From whjackson at gmail.com  Fri Feb 13 10:36:33 2009
From: whjackson at gmail.com (Whitney Jackson)
Date: Fri, 13 Feb 2009 12:36:33 -0600
Subject: [WindyCity-pm] installing the same module twice
Message-ID: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>

It turns out there are two ways (at least) to build DBD::Sybase:

1) Against FreeTDS Libraries
2) Against Sybase Open Client Libraries

When you choose option 1 you can talk to MSSQL as well as Sybase.
This works b/c both DBs use the TDS protocol.

With option 2 you can only talk to Sybase DBs but you have the ability
to use additional features like bulk writes not supported with
FreeTDS.

My problem is that I need the benefits of both.  When I use DBI to
connect to a Sybase DB I want the module built against the Sybase Open
Client libraries.  On the other hand, a DBI connect against a MSSQL DB
should use the DBD::Sybase built against FreeTDS.

Does anyone know a clean way to install DBD::Sybase twice and have the
right one to load when I do a DBI connect?

Whitney

From jeremy at marzhillstudios.com  Fri Feb 13 13:46:44 2009
From: jeremy at marzhillstudios.com (Jeremy Wall)
Date: Fri, 13 Feb 2009 15:46:44 -0600
Subject: [WindyCity-pm] installing the same module twice
In-Reply-To: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
References: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
Message-ID: <69d143cd0902131346s5db407b0s1990d63fd103848b@mail.gmail.com>

That sounds like a sticky requirement there. Maybe you could hack something
together with fancy @INC hacking but it sounds like you are trying to make
DBD::Sybase do something it really shouldn't be used for. Is there a FreeTDS
module out there you can use for the MSSQL db rather than trying to have two
different varieties sharing the same namespace?

On Fri, Feb 13, 2009 at 12:36 PM, Whitney Jackson <whjackson at gmail.com>wrote:

> It turns out there are two ways (at least) to build DBD::Sybase:
>
> 1) Against FreeTDS Libraries
> 2) Against Sybase Open Client Libraries
>
> When you choose option 1 you can talk to MSSQL as well as Sybase.
> This works b/c both DBs use the TDS protocol.
>
> With option 2 you can only talk to Sybase DBs but you have the ability
> to use additional features like bulk writes not supported with
> FreeTDS.
>
> My problem is that I need the benefits of both.  When I use DBI to
> connect to a Sybase DB I want the module built against the Sybase Open
> Client libraries.  On the other hand, a DBI connect against a MSSQL DB
> should use the DBD::Sybase built against FreeTDS.
>
> Does anyone know a clean way to install DBD::Sybase twice and have the
> right one to load when I do a DBI connect?
>
> Whitney
> _______________________________________________
> WindyCity-pm mailing list
> WindyCity-pm at pm.org
> http://mail.pm.org/mailman/listinfo/windycity-pm
>



-- 
Jeremy Wall
http://jeremy.marzhillstudios.com
Jeremy at marzhillstudios.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.pm.org/pipermail/windycity-pm/attachments/20090213/1bbe353e/attachment.html>

From whjackson at gmail.com  Fri Feb 13 15:04:01 2009
From: whjackson at gmail.com (Whitney Jackson)
Date: Fri, 13 Feb 2009 17:04:01 -0600
Subject: [WindyCity-pm] installing the same module twice
In-Reply-To: <69d143cd0902131346s5db407b0s1990d63fd103848b@mail.gmail.com>
References: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
	<69d143cd0902131346s5db407b0s1990d63fd103848b@mail.gmail.com>
Message-ID: <a9fdc4d30902131504p12448b8vd2ff6184db99cb83@mail.gmail.com>

> Is there a FreeTDS module out there you can use for the MSSQL db rather
> than trying to have two different varieties sharing the same namespace?

This is the list of APIs freetds supports:  http://www.freetds.org/software.html

The only thing that looks like a possibility is DBD::ODBC on top of
unixODBC on top of freetds.  I could try using that solution for
everything that needs FreeTDS.  I don't have any experience with
unixODBC but I suppose it might work.

Alternatively, I could create DBD::FreeTDS by copying DBD::Sybase and
renaming everything appropriately.  That seems like a real pain though
and would of course need to be maintained over time.

