[rt-users] upgrading 2.0.7 to 2.0.15, fixdeps, and good stuff

Forrest Stanley fstanley at netburner.com
Wed Oct 30 13:51:10 EST 2002



At 09:16 AM 10/30/2002 -0800, Forrest Stanley wrote:
>As far as mason go's, be sure you are running Mason 1.05. I was running 
>1.1+ in my first installation of RT. I was getting several errors, 
>sometimes fixable with hacks, sometimes not. Upon scouring this list, I 
>learned that Mason 1.05 seems to be the optimal version of mason with any 
>RT version. Since you hacked the RT Makefile to download Mason via CPAN, 
>I'm guessing that you have the 1.1+ series of Mason installed.
>
>-Forrest


Furthermore, here is part of my own (working) installation procedure:

- Install prereqs for HTML mason
         perl -MCPAN -e shell
         install Params::Validate
         install Exception::Class
         install Class::Container
         install Scalar::Util
         install File::Spec

- Download and Install HTML-Mason 1.05 (must be 1.05, do not d/l via CPAN)
         gunzip -c HTML-Mason-1.05.tar.gz | tar xvf -
         cd HTML-Mason-1.05
         perl Makefile.PL
         make
         make install








>At 04:18 PM 10/29/2002 -0800, you wrote:
>>Replying to myself with two issues discovered.
>>
>> >
>> > We had a report of spam-blockers not liking the "Sender:" line
>> > that differs from the From:. Thanks to the rt-users, using sendmailpipe
>> > instead of sendmail got rid of the line.
>>
>>The Sender: line is gone, but our customer is complaining that the
>> >From line at the top of the header is different from the From: line
>>that RT adds (causing the customer's spam filter to kick in).
>>(The first one is the username of the httpd process)
>>
>>Is there any way with sendmail/sendmailpipe to get these two From
>>lines to agree?
>>
>> > If anyone can tell me if I'm hosed because of the cpan forced
>> > install, I'd appreciate it.
>>
>>I'm getting some spurious failures in HTML::Mason.  Only one user
>>can reproduce it, the rest saw it once after the upgrade.  No meaningful
>>error in rt.log. or apache's logs.  Anyone seen this before?  Google
>>seems to reference perl 5.6.0 and/or some modules as guilty parties.
>>It looks like Mason is already dying for some reason and then chokes
>>on the die command.
>>
>>----
>>
>>System error
>>
>>error:  Bizarre copy of ARRAY in aassign at 
>>/usr/lib/perl5/site_perl/5.6.0/Devel/StackTrace.pm line 67, line 2.
>>
>>context:  ...
>>314:  } else {
>>315:  eval {$self->comp({base_comp=>$request_comp}, $first_comp, 
>>@request_args)};
>>316:  }
>>317:  select $old;
>>318:  die $@ if $@;
>>319:  }
>>320:  };
>>321:
>>322:  # Handle errors.
>>...
>>
>>code stack:  /usr/lib/perl5/site_perl/5.6.0/HTML/Mason/Request.pm:318
>>
>>
>>
>>--
>>--------------------------------------------------------
>>Brian McClendon     bam at intrinsic.com       650-966-6804
>>--------------------------------------------------------
>>_______________________________________________
>>rt-users mailing list
>>rt-users at lists.fsck.com
>>http://lists.fsck.com/mailman/listinfo/rt-users
>>
>>Have you read the FAQ? The RT FAQ Manager lives at http://fsck.com/rtfm
>
>_______________________________________________
>rt-users mailing list
>rt-users at lists.fsck.com
>http://lists.fsck.com/mailman/listinfo/rt-users
>
>Have you read the FAQ? The RT FAQ Manager lives at http://fsck.com/rtfm




More information about the rt-users mailing list