From toby.corkindale at strategicdata.com.au  Mon May  7 18:34:26 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Tue, 08 May 2012 11:34:26 +1000
Subject: [Melbourne-pm] Melbourne Perl mongers May meeting is tomorrow
Message-ID: <4FA87822.5040403@strategicdata.com.au>

Hello,
The Melbourne Perl Mongers will be meeting tomorrow, on Wednesday the 
9th of May, 2012 and will start around 6:30pm at this location:

   Strategic Data
   Level 2 51-55
   Johnston street
   Fitzroy

I believe we have the following agenda:

  * Kahlil Hodgson presenting on "Perl for profit" - experiences of 
using Perl in a new technology start-up.

  * Tony Smith, leading a discussion on best practices when working with 
(and hopefully upgrading) 20th century Perl code.


It doesn't look like Perl 5.16 will be released by tomorrow, so we'll 
have to hold the party for that next month :)

Cheers,
Toby

From toby.corkindale at strategicdata.com.au  Tue May  8 18:24:31 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Wed, 09 May 2012 11:24:31 +1000
Subject: [Melbourne-pm] TONIGHT - Melbourne Perlmongers
In-Reply-To: <4FA87822.5040403@strategicdata.com.au>
References: <4FA87822.5040403@strategicdata.com.au>
Message-ID: <4FA9C74F.4020703@strategicdata.com.au>

Don't forget, tonight's meeting is.. Tonight!

Toby

-------- Original Message --------
Subject: [Melbourne-pm] Melbourne Perl mongers May meeting is tomorrow
Date: Tue, 08 May 2012 11:34:26 +1000
From: Toby Corkindale <toby.corkindale at strategicdata.com.au>
To: melbourne-pm <melbourne-pm at pm.org>

Hello,
The Melbourne Perl Mongers will be meeting tomorrow, on Wednesday the
9th of May, 2012 and will start around 6:30pm at this location:

   Strategic Data
   Level 2 51-55
   Johnston street
   Fitzroy

I believe we have the following agenda:

  * Kahlil Hodgson presenting on "Perl for profit" - experiences of
using Perl in a new technology start-up.

  * Tony Smith, leading a discussion on best practices when working with
(and hopefully upgrading) 20th century Perl code.


It doesn't look like Perl 5.16 will be released by tomorrow, so we'll
have to hold the party for that next month :)

Cheers,
Toby
_______________________________________________
Melbourne-pm mailing list
Melbourne-pm at pm.org
http://mail.pm.org/mailman/listinfo/melbourne-pm

From kahlil.hodgson at dealmax.com.au  Fri May 11 01:54:27 2012
From: kahlil.hodgson at dealmax.com.au (Kahlil Hodgson)
Date: Fri, 11 May 2012 18:54:27 +1000
Subject: [Melbourne-pm] Working with old Perl code
Message-ID: <4FACD3C3.5050603@dealmax.com.au>

Hi All,

After Tony's discussion at Wednesdays meeting, I've had some thoughts about
generic strategies for cleaning up old perl code.  I totally agree with 
Tony's
reluctance to mess with the structure of the code too early: the 
potential to
change or break functionality before it is well understood is just to 
great.
One fairly safe strategy, however, is to clean up the use of variables and
references.

The code in Tony's example was clearly from a time before lexical variables
either existed or where in common use.  Simply walking through this code and
inserting 'my' in the appropriate places to bring non-global variables 
into a
lexical scope should be fairly safe (provided you carefully check that the
variables are not being used globally).  More importantly, the process helps
to pin-point the nasty global variables, which you can mark and fix up 
later.
As part of this code walk, you will probably find uses of 'local' that are
really just being used for scoping.  Figuring out which instances can be
replaced with 'my' helps to point out the genuine uses of localization (if
any) which you may also decide to mark and fix up later.

This 'first pass' can be done fairly quickly and will make the code look and
behave a lot more like a modern perl

Older code sometimes uses *TYPEGLOBs instead of variable references. 
Now, for
those who don't know, *TYPEGLOB's are references into the symbol table (the
place where we keep global variables) and simultaneously point to all 
'things'
(or types?) which match (or glob to?) a particular symbol name.  Okay, 
that's
a _very_ rough definition, but *TYPEGLOBs (or *TYPEGOBLINs?) are _very_ odd
creatures.  I found the first chapter of "Advanced Perl Programming" by 
Simon
Cozens particularly enlightening on this topic (contact me off-list if 
you want
to borrow a copy).

Believe it or not, there are sensible places to use *TYPEGLOBs, but they are
relatively rare in modern perl and mostly live in CPAN modules (mostly)
written by people who "really know what they are doing".

