2008.02.06 17:14 "[Tiff] Re: Tiff Digest, Vol 45, Issue 3", by Gary McGath

"Anders Sewerin Johansen" <asjo@kb.dk> wrote:

> The reason for the question is that several parameters are relevant

for picking a format for our purposes. Some are obvious to outsiders of final file, CPU requirements for rendering...), while others >(size may not be (openness of format, availability of several open source renderes, patent issues). In particular the tiff/lzw format, while suitable on most counts (stable, well-known...) is discounted due to the patent issues past, present and future of the lzw compression algorithm/implementation. The tiff/zip format would not be discounted on the patent issue as far as I know, but might be discounted with regards to cross-platform usability, if the default Linux/UNIX system configuration did not allow for reading and writing it.

One issue which concerns me, specifically with regard to long-term preservation, is that there is no vendor-independent specification for ZIP. This doesn't matter for most users, since the format is well-documented, but it has "reserved" fields, which raise questions for the reconstruction of ZIP archives decades later. I discussed that issue a little here: http://fileformats.blogspot.com/2007/08/should-there-be-iso-zip-standard.html

Of course, ODF and other formats which subsume ZIP compression have the same issue, so there's a large enough body of ZIP compressed documents that my concern is probably just academic.

--
Gary McGath
Digital Library Software Engineer

Harvard University Library Office for Information Systems http://hul.harvard.edu/~gary/index.html