Oslo QA Hackathon 2008 : Attendees

From Perl QA
Jump to: navigation, search

Contents

Overview

To register, please add yourself to the "nutshell" table below and add some background about yourself in the "Confirmed Attendees" section.

For other info about the event, see the Oslo QA Hackathon 2008 page.

Attendees in a nutshell

Name alias interested in
Adam Kennedy Alias PITA, PPI, Strawberry April release, toolchain
Adrian Howard adrianh Test::Class, TDD, agile, UX
Andy Armstrong TAP:Parser
Anton Berezin freebsd perl
brian d foy
Curtis Poe Ovid TAP::Parser
Dag-Erling Smørgrav des FreeBSD, continuous integration
David Golden xdg CPAN::Reporter, smoking, toolchain
Gabor Szabo TAP, p5p
H. Merijn Brand Tux perl Configure
Jeremiah Foster debian-perl
Jos Boumas kane CPANPLUS
Lars Thegler tagg freebsd perl
Mark Overmeer MarkOv CPAN6
Michael Peters wonko Smolder
Michael Schwern TAP, Test::, toolchain
Peter Makholm debian-perl
Ricardo Signes rjbs CPAN::, Test::, toolchain
Thomas Klausner domm CPANTS
Jonas B. Nielsen jonasbn Module::Build, Test::Builder and automated smoke testing
Cosimo Streppone cosimo Test::Class, Selenium, PITA, Perl6 test suite, Win32::API in Strawberry :)
Nadim Khemir nkh Test::Cookbook, Process automation, coverage and development tools in general
Trond Michelsen trondmm
Øystein Torget The revised TAP specification and a GUI for displaying TAP results. Specifically some sort of web interface.

Confirmed Attendees

Jeremiah Foster <jeremiah@jeremiahfoster.com>

Talk: Squeezing CPAN into debian

Description
How do you get thousands of perl modules to play nicely with thousands of debian packages?
The debian-perl group integrates perl source code from CPAN into the wider debian distribution allowing users to easily handle dependencies and installation. Relying on a relatively standardized interface (CPAN) the group uses tools, often written in perl, to create packages that not only fit the debian policies on 'freedom' and license but also are easy to manage. Active maintenance and bug tracking are included in this process allowing debian to be up-to-date with a wide selection of packaged perl modules.
Length
20 to 30 minutes
Speaker
Jeremiah C. Foster. Member of the debian-perl team.
Audience
This should appeal to anyone who is interested in how the CPAN interface is used, what requirements debian has for software, and those who wish to integrate their perl modules into another distribution system.
Location
Cyberspace or wherever; have plane ticket, will travel.
Availability
I think this will be fun so I am funding my own damn self!

Peter Makholm <peter@makholm.net>

  • Member of the debian-perl packaging team in debian.
  • http://peter.makholm.net/
  • Interested in (in order of interest)
    • Discussing packaging modules for Debian and other distributions
    • learning more about testing and QA for internal use
    • Being a butterfly
 Depending on the other Debian packagers I would be able to talk about
 packageing perl modules for Debian. But others (Jeremiah Foster) would
 probably be better to talk about the workings of the Debian Perl Group
 as a group.
 
 //Makholm

Thomas Klausner <domm@cpan.org>

  • CPANTS / Vienna.pm
 [...] I'm intrigued :-)
 
 For one thing, I'd love to come and talk about CPANTS (and I'd love even
 more to have some dedicated time to hack on CPANTS :-). The only thing
 that keeps me from a completet commitment is the fact that I'd have to
 fly to Oslo. And I don't like flying (for ecologic reasons...).

Just bought my plane tickets (and now I'm looking for one of those we-plant-trees-for-your-co2-thingies...)

H. Merijn 'Tux' Brand <h.m.brand@xs4all.nl> (Committed and sponsored)

  • Designed and created the bleadperl smoke process, which is currently maintained by Abe
  • Configure-pumpking: maintains Configure for perl-devel and perl-maint
  • maintains the defined-or patch for perl-maint, which the FreeBSD Perl packaging guy includes for perl on FreeBSD
  • Can give free software-related talks about (provided I can attend):
    • The perl smoke process
    • The perl configuration
    • Converting SCCS to git :p
  • http://cpansearch.perl.org/~hmbrand/
  • http://mirrors.develooper.com/hpux/
  • Emailed and chatted on IRC. Positive feedback. Sponsored by Procura BV.