Whitney

On Fri, Feb 13, 2009 at 3:46 PM, Jeremy Wall <jeremy at marzhillstudios.com> wrote:
> That sounds like a sticky requirement there. Maybe you could hack something
> together with fancy @INC hacking but it sounds like you are trying to make
> DBD::Sybase do something it really shouldn't be used for. Is there a FreeTDS
> module out there you can use for the MSSQL db rather than trying to have two
> different varieties sharing the same namespace?
>
> On Fri, Feb 13, 2009 at 12:36 PM, Whitney Jackson <whjackson at gmail.com>
> wrote:
>>
>> It turns out there are two ways (at least) to build DBD::Sybase:
>>
>> 1) Against FreeTDS Libraries
>> 2) Against Sybase Open Client Libraries
>>
>> When you choose option 1 you can talk to MSSQL as well as Sybase.
>> This works b/c both DBs use the TDS protocol.
>>
>> With option 2 you can only talk to Sybase DBs but you have the ability
>> to use additional features like bulk writes not supported with
>> FreeTDS.
>>
>> My problem is that I need the benefits of both.  When I use DBI to
>> connect to a Sybase DB I want the module built against the Sybase Open
>> Client libraries.  On the other hand, a DBI connect against a MSSQL DB
>> should use the DBD::Sybase built against FreeTDS.
>>
>> Does anyone know a clean way to install DBD::Sybase twice and have the
>> right one to load when I do a DBI connect?
>>
>> Whitney
>> _______________________________________________
>> WindyCity-pm mailing list
>> WindyCity-pm at pm.org
>> http://mail.pm.org/mailman/listinfo/windycity-pm
>
>
>
> --
> Jeremy Wall
> http://jeremy.marzhillstudios.com
> Jeremy at marzhillstudios.com
>

From craigberry at mac.com  Fri Feb 13 16:24:28 2009
From: craigberry at mac.com (Craig A. Berry)
Date: Fri, 13 Feb 2009 18:24:28 -0600
Subject: [WindyCity-pm] installing the same module twice
In-Reply-To: <a9fdc4d30902131504p12448b8vd2ff6184db99cb83@mail.gmail.com>
References: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
	<69d143cd0902131346s5db407b0s1990d63fd103848b@mail.gmail.com>
	<a9fdc4d30902131504p12448b8vd2ff6184db99cb83@mail.gmail.com>
Message-ID: <657166CE-FA2C-4CB4-836B-F7CFF43DA7AA@mac.com>


On Feb 13, 2009, at 5:04 PM, Whitney Jackson wrote:

>> Is there a FreeTDS module out there you can use for the MSSQL db  
>> rather
>> than trying to have two different varieties sharing the same  
>> namespace?
>
> This is the list of APIs freetds supports:  http://www.freetds.org/software.html

That's not a list of APIs; it's a list of software packages.  The API  
status is provided as a Perl script here:

<http://freetds.cvs.sourceforge.net/viewvc/freetds/freetds/doc/api_status.txt?revision=1.61&view=markup 
 >


> The only thing that looks like a possibility is DBD::ODBC on top of
> unixODBC on top of freetds.  I could try using that solution for
> everything that needs FreeTDS.  I don't have any experience with
> unixODBC but I suppose it might work.

FreeTDS's ODBC driver does work, and I think on most platforms works  
with iODBC as well as unixODBC.  If you have a Sybase Open Client ODBC  
driver and can use the FreeTDS driver for MS SQL Server, then using a  
driver manager to pick different drivers for different databases is a  
well-traveled path.  Loading a different library at run-time based on  
directives in your connection string is the whole point of ODBC driver  
managers.

Whether this "bulk write" feature that is apparently missing when  
using DBD::Sybase with FreeTDS would be supported by any given ODBC  
driver or driver manager is an open question.  Or by DBD::ODBC for  
that matter.  Could you say more about what exactly is missing in  
FreeTDS that is present in the Sybase Open Client libraries?

>>> Does anyone know a clean way to install DBD::Sybase twice and have  
>>> the
>>> right one to load when I do a DBI connect?


