Version 1.5.17

- Updated the Swedish OSD texts (thanks to Tomas Berglund).
- Made the 'pic2mpg' script of the 'pictures' plugin work with uppercase filename
  extensions and relative paths (thanks to Stefan Wagner for reporting this one).
- Updated the Romanian OSD texts (thanks to Lucian Muresan).
- Updated the Dutch OSD texts (thanks to Johan Schuring).
- Stripping control codes 0x86 and 0x87 from SI strings.
- Updated French language texts (thanks to Jean-Claude Repetto).
- Fixed handling 3 and 4 byte UTF-8 symbols in Utf8CharGet() (thanks to Andreas
  Mair).
- Fixed a crash in cFreetypeFont::DrawText() if an unknown symbol is encountered
  (thanks to Tobias Grimm). Unknown symbols are replaced with a '?'.
- Updated the Slovenian OSD texts (thanks to Matjaz Thaler).
- Updated the Czech OSD texts (thanks to Vladimír Bárta and Jiri Dobry).
- Updated the Turkish OSD texts (thanks to Oktay Yolgeçen).
- The 'plugins' target in the Makefile now returns an error exit code if one of the
  plugins failed to compile (suggested by Tobias Grimm).
- Rendering the non-breaking space symbol as a blank (thanks to Tobias Grimm).
- Changed the default character set for SI data from ISO6937 (as required by the DVB
  standard ETSI EN 300 468) to ISO-8859-9, in order to work around the stupidity of
  some providers, who actually use ISO-8859-9, but fail to correctly announce that.

libsi/si.c

@@ -6,7 +6,7 @@
  *   the Free Software Foundation; either version 2 of the License, or     *
  *   (at your option) any later version.                                   *
  *                                                                         *
- *   $Id: si.c 1.22 2007/07/21 13:49:48 kls Exp $
+ *   $Id: si.c 1.24 2008/03/01 12:02:01 kls Exp $
  *                                                                         *
  ***************************************************************************/
 
@@ -340,6 +340,9 @@ bool SetSystemCharacterTable(const char *CharacterTable) {
 // and length are adjusted accordingly.
 static const char *getCharacterTable(const unsigned char *&buffer, int &length, bool *isSingleByte = NULL) {
    const char *cs = "ISO6937";
+   cs = "ISO-8859-9"; // Workaround for broadcaster stupidity: according to
+   // "ETSI EN 300 468" the default character set is ISO6937. But unfortunately some
+   // broadcasters actually use ISO-8859-9, but fail to correctly announce that.
    if (isSingleByte)
       *isSingleByte = false;
    if (length <= 0)
@@ -415,7 +418,6 @@ void String::decodeText(char *buffer, int size) {
       if (    ((' ' <= *from) && (*from <= '~'))
            || (*from == '\n')
            || (0xA0 <= *from)
-           || (*from == 0x86 || *from == 0x87)
          )
          *to++ = *from;
       else if (*from == 0x8A)