I'm editing an HTML file in Bluefish and I use the text plug-‏‏in, quoted exactly here from the editor, and in Konqueror it's showing up as plug-â€�â€�in, with boxes where the � are. I know I'm using a proper - because I keep typing it in, but the only way to fix the apparent problem seems to be deleting g-i and retyping it. This hasn't been a problem with viewing in Firefox, which probably just ignores it, so I haven't noticed until now. Encoding is UTF-8. This is showing up every place I use -. I originally wrote a lot of the text in OpenOffice.org Writer and pasted later directly into Bluefish. I guess I just assumed that it would at least put a placeholder where an unprintable character was. Sure enough, Kate shows the extra characters. Anyone know how to get rid of them, or at least see them in Bluefish? Can sed handle upper-half characters if I use hex codes? Thanks.
ta0kira

Perhaps something was wrong with your X input method (or since you are using Bluefish, your GTK input method).

Btw, GNU sed 4.1.5 works as expected (at least on my machine). E.g.,
removes all those garbage characters.

Thanks. You must have used a script for that, unless Konsole just sucks. It will only allow pasting of â and not the other two. Either way, I took care of it in Kate. Surprisingly, I had about 500 instances of that particular set of three characters. Pretty much everywhere - showed up in the Writer document.

It could also be that GTK and Klipper don't mix well, but that doesn't explain why Bluefish fails to show them. Pasting those characters into Bluefish makes them show up, but it probably does some sort of translation. If you put those three characters into a UTF-8 file with Kate then open it with Bluefish they won't show up. I've reported it as a Bluefish bug.
ta0kira

I just pasted directly into an xterm. The � character does not show up, but it is still “there” on the commandline, and expected results occur.