You should be able to create two separate installations by building  
twice using two different variations of:

$ perl Makefile.PL INSTALL_BASE=/path/to/somewhere

How to get the right one at run-time depends on whether you want to  
fiddle with PERL5LIB in the environment, "use lib" in your code, -Ilib  
on the command line, etc.

________________________________________
Craig A. Berry
mailto:craigberry at mac.com

"... getting out of a sonnet is much more
  difficult than getting in."
                  Brad Leithauser


From whjackson at gmail.com  Fri Feb 13 17:21:06 2009
From: whjackson at gmail.com (Whitney Jackson)
Date: Fri, 13 Feb 2009 19:21:06 -0600
Subject: [WindyCity-pm] installing the same module twice
In-Reply-To: <657166CE-FA2C-4CB4-836B-F7CFF43DA7AA@mac.com>
References: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
	<69d143cd0902131346s5db407b0s1990d63fd103848b@mail.gmail.com>
	<a9fdc4d30902131504p12448b8vd2ff6184db99cb83@mail.gmail.com>
	<657166CE-FA2C-4CB4-836B-F7CFF43DA7AA@mac.com>
Message-ID: <a9fdc4d30902131721u82b3dkcf4e8b8d999d198b@mail.gmail.com>

> If you have a Sybase Open Client ODBC driver and can use the FreeTDS driver
> for MS SQL Server, then using a driver manager to pick different drivers for different
> databases is a well-traveled path.

I went ahead and installed unixODBC and DBD::ODBC.   As you point out
this is just what I need since it allows you to specify multiple
drivers and refer to the one you need when you configure your DSN.  I
just setup a test DSN and FreeTDS driver.  I'm able to successfully
connect using DBI/DBD::ODBC and run basic queries.  Thanks for the
advice!

> Could you say more about what exactly is missing in FreeTDS that is present in the
> Sybase Open Client libraries?

>From the DBD::Sybase documentation under 'Bulk-Load Functionality':

"NOTE: This feature requires that the libblk.a library be available at
build time. This is not always the case if the Sybase SDK isn't
installed. You can test the $dbh->{syb_has_blk} attribute to see if
the BLK api calls are available in your copy of DBD::Sybase."

This feature is labelled "Experimental" but I've had pretty good luck
with it so far.  It pretty much allows you to do the equivalent of a
bcp without having to create a bcp file and run an external command.

The other issue for me, is that it may be harder to get support from
my DBAs and ultimately Sybase for some issues when using FreeTDS.
That's just theoretical though.  I've been using FreeTDS for a while
now and it's never been an issue.

Thanks again for the help!

Whitney

On Fri, Feb 13, 2009 at 6:24 PM, Craig A. Berry <craigberry at mac.com> wrote:
>
> On Feb 13, 2009, at 5:04 PM, Whitney Jackson wrote:
>
>>> Is there a FreeTDS module out there you can use for the MSSQL db rather
>>> than trying to have two different varieties sharing the same namespace?
>>
>> This is the list of APIs freetds supports:
>>  http://www.freetds.org/software.html
>
> That's not a list of APIs; it's a list of software packages.  The API status
> is provided as a Perl script here:
>
> <http://freetds.cvs.sourceforge.net/viewvc/freetds/freetds/doc/api_status.txt?revision=1.61&view=markup>
>
>
>> The only thing that looks like a possibility is DBD::ODBC on top of
>> unixODBC on top of freetds.  I could try using that solution for
>> everything that needs FreeTDS.  I don't have any experience with
>> unixODBC but I suppose it might work.
>
> FreeTDS's ODBC driver does work, and I think on most platforms works with
> iODBC as well as unixODBC.  If you have a Sybase Open Client ODBC driver and
> can use the FreeTDS driver for MS SQL Server, then using a driver manager to
> pick different drivers for different databases is a well-traveled path.
>  Loading a different library at run-time based on directives in your
> connection string is the whole point of ODBC driver managers.
>
> Whether this "bulk write" feature that is apparently missing when using
> DBD::Sybase with FreeTDS would be supported by any given ODBC driver or
> driver manager is an open question.  Or by DBD::ODBC for that matter.  Could
> you say more about what exactly is missing in FreeTDS that is present in the
> Sybase Open Client libraries?
>
>>>> Does anyone know a clean way to install DBD::Sybase twice and have the
>>>> right one to load when I do a DBI connect?
>
>
> You should be able to create two separate installations by building twice
> using two different variations of:
>
> $ perl Makefile.PL INSTALL_BASE=/path/to/somewhere
>
> How to get the right one at run-time depends on whether you want to fiddle
> with PERL5LIB in the environment, "use lib" in your code, -Ilib on the
> command line, etc.
>
> ________________________________________
> Craig A. Berry
> mailto:craigberry at mac.com
>
> "... getting out of a sonnet is much more
>  difficult than getting in."
>                 Brad Leithauser
>
>

