some old fm books are full of stupid iso-8859-1 and such

[douglas]

and when these get imported into booki, they break, because booki wants utf-8. At present, as a workaround, these books are fixed in a slightly hacky way as they leave booki to go to objavi. That means they are broken when you look at them inside booki. It would be better for them to be fixed in booki-twiki-gateway.cgi and/or booki import, and then we'd never have to think about character encoding ever again.

discussed on irc a few days ago:

21:04 < Aco> ah yes.. we had such terrible fail with not encoding a lot of strange unicode characters
21:04 < Aco> funny quotes and stuff
21:04 < eset> yeah
21:04 < eset> whacky
21:04 < Aco> i kind of fixed it now, but it is not final
21:04 < eset> nice to have doen tho
21:05 < Aco> i should probably do it after the save. not now, during the export.