ARCHIVED: What is LZW and what are the issues surrounding it?

LZW (Lempel-Ziv-Welch) is a popular compression algorithm used by a number of formats, including GIF, TIFF, PostScript, PDF, Unix Compress, and V.42bis. It is based on LZ77 and LZ78, methods developed by Abraham Lempel and Jacob Ziv in the 1970s, and was later refined into LZW by Terry Welch. It effectively compresses repetitive data and does so with minimal computational overhead.

Unisys used to hold the patent for LZW, though many companies and developers, including CompuServe, mistakenly believed it to be in the public domain. Unisys had always required licenses from companies that used LZW in their hardware (e.g., modem manufacturers), but for many years it overlooked LZW implementations in software. When it came to light that LZW was used in CompuServe's GIF image format, Unisys began to require that developers pay licensing fees for programs that could display or create GIF files. Although it did not require end users or the creators of freeware and other non-profit software to acquire licenses, Unisys was heavily criticized for its actions. Critics accused Unisys of trying to cash in on a format that, with the explosive growth of the World Wide Web, had become a universal standard. Unisys has denied these charges.

The US patent that Unisys held expired in June 2003, and patents in other parts of the world had all expired by July 7, 2004. Unisys currently holds and has patents pending on improvements to the LZW compression algorithm.

For more information about LZW and other compression formats, read the FAQ for the Usenet newsgroup comp.compression at:

For more information about the LZW patent, see:

Note: Popular formats that do not rely on LZW include JPEG, PNG, GNU zip (Gzip), StuffIt, and zip.