diff libpurple/ft.c @ 21121:35b4f1dc4c8d

replace most calls to strerror with calls to g_strerror. strerror will return a locale-specific string in the locale-specific encoding, which isn't guaranteed to be UTF-8. g_strerror will always return a UTF-8 string. I left gg and zephyr untouched, since gg doesn't include glib headers yet, and zephyr does something weird with a #define for strerror. Someone more familliar with those should take a look. And the win32 guys should check and see if I screwed something up, since they had strerror #defined to something else. This should fix #2247 (and maybe some mystery crashes)

--- a/libpurple/ft.c	Sat Nov 03 17:04:25 2007 +0000
+++ b/libpurple/ft.c	Sat Nov 03 17:52:28 2007 +0000
@@ -207,15 +207,15 @@
 	switch(xfer_type) {
 		case PURPLE_XFER_SEND:
 			msg = g_strdup_printf(_("Error reading %s: \n%s.\n"),
-								  utf8, strerror(err));
+								  utf8, g_strerror(err));
 			break;
 		case PURPLE_XFER_RECEIVE:
 			msg = g_strdup_printf(_("Error writing %s: \n%s.\n"),
-								  utf8, strerror(err));
+								  utf8, g_strerror(err));
 			break;
 		default:
 			msg = g_strdup_printf(_("Error accessing %s: \n%s.\n"),
-								  utf8, strerror(err));
+								  utf8, g_strerror(err));
 			break;
 	}
 	g_free(utf8);