Using *TYPEGLOBs in place of variable references is _not_ a sensible 
thing to
do in modern perl.  The code that Tony's is dealing with is a good 
example.

 From memory we had something like:

     &read_config(*my_config);
         ...
     %my_config{'some_key'} = 'some_value';

We could try to replace this with

     my $my_config_ref = {};
     &read_config($my_config_ref);
         ...
     $my_config_ref->{'some_key'} = 'some_value';

But then we're going to have to change the code in 'read_config' to get 
it to
work, and that's not in spirit of making small changes and not breaking
functionality.  You can get around this with the following trick:

     my $my_config_ref = {};
     *my_config = $my_config_ref; # temp until we fix read_config
     &read_config(*my_config);
         ...
     $my_config_ref->{'some_key'} = 'some_value';

The trick changes the 'hash' slot for the symbol 'my_config' to point to the
same hash that $my_config_ref points to.

The following little test script may clarify for those who doubt

     #!/usr/bin/perl

     use v5.10;

     sub read_config {
         local (*my_config) = @_; # NB: can't use 'my' on a TYPEGLOB
         $my_config{'pre'} = 'from config';
     }

     my $my_config_ref = {};
     *my_config = $my_config_ref;
     read_config(*my_config);

     say $my_config_ref->{'pre'};

     $my_config_ref->{'post'} = 'after config';
     say $my_config_ref->{'post'};

     say $my_config_ref->{'pre'};

Once you are happy that this is all working correctly, you can then go into
'read_config' and replace the *TYPEGOBLIN and get rid of the trick (I'm
assuming nothing else calls read_config this way).

Getting rid of the superfluous '&' in front of subroutine calls is another
quick fix.

Finally, after doing all that you can probably try

     use strict;
     use warnings;

and see if the remaining badness is manageable.

Thanks to Tony for bringing up this topic and I hope my suggestions help.

Happy Hacking!

Kal

-- 
Kahlil (Kal) Hodgson                       GPG: C9A02289
Head of Technology                         (m) +61 (0) 4 2573 0382
DealMax Pty Ltd                            (w) +61 (0) 3 9008 5281

Suite 1415
401 Docklands Drive
Docklands VIC 3008 Australia

"All parts should go together without forcing.  You must remember that
the parts you are reassembling were disassembled by you.  Therefore,
if you can't get them together again, there must be a reason.  By all
means, do not use a hammer."  -- IBM maintenance manual, 1925


From toby.corkindale at strategicdata.com.au  Mon May 14 23:12:17 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Tue, 15 May 2012 16:12:17 +1000
Subject: [Melbourne-pm] Next meeting - 14th June
Message-ID: <4FB1F3C1.7090706@strategicdata.com.au>

Hi all,
our next meeting is scheduled to be Wednesday 14th of June, 2012.

Potential talks I have so far are:
  * What's new in Perl 5.16
  * Scott may be ready for a talk on Perl on BeagleBone by then (and if 
not, I might do a ten minute overview of what the BB is, at least)

Does anyone else have a presentation they'd like to present at the next 
meeting?

Regarding Perl 5.16:
The 2nd release candidate came out today, and if no showstopper bugs are 
found, 5.16.0 should be out in a week from now.


Cheers,
Toby

From christopher.short at monash.edu  Tue May 15 22:51:31 2012
From: christopher.short at monash.edu (Christopher Short)
Date: Wed, 16 May 2012 15:51:31 +1000
Subject: [Melbourne-pm] XML::Xerces
Message-ID: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>

Howdy folks,

have any of you managed to successfully install XML::Xerces under Perl 5.14 ?

I gather it has issues after Perl 5.8

cheers,
Christopher

From rob at eatenbyagrue.org  Wed May 16 00:07:06 2012
From: rob at eatenbyagrue.org (Robert Norris)
Date: Wed, 16 May 2012 17:07:06 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
Message-ID: <CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>

On Wed, May 16, 2012 at 3:51 PM, Christopher Short
<christopher.short at monash.edu> wrote:
> have any of you managed to successfully install XML::Xerces under Perl 5.14 ?

Just worked through it now. Its fiddly, but it appears to work.

If you haven't already got Xerces-C installed, you'll need to build
it. XML::Xerces requires 2.7.0 only, so the latest version isn't much
good to you. Grab it here:

http://archive.apache.org/dist/xml/xerces-c/Xerces-C_2_7_0/source/xerces-c-src_2_7_0.tar.gz

Unpack and build, setting XERCESROOT along the way (critical to the
Xerces-C and Xerces::XML builds). Explicitly specifying the compilers
in runConfigure is important, as Xerces::XML will fail to build if
Xerces was not built with g++, but on many Linuxes configure finds
c++. Explicit is good!

