Services & ResourcesWolfram ForumsMathGroup Archive

[Date Index] [Thread Index] [Author Index]

undocumented limits on subprocesses?

• To: mathgroup at smc.vnet.net
• Subject: [mg121090] undocumented limits on subprocesses?
• From: algal <alexisgallagher at gmail.com>
• Date: Fri, 26 Aug 2011 05:26:47 -0400 (EDT)
• Delivered-to: l-mathgroup@mail-archive0.wolfram.com

Hi,

I am trying to understand why a command-line utility I have written
always crashes only when I try to call it from Mathematica. Are there
little-documented limits on what child processes can do -- e.g., on
memory allocation, handling of the environment, etc.?

The command-line utility is a large (70MB) standalone binary
executable produced by Steel Bank Common Lisp (SBCL). It queries the
network and prints csv-formatted data to stdout. I've been able to run
it successfully from the command line, from shell scripts, from R, and
from Python. But when I try to run it from Mathematica, using Export
or OpenRead, the binary crashes.

I don't have this problem when I use Mathematica merely to call a
similar executable which only prints  prints a thousand lines of
"Hello, World" (roughly the number of lines of the actual utility), so
the problem depends on the computation done by the binary.

Any suggestions or pointers appreciated.

• Follow-Ups:
  • Re: undocumented limits on subprocesses?
    • From: "McHale, Paul" <Paul.McHale@excelitas.com>