From jon at jrock.us  Fri Feb 13 20:08:20 2009
From: jon at jrock.us (Jonathan Rockway)
Date: Fri, 13 Feb 2009 22:08:20 -0600
Subject: [WindyCity-pm] installing the same module twice
In-Reply-To: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
	(Whitney Jackson's message of "Fri, 13 Feb 2009 12:36:33 -0600")
References: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
Message-ID: <87ab8p4q4r.fsf@bar.jrock.us>

* On Fri, Feb 13 2009, Whitney Jackson wrote:
> Does anyone know a clean way to install DBD::Sybase twice and have the
> right one to load when I do a DBI connect?

Search and replace? (i.e. replace Sybase with Sybase2, and build Sybase2
differently; now you have DBD::Sybase and DBD::Sybase2).

If you don't need both in the same app, then just use local::lib to
install them to different places, then use local::lib to pick the right
one at runtime.

--
print just => another => perl => hacker => if $,=$"

From craigberry at mac.com  Fri Feb 13 21:03:04 2009
From: craigberry at mac.com (Craig A. Berry)
Date: Fri, 13 Feb 2009 23:03:04 -0600
Subject: [WindyCity-pm] installing the same module twice
In-Reply-To: <a9fdc4d30902131721u82b3dkcf4e8b8d999d198b@mail.gmail.com>
References: <a9fdc4d30902131036i5fe80ed3i8bc93bd7210672cf@mail.gmail.com>
	<69d143cd0902131346s5db407b0s1990d63fd103848b@mail.gmail.com>
	<a9fdc4d30902131504p12448b8vd2ff6184db99cb83@mail.gmail.com>
	<657166CE-FA2C-4CB4-836B-F7CFF43DA7AA@mac.com>
	<a9fdc4d30902131721u82b3dkcf4e8b8d999d198b@mail.gmail.com>
Message-ID: <246C8FD0-9277-46F6-920A-6B4AE252193E@mac.com>


On Feb 13, 2009, at 7:21 PM, Whitney Jackson wrote:

>> Could you say more about what exactly is missing in FreeTDS that is  
>> present in the
>> Sybase Open Client libraries?
>
> From the DBD::Sybase documentation under 'Bulk-Load Functionality':
>
> "NOTE: This feature requires that the libblk.a library be available at
> build time. This is not always the case if the Sybase SDK isn't
> installed. You can test the $dbh->{syb_has_blk} attribute to see if
> the BLK api calls are available in your copy of DBD::Sybase."
>

And it claims to support this by using the CS_BULK_LOGIN attribute.   
If you grep the FreeTDS sources for CS_BULK_LOGIN, you'll definitely  
find something that looks like support.  If some aspect of it doesn't  
work, or if the DBD::Sybase configuration process doesn't detect it  
correctly, please report it to the relevant list.  At first blush it  
really looks like the Makefile.PL for DBD::Sybase is at fault:

http://cpansearch.perl.org/src/MEWP/DBD-Sybase-1.09/Makefile.PL

You'll see that it configures with -DNO_BLK=1 unless it finds a  
library named blk or libblk or sybblk.  I could be wrong and I don't  
have an installation handy to check, but IIRC, FreeTDS just includes  
the bcp API as part of dblib (or libsybdb?) and does not put it in a  
separate library.  If you can confirm that please report it on RT at:

http://rt.cpan.org/Public/Dist/Display.html?Name=DBD-Sybase

> This feature is labelled "Experimental" but I've had pretty good luck
> with it so far.  It pretty much allows you to do the equivalent of a
> bcp without having to create a bcp file and run an external command.