$ tar xvfz xerces-c-src_2_7_0.tar.gz
$ cd xerces-c-src_2_7_0
$ export XERCESROOT=`pwd`
$ cd src/xercesc
$ ./runConfigure -p linux -c gcc -x g++

You'll probably want to install this somewhere, but I didn't for now.
XML::Xerces will link to wherever its built so it will work properly,
but you need to keep the build dir around if you don't install it. Or
you can do crazy shenanigans with static linking and such. Up to you
:)

Now XML::Xerces. For Perl later that 5.9.notsure it needs to rebuild
the SWIG bindings. You'll need SWIG on your path, and XML::Xerces
needs the XERCES_DEVEL=1 environment variable to tell it what to do.

$ XERCES_DEVEL=1 cpanm XML::Xerces.

That completed successfully for me. I haven't verified it other than
to note that Perl can load it ok:

$ perl -MXML::Xerces -e1
$ echo $?
0

Good luck!
Rob.

From christopher.short at monash.edu  Wed May 16 01:19:30 2012
From: christopher.short at monash.edu (Christopher Short)
Date: Wed, 16 May 2012 18:19:30 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
Message-ID: <CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>

Hey there Rob!
Sorry went down the path of "rebuilding SWIG bindings" on Monday and
couldn't get Swig to install :-( Tried again right now and no dice.

So ... following that we regressed to Xerces 1.5.2 (and discovered
Rus, our predecessor had also done that for reason long forgotten).

But 1.5.2 doesn't want to install either, despite not requiring Swig.

:-(

Marc may have found an interesting patch for XML::Xerces in a Fedora
package, so that may be worth investigating.

Christopher

On 16 May 2012 17:07, Robert Norris <rob at eatenbyagrue.org> wrote:
> On Wed, May 16, 2012 at 3:51 PM, Christopher Short
> <christopher.short at monash.edu> wrote:
>> have any of you managed to successfully install XML::Xerces under Perl 5.14 ?
>
> Just worked through it now. Its fiddly, but it appears to work.
>
> If you haven't already got Xerces-C installed, you'll need to build
> it. XML::Xerces requires 2.7.0 only, so the latest version isn't much
> good to you. Grab it here:
>
> http://archive.apache.org/dist/xml/xerces-c/Xerces-C_2_7_0/source/xerces-c-src_2_7_0.tar.gz
>
> Unpack and build, setting XERCESROOT along the way (critical to the
> Xerces-C and Xerces::XML builds). Explicitly specifying the compilers
> in runConfigure is important, as Xerces::XML will fail to build if
> Xerces was not built with g++, but on many Linuxes configure finds
> c++. Explicit is good!
>
> $ tar xvfz xerces-c-src_2_7_0.tar.gz
> $ cd xerces-c-src_2_7_0
> $ export XERCESROOT=`pwd`
> $ cd src/xercesc
> $ ./runConfigure -p linux -c gcc -x g++
>
> You'll probably want to install this somewhere, but I didn't for now.
> XML::Xerces will link to wherever its built so it will work properly,
> but you need to keep the build dir around if you don't install it. Or
> you can do crazy shenanigans with static linking and such. Up to you
> :)
>
> Now XML::Xerces. For Perl later that 5.9.notsure it needs to rebuild
> the SWIG bindings. You'll need SWIG on your path, and XML::Xerces
> needs the XERCES_DEVEL=1 environment variable to tell it what to do.
>
> $ XERCES_DEVEL=1 cpanm XML::Xerces.
>
> That completed successfully for me. I haven't verified it other than
> to note that Perl can load it ok:
>
> $ perl -MXML::Xerces -e1
> $ echo $?
> 0
>
> Good luck!
> Rob.

From christopher.short at monash.edu  Wed May 16 18:46:48 2012
From: christopher.short at monash.edu (Christopher Short)
Date: Thu, 17 May 2012 11:46:48 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
	<CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
Message-ID: <CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>

Well, we finally managed to get Swig and Xerces C and XML::Xerces 2.7
installed ...

... but then we found that they'd removed DOMParser which we needed,
so we're now trying to reinstall an earlier, buggier version. grr

Christopher

From christopher.short at monash.edu  Mon May 21 20:59:41 2012
From: christopher.short at monash.edu (Christopher Short)
Date: Tue, 22 May 2012 13:59:41 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
	<CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
	<CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>
Message-ID: <CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>

Well, we've not managed to get this working.

