Hello all, i've investigated half a day on my accented problem on zebra : I had a specific default.idx, mapped to french-specific word-phrase-utf.chr config files, but... when searching for "helene" I had not the same results as when searching for "helene". This makes me mad, until I tought I was looking in the wrong direction. In fact, there is a default.idx file in /usr/local/share/zebra, and my zebra-biblio.cfg had : profilePath:${srcdir:-.}:/usr/local/share/idzebra/tab/:/home/paul/koha.dev/dev_week/zebraplugin/zebradb/biblios/tab/:${srcdir:-.}/tab/ I inverted my path and /usr/local : profilePath:${srcdir:-.}:/home/paul/koha.dev/dev_week/zebraplugin/zebradb/biblios/tab/:/usr/local/share/idzebra/tab/:${srcdir:-.}/tab/ and ... it works. Thus, it seems that profilePath directories are solved : "the 1st that fits is choosen" instead of "the last that fits has won" (like for cascading stylesheets for example) So my 2 questions : * did you encounter this problem and had the same conclusion ? * don't you think it could be reported as a problem to indexdata ? * isn't the wiki wrong on this ? (I can change it if you confirm) and yes, I know, that's 3 questions ;-) -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)