FreeTDS definitely implements the bcp API:

http://www.freetds.org/reference/a00277.html

________________________________________
Craig A. Berry
mailto:craigberry at mac.com

"... getting out of a sonnet is much more
  difficult than getting in."
                  Brad Leithauser


From joshua.mcadams at gmail.com  Tue Feb 17 08:07:46 2009
From: joshua.mcadams at gmail.com (Joshua McAdams)
Date: Tue, 17 Feb 2009 16:07:46 +0000
Subject: [WindyCity-pm] Chicago.pm Meeting February 24th - Using Perl to
	Manage Software Application Delivery
Message-ID: <000e0cd47b0605a4b304631f81a9@google.com>

If you have trouble viewing or submitting this form, you can fill it out  
online:
https://spreadsheets.google.com/viewform?formkey=cE9DbVRDTjZWV0RhZnZfVHdEaGg5Snc6MA..


Chicago.pm Meeting February 24th - Using Perl to Manage Software  
Application Delivery


Why Java is the wrong language for building and deploying Java applications  
and why Perl is probably the best? Sean Blanton will explain how OpenMake  
Software uses Perl to build tens of thousands of Java , .NET, C/C++, COBOL  
business applications in the financial, retail, transportation and software  
industries and will illustrate how Perl is used for software configuration  
management by IBM, CA and open source projects.

Sean works for OpenMake, a successful local software company, large Perl  
user, and a YAPC Chicago sponsor and supporter.

The meeting will start at 7pm and will be on the 17th floor of the Google  
office at 20 W Kinzie. Please RSVP so that you don't get hassled at the  
door :)


Name



Powered by Google Docs

Terms of Service - Additional Terms

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.pm.org/pipermail/windycity-pm/attachments/20090217/ae2022f5/attachment.html>

From joshua.mcadams at gmail.com  Mon Feb 23 20:45:46 2009
From: joshua.mcadams at gmail.com (Joshua McAdams)
Date: Mon, 23 Feb 2009 22:45:46 -0600
Subject: [WindyCity-pm] Fwd: Chicago.pm Meeting February 24th - Using Perl
	to Manage Software Application Delivery
In-Reply-To: <000e0cd47b0605a4b304631f81a9@google.com>
References: <000e0cd47b0605a4b304631f81a9@google.com>
Message-ID: <49d805d70902232045w7539c06j7975a55a4b0a000f@mail.gmail.com>

Just a reminder, the meeting is Tuesday.  Please RSVP by noon!


---------- Forwarded message ----------
From: Joshua McAdams <joshua.mcadams at gmail.com>
Date: Tue, Feb 17, 2009 at 10:07 AM
Subject: [WindyCity-pm] Chicago.pm Meeting February 24th - Using Perl
to Manage Software Application Delivery
To: windycity-pm at pm.org


If you have trouble viewing or submitting this form, you can fill it out online:
https://spreadsheets.google.com/viewform?formkey=cE9DbVRDTjZWV0RhZnZfVHdEaGg5Snc6MA..

Chicago.pm Meeting February 24th - Using Perl to Manage Software
Application Delivery

Why Java is the wrong language for building and deploying Java
applications and why Perl is probably the best? ? Sean Blanton will
explain how OpenMake Software uses Perl to build tens of thousands of
Java , .NET, C/C++, COBOL business applications in the financial,
retail, transportation and software industries and will illustrate how
Perl is used for software configuration management by IBM, CA and open
source projects.

Sean works for OpenMake, a successful local software company, large
Perl user, and a YAPC Chicago sponsor and supporter.

The meeting will start at 7pm and will be on the 17th floor of the
Google office at 20 W Kinzie. ?Please RSVP so that you don't get
hassled at the door :)

________________________________
Name

Powered by Google Docs

Terms of Service - Additional Terms

_______________________________________________
WindyCity-pm mailing list
WindyCity-pm at pm.org
http://mail.pm.org/mailman/listinfo/windycity-pm

