Lennart Poettering wrote: > At two places: > get_service() > and in: > pa__init() Attached patches to try and fix it ... the utf8 stuff is "borrowed" from glib. I haven't actually tried it with UTF-8 data, but it should be ok. I saw that the utf8 functions in pulse came from glib so I thought it'd be ok to copy some more from there. > (And while we're at it. Avahi's current implementation of > avahi_alternative_service_name() which generates an alternative > service name when a name conflict happens and basically just attaches > " #2", " #3" to it, generates a 66 character string when you pass a 63 > character string to it. Which is obviously a bug in Avahi. I've posted > this as bug #151 in avahi BT now.) I tried looking through avahi, but it seems that the function used in avahi counts chars using strlen() which isn't utf8 aware either. Either way, avahi could also use something like the first patch to count UTF8 chars properly. But this is just with a quick look :) I could be way off. Cheers, R?mi -------------- next part -------------- A non-text attachment was scrubbed... Name: pa-1-utf8.patch Type: text/x-patch Size: 2020 bytes Desc: not available URL: <http://lists.freedesktop.org/archives/pulseaudio-discuss/attachments/20070816/bb8f77cf/attachment.bin> -------------- next part -------------- A non-text attachment was scrubbed... Name: pa-2-truncate.patch Type: text/x-patch Size: 1630 bytes Desc: not available URL: <http://lists.freedesktop.org/archives/pulseaudio-discuss/attachments/20070816/bb8f77cf/attachment-0001.bin>
