LOGFILE=/root/tweets/cron.log
> $FILE
curl https://nitter.net/keshinko | grep tweet-content | head -n 1 | sed -e 's/<[^>]*>//g' -e 's/^[ \t]*\(.*$\)/\1/' >> $FILE
-echo "$(tail -n 1000 $LOGFILE)" > $LOGFILE
+tail -n 1000 $LOGFILE > $LOGFILE
date -R >> $LOGFILE