Date: Fri, 17 Dec 2004 00:07:32 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [Ux2bs_Archive] No. 377 ************************************************** Thursday 16 December 2004 Number 377 ************************************************** Subjects for today 1 Re: Perl 5.8.6 : John Poltorak 2 Re: Perl 5.8.6 : Henry Sobotka 3 Re: Perl 5.8.6 : Henry Sobotka 4 Perl help needed! : John Poltorak 5 Re: Perl help needed! : Henry Sobotka 6 Re: EMXBIND problem : Dave Yeo" **= Email 1 ==========================** Date: Wed, 15 Dec 2004 18:49:58 +0000 From: John Poltorak Subject: Re: Perl 5.8.6 On Wed, Dec 15, 2004 at 07:31:09AM -0500, Henry Sobotka wrote: > John Poltorak wrote: > > > > Can anyone suggest how I can tell whether the os2_ea.t fails because > > it runs on JFS? > > Actually it fails because the test uses '+', which is an illegal char > for EAs. The Extended Attribute Naming Conventions section of the > Control Programming Guide and Reference lists '+' among chars "reserved > for system use". Change the +'s to any other legal char and it works > fine on JFS. Presumably it passes on HPFS because '+' became illegal > with the introduction of JFS. I'm not sure I follow that... Are you saying that it passes on HPFS because '+' is legal on JFS? How do these tests differentiate between HPFS and JFS? Here is the os2_ea.t test... What changes do you think are required? Maybe we can get the test amended... BEGIN { chdir 't' if -d 't/lib'; at INC = '../lib' if -d 'lib'; require Config; import Config; if (-d 'lib' and $Config{'extensions'} !~ /\bOS2(::|\/)REXX\b/) { print "1..0\n"; exit 0; } } # Before `make install' is performed this script should be runnable with # `make test'. After `make install' it should work as `perl test.pl' ######################### We start with some black magic to print on failure. # Change 1..1 below to 1..last_test_to_print . # (It may become useful if the test is moved to ./t subdirectory.) BEGIN { $| = 1; print "1..21\n"; } END {print "not ok 1\n" unless $loaded;} use OS2::ExtAttr; $loaded = 1; print "ok 1\n"; ######################### End of black magic. # Insert your test code below (better if it prints "ok 13" # (correspondingly "not ok 13") depending on the success of chunk 13 # of the test code): unlink 't.out' if -f 't.out'; system 'cmd', '/c', 'echo OK > t.out'; { my %a; tie %a, 'OS2::ExtAttr', 't.out'; print "ok 2\n"; keys %a == 0 ? print "ok 3\n" : print "not ok 3\n"; $a{'++'} = '---'; print "ok 4\n"; $a{'AAA'} = 'xyz'; print "ok 5\n"; } { my %a; tie %a, 'OS2::ExtAttr', 't.out'; print "ok 6\n"; my $c = keys %a; $c == 2 ? print "ok 7\n" : print "not ok 7\n# c=$c\n"; my at b = sort keys %a; " at b" eq '++ AAA' ? print "ok 8\n" : print "not ok 8\n# keys=` at b'\n"; $a{'++'} eq '---' ? print "ok 9\n" : print "not ok 9\n";; $a{'AAA'} eq 'xyz' ? print "ok 10\n" : print "not ok 10\n# aaa->`$a{AAA}'\n"; $c = delete $a{'++'}; $c eq '---' ? print "ok 11\n" : print "not ok 11\n# deleted->`$c'\n";; } print "ok 12\n"; { my %a; tie %a, 'OS2::ExtAttr', 't.out'; print "ok 13\n"; keys %a == 1 ? print "ok 14\n" : print "not ok 14\n"; my at b = sort keys %a; " at b" eq 'AAA' ? print "ok 15\n" : print "not ok 15\n"; $a{'AAA'} eq 'xyz' ? print "ok 16\n" : print "not ok 16\n";; ! exists $a{'+'} ? print "ok 17\n" : print "not ok 17\n";; ! defined $a{'+'} ? print "ok 18\n" : print "not ok 18\n# ->`$a{'++'}'\n";; ! exists $a{'++'} ? print "ok 19\n" : print "not ok 19\n";; ! defined $a{'++'} ? print "ok 20\n" : print "not ok 20\n# ->`$a{'++'}'\n";; } print "ok 21\n"; unlink 't.out'; > h~ -- John _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 2 ==========================** Date: Wed, 15 Dec 2004 15:02:12 -0500 From: Henry Sobotka Subject: Re: Perl 5.8.6 John Poltorak wrote: > > I'm not sure I follow that... Are you saying that it passes on HPFS > because '+' is legal on