To give you some background, we've decided it's probably time to
upgrade our website from perl 5.8 to 5.14 ;-)

I'm installing a copy of our site onto a fresh server and so am
methodically plodding through the site to find required modules etc.

Part of the site uses XML::Xerces and XML::Xerces::DOMParser so I need
to get those working.

As per the requirements for this package, I am indeed installing the
matching version of Xerces-C first, as well as SWIG 2.0.6 to
"regenerate the bindings" using the XERCES_DEVEL = 1 trick mentioned
by Rob.

Version 2.7.0
=> everything installs
=> but DOMParser has been removed

Version 1.7.0
=> includes DOMParser
=> fails during the build of Xerces-C
"UnixHTTPURLInputStream.cpp:99:22: error: iostream.h: No such file or directory"

Version 1.5.2
=> includes DOMParser
=> Xerces-C builds and installs
=> fails during the build of XML::Xerces
"error: iostream.h: No such file or directory"
Also many complaints about
"deprecated conversion from string constant to "char *"
and
"cannot convert "char*" to "PerlInterpreter*"

Bleh!
Anyone have any tips?
Christopher

On 17 May 2012 11:46, Christopher Short <christopher.short at monash.edu> wrote:
> Well, we finally managed to get Swig and Xerces C and XML::Xerces 2.7
> installed ...
>
> ... but then we found that they'd removed DOMParser which we needed,
> so we're now trying to reinstall an earlier, buggier version. grr
>
> Christopher

From toby.corkindale at strategicdata.com.au  Mon May 21 21:11:25 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Tue, 22 May 2012 14:11:25 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
	<CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
	<CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>
	<CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
Message-ID: <4FBB11ED.1030505@strategicdata.com.au>

On 22/05/12 13:59, Christopher Short wrote:
> Well, we've not managed to get this working.
>
> To give you some background, we've decided it's probably time to
> upgrade our website from perl 5.8 to 5.14 ;-)
>
> I'm installing a copy of our site onto a fresh server and so am
> methodically plodding through the site to find required modules etc.
>
> Part of the site uses XML::Xerces and XML::Xerces::DOMParser so I need
> to get those working.

It looks like they removed DOMParser from that module ten years ago, so 
it really doesn't look like it was popular enough to get reincarnated 
elsewhere.

Could you re-write the bits of the site that use DOMParser to use a 
current API? (It doesn't look like DOMParser provided anything 
particularly special)


> As per the requirements for this package, I am indeed installing the
> matching version of Xerces-C first, as well as SWIG 2.0.6 to
> "regenerate the bindings" using the XERCES_DEVEL = 1 trick mentioned
> by Rob.
>
> Version 2.7.0
> =>  everything installs
> =>  but DOMParser has been removed
>
> Version 1.7.0
> =>  includes DOMParser
> =>  fails during the build of Xerces-C
> "UnixHTTPURLInputStream.cpp:99:22: error: iostream.h: No such file or directory"
>
> Version 1.5.2
> =>  includes DOMParser
> =>  Xerces-C builds and installs
> =>  fails during the build of XML::Xerces
> "error: iostream.h: No such file or directory"
> Also many complaints about
> "deprecated conversion from string constant to "char *"
> and
> "cannot convert "char*" to "PerlInterpreter*"
>
> Bleh!
> Anyone have any tips?
> Christopher
>
> On 17 May 2012 11:46, Christopher Short<christopher.short at monash.edu>  wrote:
>> Well, we finally managed to get Swig and Xerces C and XML::Xerces 2.7
>> installed ...
>>
>> ... but then we found that they'd removed DOMParser which we needed,
>> so we're now trying to reinstall an earlier, buggier version. grr

-- 
.signature

From toby.corkindale at strategicdata.com.au  Mon May 21 21:13:51 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Tue, 22 May 2012 14:13:51 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <4FBB11ED.1030505@strategicdata.com.au>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
	<CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
	<CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>
	<CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
	<4FBB11ED.1030505@strategicdata.com.au>
Message-ID: <4FBB127F.1060302@strategicdata.com.au>

On 22/05/12 14:11, Toby Corkindale wrote:
> On 22/05/12 13:59, Christopher Short wrote:
>> Well, we've not managed to get this working.
>>
>> To give you some background, we've decided it's probably time to
>> upgrade our website from perl 5.8 to 5.14 ;-)
>>
>> I'm installing a copy of our site onto a fresh server and so am
>> methodically plodding through the site to find required modules etc.
>>
>> Part of the site uses XML::Xerces and XML::Xerces::DOMParser so I need
>> to get those working.
>
> It looks like they removed DOMParser from that module ten years ago, so
> it really doesn't look like it was popular enough to get reincarnated
> elsewhere.
>
> Could you re-write the bits of the site that use DOMParser to use a
> current API? (It doesn't look like DOMParser provided anything
> particularly special)

