoding UTF8 parameter set, or use a system locale which returns #utf8.

It�s not rendered in reverse order, when one does the same as Sven did with example text #2 (�Ready to continue?�), the String will render in Pharo with the question mark at the end.

Cheers,

Henry