This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Bill, > Russ, Glenn and Steve -- I made some minor changes to the znetcdf info page > and added some performance numbers: > http://snow.cit.cornell.edu/noon/z_netcdf.html > > If it looks fine, I would like to announce it generally tomorrow on the > netcdf mailing list. > > Your comments are appreciated. The compression is impressive, especially considering there is no significant difference in performance. You might change the reference to ncgzip as a "utility routine" to a "utility program", assuming it's a program you can invoke from the command line rather than a routine you have to invoke from a program. I wasn't quite clear on whether setting the Buffer Size to 8k meant the default compression factor was 0, or whether the Buffer Size could be set independently of the compression factor. Maybe my confusion is just from your example timings, where the 8k buffer size is associated with no compression. --Russ