In case it helps, this is an example of how to use the newer 
XercesDOMParser API:

http://cpansearch.perl.org/src/JASONS/XML-Xerces-2.7.0-0/samples/DOMPrint.pl

Toby


From rob at eatenbyagrue.org  Mon May 21 21:19:46 2012
From: rob at eatenbyagrue.org (Robert Norris)
Date: Tue, 22 May 2012 14:19:46 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
	<CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
	<CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>
	<CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
Message-ID: <CANRWccTj+v5LNNwy6Ac6eUs2XbY=-qJ1-chDZANi3pNKD7bpag@mail.gmail.com>

On Tue, May 22, 2012 at 1:59 PM, Christopher Short
<christopher.short at monash.edu> wrote:
> Anyone have any tips?

Rewrite the application, or at least part of it. Alternatively, move
the application off to its own ancient server (or VM, or chroot) and
let it live out its days in isolation. I'm not kidding.

Rob.

From sam at nipl.net  Tue May 22 01:02:06 2012
From: sam at nipl.net (Sam Watkins)
Date: Tue, 22 May 2012 18:02:06 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
	<CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
	<CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>
	<CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
Message-ID: <20120522080206.GB23422@opal.nipl.net>

Any XML parser much longer than this:

  /(<.*?>|[^<]+)\s*/g
  /(\w+)="(.*?)"/g

is too big, slow, and complex for me!  :p

Sam

From andrew at sericyb.com.au  Tue May 22 01:37:20 2012
From: andrew at sericyb.com.au (Andrew Pam)
Date: Tue, 22 May 2012 18:37:20 +1000
Subject: [Melbourne-pm] XML::Xerces
In-Reply-To: <20120522080206.GB23422@opal.nipl.net>
References: <CA+FD6gyEApia2sVZv1NUGv2YKUEGHRzhV103_Dbyrnt=Dvj71g@mail.gmail.com>
	<CANRWccQHGte38ZMMiN=uA1G1=f=MCqQ0V8YuMh1YVtL3808TUA@mail.gmail.com>
	<CA+FD6gzkT=k9_ZuggqYQwj8L-Qm+CLZO5sxnmNjU8hGTPbau_Q@mail.gmail.com>
	<CA+FD6gwhQbfyd1KF3gVahhurYH8TMWAMNhj9mmnvUsTpG70LNg@mail.gmail.com>
	<CA+FD6gzsCqvmC0h=nKJ0UDzYHu1VC826MK6hh-Q3_XKV68Hw1w@mail.gmail.com>
	<20120522080206.GB23422@opal.nipl.net>
Message-ID: <4FBB5040.5010609@sericyb.com.au>

On 22/05/12 18:02, Sam Watkins wrote:
> Any XML parser much longer than this:
> 
>   /(<.*?>|[^<]+)\s*/g
>   /(\w+)="(.*?)"/g
> 
> is too big, slow, and complex for me!  :p

You're a bad man.  :)

Andrew

From toby.corkindale at strategicdata.com.au  Tue May 22 19:25:41 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Wed, 23 May 2012 12:25:41 +1000
Subject: [Melbourne-pm] BeagleBone Perl library
Message-ID: <4FBC4AA5.3010005@strategicdata.com.au>

Hi,
I've been playing around with a BeagleBone for the past week, and I've 
come up with some initial Perl libraries to do GPIO and SPI, plus 
written some drivers to make it easy to talk to a certain LCD panel 
controller.

I know Scott is working with the BB as well, and maybe others?
My code so far is at:
https://github.com/TJC/BeagleBone

And a small blog post about the LCD controller is here:
http://blog.dryft.net/2012/05/ssd1306-lcd-controller-in-perl-for.html


Unlike on the Arduino, Perl seems like a decent language to work with on 
the 'bone.. Although I did break out into Inline::C for one part when 
driving the spidev device, since it meant I could copy-and-paste the 
ioctls from example C code :)


Cheers,
Toby

From scottp at dd.com.au  Tue May 22 19:45:29 2012
From: scottp at dd.com.au (Scott Penrose)
Date: Tue, 22 May 2012 21:45:29 -0500 (CDT)
Subject: [Melbourne-pm] BeagleBone Perl library
In-Reply-To: <4FBC4AA5.3010005@strategicdata.com.au>
References: <4FBC4AA5.3010005@strategicdata.com.au>
Message-ID: <B43FB168-FD39-47CC-BD20-7002B447E2E6@dd.com.au>