Talk: Testing perl development: Test::Smoke

Description
The development branch of perl is in a constant state of flux. With several people providing patches for very different corners of the perl language ranging from new features to bug fixes and from performing improvements to documentation changes, there are many things that can go wrong. The people that provide the patch are likely to have tested it on their system, in their OS and with their choice of configuration. But will it still work in a 64bit threaded environment on VMS?
The Test
:Smoke process tries to build the latest state of perl in as many as possible ways on the given architecture, and reports the results back to the perl5 development group, who see immediately if their works has broken things on the smoker's boxes.
Length
The presentation typically lasts for about 40 minutes. This can be reduced or extended as needed.
Speaker
H.Merijn (Tux) works with PROCURA B.V. in the Netherlands, a small company that does ICT for local government. He is the current holder of the Configure pumkin, and in that role responbsible for the wellbeing of the configuration process of perl.
Audience
This presentation will appeal to anyone with an interest in distributed testing and/or automated testing within many different installation environments.
Location
The Netherlands
Availability
Awaiting confirmation that my employer will be able to fund the trip. If not then I would appreciate funding to get me there. I would fly from Schiphol.


Curtis 'Ovid' Poe <curtis_ovid_poe@yahoo.com> (Committed and sponsored)

Talk: Turbo Charge Your Test Suites

Description
If your tests run faster, you're more likely to run them. If your tests are easier to run, you're more likely to rum them. If your tests are easier to write, you're more likely to write them. I will show you how to accomplish all three.
Length
45 minutes, give or take 5, depending on the audience.
Audience
Intermediate-to-advanced Perl programmers.

This follows a case study of a BBC project which was hampered by a test suite with over 10,000 tests taking almost an hour and a half to run. Pulling out all the stops, we reduced our run time to 10 minutes and in the process, made our tests easier to write.

  • Make your tests run faster.
    • Don't drop and recreate your database: rebuild it.
    • Heresy for everyone! Identifying "ok to fail" tests and skip them.
    • Run your tests in a single process.
      • Manual Aggregation.
      • Test::Aggregate.
      • Test::Class.
  • Make your programmers run faster.
    • Create a custom Test::More.
    • Refactor your tests.
    • Integrate your tests with your editor.
    • Use advanced features of the new Test::Harness.
    • Steal xUnit ideas

Adam Kennedy <cpan@ali.as> (Committed and sponsored)

  • First suggested by Ricardo Signes
  • Practical Image Testing Architecture developer, Perl toolchain hacker. The guy with the most modules on CPAN. Works also on improving the state of Perl on Windows with <http://strawberryperl.com/>.
  • Can give free software-related talks about:
    • PPI (implementing a provably impossible parser)
    • PITA (virtualised testing)
    • Nothing Can Possibly Go Wrong (reasons concepts/designs fail) Maybe the best talk for this conference. Video at <http://lca2007.linux.org.au/talk/122.html>
    • Toolchainy stuff
    • Efficiently running many many projects
  • http://search.cpan.org/~adamk/
  • Chatted on IRC. Sponsored by [1]

Talk: Nothing Can Possibly Go Wrong

Title

Nothing Can Possibly Go Wrong

Description

Every software developer and systems designer make mistakes. Most of the time they can be fixed.

But as you gain more experience and your projects get larger and larger you enter the dangerous and exotic world of large-scale design, with mistakes so seductive you won’t notice until years after you make them, and so dangerous you can’t risk fixing them even once you know about them.

Based on knowledge from the worlds of engineering, psychology, economics, and biology, this talk takes a break-neck tour of some of the worst and most seductive software and systems design mistakes, with humorous examples from Open Source and the Real World (including mine).

Speaker

Adam Kennedy is one of Perl's "Beautiful People". Brilliant and stunningly handsome, yet still modest, he particularly enjoys being asked to write his own biographies for conferences.

Adam is the author of the "theoretically impossible" Perl Parsing Interface (PPI), the first fully Open Source Perl distribution for Windows, and has released more modules on the Central Perl Archive Network than any other Perl developer. He is a member of the CPAN administration team, helps drives the evolution of the Perl toolchain, and is involved with a number of other Perl efforts. By day he works at his Perl consulting company Phase N helping a variety of companies solve difficult and "impossible" problems.

Audience

This talk will be useful to technical or policy people at any level. It does not require any specific knowledge of programming, although people with an IT background may appreciate some of the examples more. It was intentionally created to be adaptable to audiences of varying types.

