Last week during a break between talks at a science and technology conference I had a delightful chance encounter with an old acquaintance, Bill Press, who gave me a precious gift: he told me that he still uses the "Free Text Indexer/Browser" software that I wrote 20 years ago! I suspect he's the last person on the planet to do so. What he still likes about my antiquated real-time high-bandwidth information retrieval (IR) code is that it's:

  • simple (a few hundred lines of pedestrian C)
  • transparent (no mysteries about stopwords, stemming, etc.)
  • fast (sub-second response time)

It's also free and works on just about any computer system. The interface is ugly but functional. Free Text provides some powerful features that are lacking in most "modern" IR tools:

  • alphabetized lists of all words in the database, with the number of times each appears
  • fuzzy and/or/not proximity search filtering
  • instant key-word-in-context (KWIC) displays for rapid scanning
  • easy access to arbitrary chunks of text for reading and copy/paste into notes elsewhere

Hmmmmm ... I really need to dig out that old software and rewrite it — maybe in Perl, using a proper database, with a web interface, etc. Tnx for reminding me, Bill!

(cf. Free Text Archive, Free Text IR Philosophy, and FreeTextDesiderata (29 Oct 1999), KwicsChinksAndChunks (31 Jan 2000), WorldTexasHistory (15 May 2000), FreeTextFriends (25 Aug 2000), IrWishes (4 Jan 2001), PersonalComputerHistory (25 Feb 2002), ...)

TopicProgramming - TopicPersonalHistory - 2006-04-30

(correlates: IrWishes, HelloSailor, FreeTextDesiderata, ...)