JFS? How do these tests differentiate between HPFS > and JFS? No, it passes on HPFS because (presumably, don't have an older version of the Control Programming Guide readily available to check) '+' was once legal, and fails on JFS because '+' is now reserved for system use (along with $, at and &). The fact that it still works on HPFS might be considered a file-system bug. The test doesn't differentiate between the two file systems. But if you copy it to an HPFS partition and run it there, it passes. > Here is the os2_ea.t test... What changes do you think are required? Maybe > we can get the test amended... Change the +'s to anything except one off the four illegal characters, e.g. '*', and the test passes on JFS. h~ _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 3 ==========================** Date: Wed, 15 Dec 2004 17:26:42 -0500 From: Henry Sobotka Subject: Re: Perl 5.8.6 I just looked at the Visual Age 3.0 Control Programming Guide (1995, pre-JFS) and it lists '+' as one of the four characters not to be used in naming EAs ("reserved for system use"). So the Perl test is flawed in not expecting '++' to fail and treating the failure as a pass, and the fact that it happens to work on HPFS is a file-system bug. h~ _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 4 ==========================** Date: Wed, 15 Dec 2004 23:27:21 +0000 From: John Poltorak Subject: Perl help needed! It looks like my mailing list manager for the os2-unix mailing list has packed up for some mysterious reason. It is my own hacked version of Majordomo which had been managing mail for over four years, but now it doesn't seem to do anything. It uses Perl version 5.005_53 built for os2 and I'm not aware of making any changes to the environment it runs in which currently has an uptime of 175 days. Is there any debugging I can turn to force some response from this old version of Perl? I guess it could be a problem with sendmail rather than Majordomo, but the mail server appears to have been working OK... -- John _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 5 ==========================** Date: Wed, 15 Dec 2004 18:55:21 -0500 From: Henry Sobotka Subject: Re: Perl help needed! John Poltorak wrote: > > It looks like my mailing list manager for the os2-unix mailing list has > packed up for some mysterious reason. It is my own hacked version of > Majordomo which had been managing mail for over four years, but now it > doesn't seem to do anything. It uses Perl version 5.005_53 built for os2 > and I'm not aware of making any changes to the environment it runs in > which currently has an uptime of 175 days. > > Is there any debugging I can turn to force some response from this old > version of Perl? You can run perl -d to step through the script, or there may be a $debug variable at the top of Majordomo whose value can be changed from 0 to 1 to turn on debugging. But if the software has been running fine for four years, it could be stuff like a humungous log or partition filling up or resource depletion, i.e. more of a system thing. Unless you're going for a Guinness record for uptime, try C-A-D before busting your brains. h~ _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs **= Email 6 ==========================** Date: Wed, 15 Dec 2004 16:46:18 -0800 From: "Dave Yeo" Subject: Re: EMXBIND problem On Wed, 15 Dec 2004 11:15:24 +0000, John Poltorak wrote: > >When building GETTEXT there is a build failure due to EMXBIND from EMX/GCC >2.8.1. Does anyone know whether EMXBIND has been updated, or if not, then >where do I get a suitable drop in replacement? PGCC 2.95.3, GCC3.0.3 and GCC 3.2.1 all have updated EMXBIND and the other EMX tools with 3.2.1 having the newest I'd guess Dave _______________________________________________ UX2BS mailing list UX2BS at os2ports.com http://os2ports.com/mailman/listinfo/ux2bs