My suspicion, despite Pasi's statement to the contrary, is that something in the environment affected the output of the report. However, I don't know of anything that would automagically convert UTF-8 to ISO8859-1. It could be that Pasi had ISO data in the database.
Since the place where Pasi was (is?) employed at the time has moved on to Koha, I think we'd need someone to independently verify the bug again. I am setting this bug to incomplete and removing the pullrequest tag pending that confirmation. Perhaps, Ben Shum can let us know why he set the bug to confirmed.
My suspicion, despite Pasi's statement to the contrary, is that something in the environment affected the output of the report. However, I don't know of anything that would automagically convert UTF-8 to ISO8859-1. It could be that Pasi had ISO data in the database.
Since the place where Pasi was (is?) employed at the time has moved on to Koha, I think we'd need someone to independently verify the bug again. I am setting this bug to incomplete and removing the pullrequest tag pending that confirmation. Perhaps, Ben Shum can let us know why he set the bug to confirmed.