Excellent. We should merge. And I gave a domain. Perlbone

Sent from my iPhone

On 23/05/2012, at 12:26, Toby Corkindale <toby.corkindale at strategicdata.com.au> wrote:

> Hi,
> I've been playing around with a BeagleBone for the past week, and I've come up with some initial Perl libraries to do GPIO and SPI, plus written some drivers to make it easy to talk to a certain LCD panel controller.
> 
> I know Scott is working with the BB as well, and maybe others?
> My code so far is at:
> https://github.com/TJC/BeagleBone
> 
> And a small blog post about the LCD controller is here:
> http://blog.dryft.net/2012/05/ssd1306-lcd-controller-in-perl-for.html
> 
> 
> Unlike on the Arduino, Perl seems like a decent language to work with on the 'bone.. Although I did break out into Inline::C for one part when driving the spidev device, since it meant I could copy-and-paste the ioctls from example C code :)
> 
> 
> Cheers,
> Toby
> _______________________________________________
> Melbourne-pm mailing list
> Melbourne-pm at pm.org
> http://mail.pm.org/mailman/listinfo/melbourne-pm

From tconnors at rather.puzzling.org  Tue May 22 23:00:26 2012
From: tconnors at rather.puzzling.org (Tim Connors)
Date: Wed, 23 May 2012 16:00:26 +1000 (EST)
Subject: [Melbourne-pm] BeagleBone Perl library
In-Reply-To: <4FBC4AA5.3010005@strategicdata.com.au>
References: <4FBC4AA5.3010005@strategicdata.com.au>
Message-ID: <alpine.DEB.2.00.1205231536420.12626@dirac.rather.puzzling.org>

On Wed, 23 May 2012, Toby Corkindale wrote:

> Unlike on the Arduino, Perl seems like a decent language to work with on the
> 'bone.. Although I did break out into Inline::C for one part when driving the
> spidev device, since it meant I could copy-and-paste the ioctls from example C
> code :)

Thanks for Inline::C :)

But doesn't work on my debian testing/hybrid box at work (but does at
home) (output at end of this email)!

I don't *have* /usr/lib/perl/5.14/CORE/cc_runtime.h on my system.  Debian
last had a cc_runtime.h in the perl 5.10 version in stable.   It's also
not on the system I have at home where Inline::C is working!

Anyone seen or recognise this?

> perl /tmp/inlinec.pl
/usr/bin/perl /usr/share/perl/5.14/ExtUtils/xsubpp  -typemap
/usr/share/perl/5.14/ExtUtils/typemap   inlinec_pl_f5a7.xs >
inlinec_pl_f5a7.xsc && mv inlinec_pl_f5a7.xsc inlinec_pl_f5a7.c
make: *** No rule to make target `/usr/lib/perl/5.14/CORE/cc_runtime.h',
needed by `inlinec_pl_f5a7.o'.  Stop.

A problem was encountered while attempting to compile and install your
Inline
C code. The command that failed was:
  make > out.make 2>&1

The build directory was:
/home/tconnors/_Inline/build/inlinec_pl_f5a7

To debug the problem, cd to the build directory, and inspect the output
files.

 at /tmp/inlinec.pl line 0
INIT failed--call queue aborted.


-- 
Tim Connors

From andrew at sericyb.com.au  Wed May 23 21:43:53 2012
From: andrew at sericyb.com.au (Andrew Pam)
Date: Thu, 24 May 2012 14:43:53 +1000
Subject: [Melbourne-pm] perlsecret - Perl secret operators and constants
Message-ID: <4FBDBC89.3020302@sericyb.com.au>

https://github.com/mirrors/perl/blob/book/perlsecret/pod/perlsecret.pod

Enjoy!

Andrew

From toby.corkindale at strategicdata.com.au  Wed May 23 21:51:09 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Thu, 24 May 2012 14:51:09 +1000
Subject: [Melbourne-pm] BeagleBone Perl library
In-Reply-To: <B43FB168-FD39-47CC-BD20-7002B447E2E6@dd.com.au>
References: <4FBC4AA5.3010005@strategicdata.com.au>
	<B43FB168-FD39-47CC-BD20-7002B447E2E6@dd.com.au>
Message-ID: <4FBDBE3D.60902@strategicdata.com.au>

On 23/05/12 12:45, Scott Penrose wrote:
> Excellent. We should merge. And I gave a domain. Perlbone

I went looking for your code a week or two ago and couldn't find your 
library on github any more. Merging the codebases sounds sensible though.