From jon at jrock.us  Tue Feb 24 03:38:30 2009
From: jon at jrock.us (Jonathan Rockway)
Date: Tue, 24 Feb 2009 05:38:30 -0600
Subject: [WindyCity-pm] [Chicago-talk] Fwd: Chicago.pm Meeting February
	24th - Using Perl to Manage Software Application Delivery
In-Reply-To: <49d805d70902232045w7539c06j7975a55a4b0a000f@mail.gmail.com>
	(Joshua McAdams's message of "Mon, 23 Feb 2009 22:45:46 -0600")
References: <000e0cd47b0605a4b304631f81a9@google.com>
	<49d805d70902232045w7539c06j7975a55a4b0a000f@mail.gmail.com>
Message-ID: <871vtovzd5.fsf@bar.jrock.us>

* On Mon, Feb 23 2009, Joshua McAdams wrote:
> Just a reminder, the meeting is Tuesday.  Please RSVP by noon!

FWIW, if the OpenMake talk doesn't fill the time, I will give my KiokuDB
talk.  It is only about 20 minutes.

Regards,
Jonathan Rockway

--
print just => another => perl => hacker => if $,=$"

From joshua.mcadams at gmail.com  Tue Feb 24 21:23:32 2009
From: joshua.mcadams at gmail.com (Joshua McAdams)
Date: Tue, 24 Feb 2009 23:23:32 -0600
Subject: [WindyCity-pm] </chicago-pm-meeting>
Message-ID: <49d805d70902242123i40353d7fo62ccce66489c4c35@mail.gmail.com>

Whew, another Chicago.pm meeting is done.  From what I heard, this was
the best Chicago.pm meeting this month, so we have something to be
proud of!

We had two great presentations.  Kicking things off was Jon Rockway
with a talk about KiokuDB (http://www.iinteractive.com/kiokudb/) and
closing out the show was Sean Blanton with a talk about OpenMake
(http://www.openmakesoftware.com/) and how they use Perl (without
strict!) to drive massive build systems.

We also handed out a few books procured by Pete Krawczyk from the
folks at O'Reilly.  The books this time around were "Visualizing Data"
(http://oreilly.com/catalog/9780596514556/) and "Intellectual Property
and Open Source" (http://oreilly.com/catalog/9780596517960).  Thanks
O'Reilly!

After the meeting we went down the block to Dick's Last Resort were we
had a few drinks on OpenMake (thanks Sean!) and were subjected to
music from what had to be the luckiest band ever... nobody who sounded
like that should ever get a gig :)  I guess that's what we get for
having a meeting on Mardi Gras, or as Pete calls it, Paczki Day.

That makes two great Chicago Perl meet-ups in a row this year.  Let's
keep the streak going.  Our next meeting is Tuesday March 24th and
will most likely be held at the Google offices again at 20 W Kinzie.
There are no topics yet, though it was mentioned that it was high time
for the chart-off to actually happen.  Also, there is rumor that
Steven Little, creator of Moose, will be on-hand next time around...
we can always hope.

Comments, criticisms, and suggestions welcome.  Please send them to
spam at example.com.

Josh

From sean at blanton.com  Wed Feb 25 08:34:17 2009
From: sean at blanton.com (Sean Blanton, Ph.D.)
Date: Wed, 25 Feb 2009 10:34:17 -0600
Subject: [WindyCity-pm] In partial defense of no 'use strict'
Message-ID: <f5cbb6b60902250834x4d83987at96001e38db0719fd@mail.gmail.com>

In partial defense, the Perl stubs are not whole scripts and variable
declarations may be prepended later. So 'use strict' in the stub will always
cause lots of errors if you run a 'perl -c' on the stub. When editing the
stub by itself, it is still useful in to run 'perl -c' to catch other
errors.
OK, the header could have 'use strict'...

I always use strict in my normal code - saves time, plain and simple.

Sean

On Tue, Feb 24, 2009 at 11:23 PM, Joshua McAdams
<joshua.mcadams at gmail.com>wrote:

