diff --git a/scripts/odt2pdf/odt2pdf.sh b/scripts/odt2pdf/odt2pdf.sh
index befdaeedcf5..fb44b2329ab 100755
--- a/scripts/odt2pdf/odt2pdf.sh
+++ b/scripts/odt2pdf/odt2pdf.sh
@@ -5,13 +5,18 @@
 # @copyright  GPL License 2019 - Camille Lafitte - cam.lafit@azerttyu.net 
 #
 # Convert an ODT into a PDF using "native" or "jodconverter" or "pyodconverter" or "unoconv" tool.
-# Dolibarr variable MAIN_ODT_AS_PDF must be defined 
-#  to value "native" to call soffice native exporter feature  
+# Dolibarr variable MAIN_ODT_AS_PDF must be defined ...
+#  to value "libreoffice" to call soffice native exporter feature (in such a case, this script is useless)  
 #  or value "unoconv" to call unoconv CLI tool after ODT generation.
 #  or value "pyodconverter" to call DocumentConverter.py after ODT generation.
 #  or value "jodconverter" to call jodconverter wrapper after ODT generation
 #  or value "/pathto/jodconverter-cli-file.jar" to call jodconverter java tool without wrapper after ODT generation.
 # Dolibarr variable MAIN_DOL_SCRIPTS_ROOT must be defined to path of script directories (otherwise dolibarr will try to guess).
+#
+# NOTE: Using this script is depcrecated, you can now convert generated ODT to PDF on the fly by setting the value MAIN_ODT_AS_PDF
+# to 'libreoffice'. It requires only soffice (OpenOffice or LibreOffice) installed on server (use apt install soffice libreoffice-common libreoffice-writer).
+# If you got this error: javaldx failed! Warning: failed to read path from javaldx with no return to prompt when running soffice --headless -env:UserInstallation=file:"/tmp" --convert-to pdf --outdir xxx ./yyy.odt, 
+# check that directory defined into env:UserInstallation parameters exists and is writeable.
 
 
 if [ "x$1" == "x" ]