Although I haven't been trying to clone the bonescript/arduino language, 
so I have a more OO approach.

ie. the arduino-style way would be:

   setPin(13, DIGITAL);
   digitalWrite(13, HIGH);

and my Perlish way is:

   my $pin = BeagleBone::Pins->new('P9_13');
   $pin->digitalWrite(1);

(my code automatically sets the appropriate mode the first time you
call a read/write/etc function on the pin)



-Toby


> On 23/05/2012, at 12:26, Toby Corkindale<toby.corkindale at strategicdata.com.au>  wrote:
>
>> Hi,
>> I've been playing around with a BeagleBone for the past week, and I've come up with some initial Perl libraries to do GPIO and SPI, plus written some drivers to make it easy to talk to a certain LCD panel controller.
>>
>> I know Scott is working with the BB as well, and maybe others?
>> My code so far is at:
>> https://github.com/TJC/BeagleBone
>>
>> And a small blog post about the LCD controller is here:
>> http://blog.dryft.net/2012/05/ssd1306-lcd-controller-in-perl-for.html
>>
>>
>> Unlike on the Arduino, Perl seems like a decent language to work with on the 'bone.. Although I did break out into Inline::C for one part when driving the spidev device, since it meant I could copy-and-paste the ioctls from example C code :)
>>
>>
>> Cheers,
>> Toby
>> _______________________________________________
>> Melbourne-pm mailing list
>> Melbourne-pm at pm.org
>> http://mail.pm.org/mailman/listinfo/melbourne-pm


-- 
.signature

From sam at nipl.net  Thu May 24 20:42:00 2012
From: sam at nipl.net (Sam Watkins)
Date: Fri, 25 May 2012 13:42:00 +1000
Subject: [Melbourne-pm] perlsecret - Perl secret operators and constants
In-Reply-To: <20120525034144.GD27437@opal.nipl.net>
References: <4FBDBC89.3020302@sericyb.com.au>
	<20120525034144.GD27437@opal.nipl.net>
Message-ID: <20120525034200.GE27437@opal.nipl.net>

On Fri, May 25, 2012 at 01:41:44PM +1000, Sam Watkins wrote:
> Andrew Pam wrote:
> > https://github.com/mirrors/perl/blob/book/perlsecret/pod/perlsecret.pod
> > 
> > Enjoy!
> 
> LOL thanks!  I promise not to use them at work.

From andrew at sericyb.com.au  Thu May 24 20:47:09 2012
From: andrew at sericyb.com.au (Andrew Pam)
Date: Fri, 25 May 2012 13:47:09 +1000
Subject: [Melbourne-pm] perlsecret - Perl secret operators and constants
In-Reply-To: <20120525034200.GE27437@opal.nipl.net>
References: <4FBDBC89.3020302@sericyb.com.au>
	<20120525034144.GD27437@opal.nipl.net>
	<20120525034200.GE27437@opal.nipl.net>
Message-ID: <4FBF00BD.5010108@sericyb.com.au>

On 25/05/12 13:42, Sam Watkins wrote:
> LOL thanks!  I promise not to use them at work.

Actually IMHO a few of them are sufficiently useful and well known to be
appropriate for regular use.  I would include:

0+
@{[ ]}
!!
,=>
= <> =~

Possibly also () x !!

Cheers,
	Andrew

From brong at fastmail.fm  Thu May 24 23:15:22 2012
From: brong at fastmail.fm (Bron Gondwana)
Date: Fri, 25 May 2012 08:15:22 +0200
Subject: [Melbourne-pm] perlsecret - Perl secret operators and constants
In-Reply-To: <4FBF00BD.5010108@sericyb.com.au>
References: <4FBDBC89.3020302@sericyb.com.au>
	<20120525034144.GD27437@opal.nipl.net>
	<20120525034200.GE27437@opal.nipl.net>
	<4FBF00BD.5010108@sericyb.com.au>
Message-ID: <20120525061522.GE10110@launde.brong.net>

On Fri, May 25, 2012 at 01:47:09PM +1000, Andrew Pam wrote:
> On 25/05/12 13:42, Sam Watkins wrote:
> > LOL thanks!  I promise not to use them at work.
> 
> Actually IMHO a few of them are sufficiently useful and well known to be
> appropriate for regular use.  I would include:

I got told I was obfuscating code for using:

my $copy = [@$orig];

But that was a screwed up work environment towards the end.

