> $HTTP_USER_AGENT will return the browser reference string, but you may well > have to parse it quite substantially to extract the actual distinct version > of a given browser. That was it! Thanks SparkOut. As for parsing, my keitai-webpage codebooks provide the necessary regex rules.