rbriber@POLY1.NIST.GOV (12/21/90)
In reference to the following discussion about the number of command line arguments >In article <9012182223.AA05751@>, brad@lsr-vax.UUCP (Brad Zoltick - LMO) >writes: >> We are running up against the number-of-arguments limit allowed on >> the command line under the C shell. Is there a way to increase >> this limit to a much larger value. Shell files that run properly under >> SUNOS4 are failing under IRIX-3.31 with the error >> >> "argument number too large". >No matter how large you make the shell and kernel arg limits, someone will >hit them. For example, if you needed to fiddle with all of the files in a >full netnews spool tree, no plausible limit would work. > . > . > . >It should be noted that flat directories of hundreds of files are not >efficient. Trees are much better for computers as well as people. We are also doing scientific analysis and the the result is directories containing many hundreds of files (or more) for short peroids of time (in our case it's during feature extraction from digitized images). While directory trees may be "better" it is not always practical to make a bunch of subdirectories for the analysis files and then remove them all afterwards. I think the question being posed in the original posting :^) is whether the end user has a way to increase the number of arguments that the shell will accept when expanding out wild cards or if this is fixed by Irix. -- ------------------------------------------------------------------------ | Adios Amebas, | "I've tried and tried and I'm still mystified, | | Rob Briber | I can't do it anymore and I'm not satisfied" | | NIST 224/B210 | -Elvis | | Gaithersburg, MD 20899| rbriber@poly1.nist.gov (Internet) | | (301) 975-6775 (voice)| rbriber@enh.nist.gov (Internet) | | (301) 975-2128 (fax) | rbruber@nbsenh.bitnet (Bitnet) | ------------------------------------------------------------------------