> Whew, another Chicago.pm meeting is done.  From what I heard, this was
> the best Chicago.pm meeting this month, so we have something to be
> proud of!
>
> We had two great presentations.  Kicking things off was Jon Rockway
> with a talk about KiokuDB (http://www.iinteractive.com/kiokudb/) and
> closing out the show was Sean Blanton with a talk about OpenMake
> (http://www.openmakesoftware.com/) and how they use Perl (without
> strict!) to drive massive build systems.
>
> We also handed out a few books procured by Pete Krawczyk from the
> folks at O'Reilly.  The books this time around were "Visualizing Data"
> (http://oreilly.com/catalog/9780596514556/) and "Intellectual Property
> and Open Source" (http://oreilly.com/catalog/9780596517960).  Thanks
> O'Reilly!
>
> After the meeting we went down the block to Dick's Last Resort were we
> had a few drinks on OpenMake (thanks Sean!) and were subjected to
> music from what had to be the luckiest band ever... nobody who sounded
> like that should ever get a gig :)  I guess that's what we get for
> having a meeting on Mardi Gras, or as Pete calls it, Paczki Day.
>
> That makes two great Chicago Perl meet-ups in a row this year.  Let's
> keep the streak going.  Our next meeting is Tuesday March 24th and
> will most likely be held at the Google offices again at 20 W Kinzie.
> There are no topics yet, though it was mentioned that it was high time
> for the chart-off to actually happen.  Also, there is rumor that
> Steven Little, creator of Moose, will be on-hand next time around...
> we can always hope.
>
> Comments, criticisms, and suggestions welcome.  Please send them to
> spam at example.com.
>
> Josh
> _______________________________________________
> Chicago-talk mailing list
> Chicago-talk at pm.org
> http://mail.pm.org/mailman/listinfo/chicago-talk
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.pm.org/pipermail/windycity-pm/attachments/20090225/694b9e74/attachment.html>

From Brian.Knight at us.mizuho-sc.com  Fri Feb 27 09:19:02 2009
From: Brian.Knight at us.mizuho-sc.com (Knight, Brian)
Date: Fri, 27 Feb 2009 12:19:02 -0500
Subject: [WindyCity-pm] Expect-style interaction between processes
Message-ID: <F6F6C35ABC6EA847A76D87680DD60FBD01AA56EF@EXMAIL.usi.mizuho-sc.com>

I am working on a project which involves pulling a configuration file off of
a legacy system through an RS-232 connection.  For transferring that file,
the legacy system supports only the XMODEM-CRC file transfer protocol.

The legacy system is attached to our serial console server, which is far more
modern and supports SSH.  The host running the scripting tasks is a RHEL 4
box.

Unfortunately the XMODEM-CRC part is not negotiable as it's a legacy system,
but the SSH connection could be converted to telnet in a pinch.

There's also some amount of send/expect scripting to do in order to tell the
system to transfer that file.  That part is easy enough.  Where I am having
trouble is, when the system is ready to send the file, I can't hand-off that
SSH connection to the XMODEM receive process.  Expect 5.42.1's "interact -u"
behavior seems to be somewhat broken, and I'm looking for another way to do
it.

Another facet of the problem is that, once the XMODEM receive is done, I need
the script to retake control of the SSH process to log off of the legacy box.

But the problem isn't with SSH; it's definitely with "interact -u".  I have
an Expect test script that literally connects a local XMODEM send process
with a local XMODEM receive process, and it doesn't work:
---------------------------------------
spawn sx -q testfile1.bin
set term_id $spawn_id
expect "XMODEM"
spawn rx -qy testfile2.bin
expect -re "rx: ready to receive.*\r\n"
interact -u $term_id
set spawn_id $term_id
---------------------------------------

I'm able to run "mkfifo ./testpipe; sx -q testfile1.bin < testpipe | rx -q
testfile2.bin > testpipe" and that works perfectly fine, so I know my XMODEM
tools aren't broken.

Any pointers on Perl modules or methods that might help attach and detach
these processes, or might help with the entire task?

Thanks,

-Brian Knight
Sr. Network Engineer
Mizuho Securities USA, Futures Division
http://www.mizuho-sc.com/

* Please note that I do not speak for my employer - only for myself.

