Hi, Patrick!

Mine is a HPUX 32 bits... :-(

I was trying your package because I intend to develop some data exchange
between a perl based webserver (Apache mod_perl) and a HP Openview Network
Node Manager (NNM) 6.2 to gather direct data from it. HP provided some
public Java API access classes. There is a ported perl API module but it is
a comercial package (http://www.cs-net.com). Inline::Java will solve my
problem at very low cost!

For instance, PerlNatives is not a issue for me...

At least I'm happy I could port "not available" perl modules at ActiveState
to my platform. Other was DBD-solid 0.20a to get access to OV NNM embedded
database. Somewhat trickly to compile, but they work.

Thanks in advance
Roberto Kodama
from Brazil.

----- Original Message ----- 
From: "Patrick LeBoutillier" <[EMAIL PROTECTED]>
To: "Roberto Kodama" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Tuesday, February 10, 2004 11:57 AM
Subject: Re: HP-UX 11 PerlNatives error test report


> Roberto,
>
> Thanks a lot. The .sl and the 'next available port numer' stuff will be
> fixed in the next version.
> As for the PerlNatives stuff, it's a bit more complicated. Since I don't
> have access
> to HPUX, I can't really debug anything.
>
> The problem is basically one of portability with the JNI/C code, possibly
> with the
> va_arg stuff. Is HPUX a 64bit architecture?
>
> Cheers,
>
> Patrick
>
>
>
>
> ----- Original Message -----
> > Dear Patrick,
> >
> > print out of requested commands issued are attached to this message.
> >
> > Thanks in advance
> >
> > Roberto Kodama
> >
> > -----Original Message-----
> > From: Patrick LeBoutillier
> > To: Roberto Kodama
> > Cc: [EMAIL PROTECTED]
> > Sent: 09/02/04 08:55
> > Subject: Re: HP-UX 11 PerlNatives error test report
> >
> > Roberto,
> >
> > Thanks a lot for this valuable feedback.
> >
> > I don't have access to HP-UX, so what does:
> >
> >   % perl -e'print "$^O\n"'
> >
> > print out? This will allow me to fix Portable.pm accordingly.
> >
> > Also can you send me the output from these commands:
> >
> > % export PERL_INLINE_JAVA_DEBUG=4
> > % perl -Mblib t/12_2_perl_natives.t
> >
> > This will give me better insight into what is happening...
> >
> > Thanks,
> >
> > Patrick
> >
> >
> > On Sat, 2004-02-07 at 19:02, Roberto Kodama wrote:
> >> Folks!
> >>
> >> as asked I'm send you "make test" log for perl_natives, where it
> > failed.
> >>
> >> Hardware: PA_RISC1.1 Workstation HP C110 / 384 MB RAM
> >> OS: HP-UX 11 patched with last quality pack (December/2003) and last
> > LD
> >> consolidated patch (PHSS_30048)
> >>
> >> Compiler: gcc 3.3.1
> >> Java 1.4.2.01 - ActiveState downloaded from HP
> >> Perl 5.6.1 - downloaded from HP
> >> Inline version is 0.44
> >> Inline::Java version is 0.45c
> >>
> >> hacks done at Inline::Java in order to be compilable and linkable:
> >>
> >> adjust Java/Portable.PM (shared libraries at HP-UX are sufixed as .sl
> >> instead .so)
> >>
> >> JVM_LIB                         =>      'libjvm.sl',
> >> JVM_SO                          =>      'libjvm.sl',
> >>
> >> perl Makefile.PL J2SDK=/opt/java1.4
> >>
> >> make (using JNI)
> >>
> >> export PERL_INLINE_JAVA_JNI=1
> >> export
> >>
> > SHLIB_PATH=/opt/java1.4/jre/lib/PA_RISC:/opt/java1.4/jre/lib/PA_RISC/ser
> > ver:
> >> /opt/java1.4/jre/lib/PA_RISC/native_threads
> >> export
> >>
> > LD_LIBRARY_PATH=/opt/java1.4/jre/lib/PA_RISC:/opt/java1.4/jre/lib/PA_RIS
> > C/se
> >> rver:/opt/java1.4/jre/lib/PA_RISC/native_threads
> >> export
> >>
> > LD_PRELOAD=/opt/java1.4/jre/lib/PA_RISC/hotspot/libjvm.sl:/opt/Inline-Ja
> > va-0
> >> .45/blib/arch/auto/Inline/Java/JNI/JNI.sl
> >>
> >> make test
> >> make install
> >>
> >> results of make test
> >>
> >> t/10_1_shared_alone.skipped test on this platform
> >> t/10_2_shared_start..skipped test on this platform
> >> t/10_3_shared_use....skipped test on this platform
> >> t/10_4_shared_stop..skipped test on this platform
> >> t/10_5_shared_fork..skipped test on this platform
> >> t/10_6_shared_sim..skipped test on this platform
> >>
> >> t/12_2_perl_natives.
> >> Note: PerlNatives is still experimental and errors here can safely
> >> be ignored if you don't plan on using this feature. However, the
> >> author would appreciate if errors encountered here were reported
> >> to the mailing list ([EMAIL PROTECTED]) along with your hardware/OS
> >> detail. Thank you.
> >> Argument "" isn't numeric in addition (+) at t/12_2_perl_natives.t
> > line 71.
> >> t/12_2_perl_natives.ok 2/5Argument "" isn't numeric in addition (+) at
> >> t/12_2_pe
> >> rl_natives.t line 71.
> >> t/12_2_perl_natives.FAILED tests 1, 3
> >>         Failed 2/5 tests, 60.00% okay
> >
>


Reply via email to