Bron.

From andrew at sericyb.com.au  Thu May 24 23:18:46 2012
From: andrew at sericyb.com.au (Andrew Pam)
Date: Fri, 25 May 2012 16:18:46 +1000
Subject: [Melbourne-pm] perlsecret - Perl secret operators and constants
In-Reply-To: <20120525061522.GE10110@launde.brong.net>
References: <4FBDBC89.3020302@sericyb.com.au>
	<20120525034144.GD27437@opal.nipl.net>
	<20120525034200.GE27437@opal.nipl.net>
	<4FBF00BD.5010108@sericyb.com.au>
	<20120525061522.GE10110@launde.brong.net>
Message-ID: <4FBF2446.5080308@sericyb.com.au>

On 25/05/12 16:15, Bron Gondwana wrote:
> I got told I was obfuscating code for using:
> 
> my $copy = [@$orig];
> 
> But that was a screwed up work environment towards the end.

At a previous employer I was warned against using "map" on the grounds
that junior developers couldn't be expected to understand it.

Thankfully I (successfully) argued that we had to expect at least some
minimal level of competence and that since we already had a mentoring
program in place, anyone who didn't understand map and couldn't figure
it out was welcome to ask me.

Andrew

From toby.corkindale at strategicdata.com.au  Sun May 27 19:31:25 2012
From: toby.corkindale at strategicdata.com.au (Toby Corkindale)
Date: Mon, 28 May 2012 12:31:25 +1000
Subject: [Melbourne-pm] Perl Foundation hacked?
Message-ID: <4FC2E37D.8090700@strategicdata.com.au>

While attempting to visit the perlfoundation.org website, Google Chrome 
told me they're a known malware site, viz:

Of the 177 pages we tested on the site over the past 90 days, 21 page(s) 
resulted in malicious software being downloaded and installed without 
user consent. The last time Google visited this site was on 2012-05-27, 
and the last time suspicious content was found on this site was on 
2012-05-27.

Malicious software includes 262 scripting exploit(s).



From tim.w.connors at gmail.com  Sun May 27 19:50:12 2012
From: tim.w.connors at gmail.com (Tim Connors)
Date: Mon, 28 May 2012 12:50:12 +1000 (EST)
Subject: [Melbourne-pm] perlsecret - Perl secret operators and constants
In-Reply-To: <4FBF2446.5080308@sericyb.com.au>
References: <4FBDBC89.3020302@sericyb.com.au>
	<20120525034144.GD27437@opal.nipl.net>
	<20120525034200.GE27437@opal.nipl.net>
	<4FBF00BD.5010108@sericyb.com.au>
	<20120525061522.GE10110@launde.brong.net>
	<4FBF2446.5080308@sericyb.com.au>
Message-ID: <alpine.DEB.2.00.1205281249490.25700@dirac.rather.puzzling.org>

On Fri, 25 May 2012, Andrew Pam wrote:

> On 25/05/12 16:15, Bron Gondwana wrote:
> > I got told I was obfuscating code for using:
> >
> > my $copy = [@$orig];
> >
> > But that was a screwed up work environment towards the end.
>
> At a previous employer I was warned against using "map" on the grounds
> that junior developers couldn't be expected to understand it.

ie, the boss didn't understand it.  Strewth.

-- 
Tim Connors

From andrew at sericyb.com.au  Sun May 27 19:51:58 2012
From: andrew at sericyb.com.au (Andrew Pam)
Date: Mon, 28 May 2012 12:51:58 +1000
Subject: [Melbourne-pm] perlsecret - Perl secret operators and constants
In-Reply-To: <alpine.DEB.2.00.1205281249490.25700@dirac.rather.puzzling.org>
References: <4FBDBC89.3020302@sericyb.com.au>
	<20120525034144.GD27437@opal.nipl.net>
	<20120525034200.GE27437@opal.nipl.net>
	<4FBF00BD.5010108@sericyb.com.au>
	<20120525061522.GE10110@launde.brong.net>
	<4FBF2446.5080308@sericyb.com.au>
	<alpine.DEB.2.00.1205281249490.25700@dirac.rather.puzzling.org>
Message-ID: <4FC2E84E.2060308@sericyb.com.au>

On 28/05/12 12:50, Tim Connors wrote:
>> At a previous employer I was warned against using "map" on the grounds
>> that junior developers couldn't be expected to understand it.
> 
> ie, the boss didn't understand it.  Strewth.

Actually the boss was the developer team lead.  He understood it just
fine, but had a low opinion of the quality of staff we were hiring.  :)

Andrew