CONFIDENTIAL: This e-mail, including its contents and attachments,
if any, are confidential. It is neither an offer to buy or sell,
nor a solicitation of an offer to buy or sell, any securities or
any related financial instruments mentioned in it. If you are not
the named recipient please notify the sender and immediately delete
it. You may not disseminate, distribute, or forward this e-mail
message or disclose its contents to anybody else. Unless otherwise
indicated, copyright and any other intellectual property rights in
its contents are the sole property of Mizuho Securities USA Inc.
     E-mail transmission cannot be guaranteed to be secure or
error-free. The sender therefore does not accept liability for any
errors or omissions in the contents of this message which arise as
a result of e-mail transmission.  If verification is required
please request a hard-copy version.
     Although we routinely screen for viruses, addressees should
check this e-mail and any attachments for viruses. We make no
representation or warranty as to the absence of viruses in this
e-mail or any attachments. Please note that to ensure regulatory
compliance and for the protection of our customers and business, we
may monitor and read e-mails sent to and from our server(s).
#####################################################################################

From frag at ripco.com  Fri Feb 27 13:21:52 2009
From: frag at ripco.com (Mike Fragassi)
Date: Fri, 27 Feb 2009 15:21:52 -0600 (CST)
Subject: [WindyCity-pm] Expect-style interaction between processes
In-Reply-To: <F6F6C35ABC6EA847A76D87680DD60FBD01AA56EF@EXMAIL.usi.mizuho-sc.com>
References: <F6F6C35ABC6EA847A76D87680DD60FBD01AA56EF@EXMAIL.usi.mizuho-sc.com>
Message-ID: <Pine.GSO.4.61.0902271433470.14353@shell3>


On Fri, 27 Feb 2009, Knight, Brian wrote:

> Any pointers on Perl modules or methods that might help attach and detach
> these processes, or might help with the entire task?

My Expect fu is marginal, but in the past when it came up I used 
Expect::Simple.  Also I've recently had success with using 
Net::SFTP::Foreign, Net::OpenSSH, and IPC::Open3::Simple.  For a 
nigh-complete list of IO capturing modules (although it probably isn't 
what you're really looking for), see:

http://search.cpan.org/~dagolden/Capture-Tiny-0.04/lib/Capture/Tiny.pod#SEE_ALSO

Finally, I ran a quick search and came across this on Perlmonks, 
which might be relevant:

   Module Net::SSH::Expect - Cannot interact
   http://www.perlmonks.org/?node_id=683629

-- Mike F.

From joshua.mcadams at gmail.com  Sat Feb 28 15:46:34 2009
From: joshua.mcadams at gmail.com (Joshua McAdams)
Date: Sat, 28 Feb 2009 17:46:34 -0600
Subject: [WindyCity-pm] Fwd: [Chicago] Flourish 2009 Registration Now Open
In-Reply-To: <1235102418.9778.76.camel@stox.dyndns.org>
References: <1235102418.9778.76.camel@stox.dyndns.org>
Message-ID: <49d805d70902281546t5dac1b39s5040402b2e5729b6@mail.gmail.com>

Not sure if this has beet passed to the list yet, but Flourish is
coming up and the registration is open....


FLOURISH 2009 BRINGS TOGETHER OPEN-SOURCE EXPERTS
FROM AROUND CHICAGO, THE MIDWEST, AND AROUND THE
WORLD



WHAT

? ? ? ?Flourish 2009


? ? ? ?Flourish is an annual conference dedicated
? ? ? ?to bringing together experts and enthusiasts
? ? ? ?in Free, Libre, Open Source Software (FLOSS).

? ? ? ?Attendees of Flourish have an opportunity to
? ? ? ?see how using, creating and supporting FLOSS
? ? ? ?can enhance their careers, businesses and
? ? ? ?academic aspirations.

? ? ? ?Admission always free.

WHO

The University of Illinois at Chicago's ACM and IEEE chapters, with
sponsorship from Motorola, Sun Microsystems, Orbitz, and others.

WHY

- Key open-source topics such as Google Android, Kubuntu, the Linux
kernel, open mapping, user interfaces, music collaboration, and
energy-efficient supercomputing.


- Speakers like Ruby On Rails creator David Heinemeier Hansson,
open-source law expert Daliah Sapers, and kernel hacker Christoph
Lameter,

- In-depth workshops on Erlang, Drupal, OpenSolaris, and Processing.


WHEN

? ? ? ?April 3 - 4, 2009
? ? ? ?9am - 5pm

WHERE

? ? ? ?UIC Student Center East
? ? ? ?750 South Halsted

HOW

Register at:

http://www.flourishconf.com/register

Follow updates via Twitter: http://twitter.com/flourishconf/