in reply to Testing for read() failures on different platforms

What is your LANG environment variable set to on each platform? RHEL3 sets it to e.g. en_US.utf8 by default. The unicode locale causes lots of new, different, and possibly buggy behavior when reading data from a file.
"Even if you are on the right track, you'll get run over if you just sit there." - Will Rogers
  • Comment on Re: Testing for read() failures on different platforms

Replies are listed 'Best First'.
Re^2: Testing for read() failures on different platforms
by leriksen (Curate) on Jul 27, 2004 at 05:14 UTC 
    [le6303@itdevtst src]$ env | grep LANG
LANG=en_US

    [download]
    I set it to this cause the utf8 stuff was causing Tk to not compile. See the bug report

    +++++++++++++++++
    #!/usr/bin/perl
    use warnings;use strict;use brain;

[reply]
[d/l]
      I assume it's the same on both platforms, then? What happens if you set it to "C"?

      How about the modules involved? All the same versions? Can you use the same version of Perl, preferably from the same source tree and with build options as similar as possible? This is just the standard troubleshooting checklist that I never resort to unless I can't solve a problem through dumb luck and/or inspiration. 8)

      "Even if you are on the right track, you'll get run over if you just sit there." - Will Rogers
[reply]
        Actually it is C on Solaris. Setting it to C on Linux makes no difference.

        Yes it is possible to set the two platforms to the same version of Perl, but I dont think that will help. I think the difference is due to the OS's, not the Perl versions (which only differ by point releases, not a significant difference)

        I guess what I'm looking for is something more robust than trying a failing read with dirhandles - that's what I am looking for sugestions on.

        +++++++++++++++++
        #!/usr/bin/perl
        use warnings;use strict;use brain;

[reply]

In Section Seekers of Perl Wisdom