I have an open source desktop search. Instead of indexed search. It does an append operation putting all the cls frm txt htm and other text format data into 1 huge file. Then search that file without any directory overhead. The file names seperates the data in the big file. That name is displayed in the form header. It's plenty fast. Searching at 20,000,000 cps and beyond. A key text search function is marching lines only versus context results. All matches hi-lited. Data extracts and more. Much more.