Preview

http://mirror.linux.org.au/pub/linux.conf.au/2007/video/talks/122.ogg

Talk: PITA - Ridiculously Large Scale Testing

Title
PITA - Ridiculously Large Scale Testing
Description

The Perl testing modules, unified through TAP (the Test Anything Protocol) have been hugely successful.

But one large problem remains. Perl runs on 100+ platforms, and could be any of 20+ versions on each. This incredible diversity means that it has become a massive task to test even a single module on a reasonable set of platforms.

This talk covers the Practical Image Testing Architecture (PITA), an automated mass-testing toolkit based on the concept of completely virtualised testing. The goal, to test any variation of any software package, in any language, on any operating system, on any hardware. And then to test everything, everywhere.

Speaker

Adam Kennedy is one of Perl's "Beautiful People". Brilliant and stunningly handsome, yet still modest, he particularly enjoys being asked to write his own biographies for conferences.

Adam is the author of the "theoretically impossible" Perl Parsing Interface (PPI), the first fully Open Source Perl distribution for Windows, and has released more modules on the Central Perl Archive Network than any other Perl developer. He is a member of the CPAN administration team, helps drives the evolution of the Perl toolchain, and is involved with a number of other Perl efforts. By day he works at his Perl consulting company Phase N helping a variety of companies solve difficult and "impossible" problems.

Audience

This talk will be useful to people with an interest and knowledge of software testing, and/or an interest and knowledge of virtualisation.

Talk: Parsing, Analysing and Manipulating Perl (without perl)

Title
Parsing, Analyzing, and Manipulating Perl (without Perl)
Description

"Only Perl can parse Perl."

For over a decade the creation of a Perl parser written in Perl, without having to use the underlying interpreter itself, was a glittering, wonderful, intangible, provably-impossible dream...

Starting with the reasons why a Perl parser was only "mostly impossible", this talk provides an introduction to the pure-Perl Perl parser PPI, and explains how it was created.

PPI now serves as the basis for a number of now-essential Quality Assurance tools, such as Perl Critic.

The talk will also briefly cover the main PPI APIs, show some examples of it in action, and demonstrate some of the amazing things that can be done with it, both good AND evil.

Speaker

Adam Kennedy is one of Perl's "Beautiful People". Brilliant and stunningly handsome, yet still modest, he particularly enjoys being asked to write his own biographies for conferences.

Adam is the author of the "theoretically impossible" Perl Parsing Interface (PPI), the first fully Open Source Perl distribution for Windows, and has released more modules on the Central Perl Archive Network than any other Perl developer. He is a member of the CPAN administration team, helps drives the evolution of the Perl toolchain, and is involved with a number of other Perl efforts. By day he works at his Perl consulting company Phase N helping a variety of companies solve difficult and "impossible" problems.

Audience

This talk will be primarily of interest to Perl programmers, but may also be of interest to other software developers.


Andy Armstrong <andy@hexten.net>


Michael Schwern <schwern@pobox.com>

  • Perl "toolchain" developer and more.
  • Can give free software-related talks about:
    • [TAP]
    • ExtUtils::MakeMaker
    • META.yml
    • Perl QA and Testing
    • Introductory human design principles
  • +1 412 608 3812 (please send text messages, international roaming is expensive)
  • http://search.cpan.org/~mschwern/
 I can think of ongoing issues we can hammer out:
 
   Furthering Module::Build
   Simplifying CPAN module installation
   Module metadata issues
   Radical departures from the current system
 
 As for talking with the toolchain developers from other projects, that would
 be great.  Perl has always had trouble coordinating its releases and modules
 with other projects, there doesn't seem to be a lot of cross-communication and
 we do things differently than everyone else.
 
 In particular, I have an idea to make life easier for packagers of Perl
 modules by developing an autoconf adapter for installing Perl modules so
 packagers don't have to write so much special case code.
 If you can get me there, I'll come.
 --Schwern 20:32, 27 November 2007 (UTC)

Talk: Coordinating Reusable Code Without Centralization

Title
Coordinating Reusable Code Without Centralization
Description

Municipal IT departments could benefit sharing high quality, reusable software. But how does that get built? How do you encourage an individual IT department to go the extra step and release their work? And how do you find it? And how do you share it? The impulse might be to set up a standards committee and work out protocols, but that all gets very top heavy and there's a good chance agreement will never be reached. The programmers who actually need the code will probably never even be involved.

CPAN (the Comprehensive Perl Archive Network) solves this problem by instead creating a meritocracy, where the ones who do the work are the ones who have the power to solve the problem. There are no standards for code, bad code is simply ignored by the community of users. All APIs and protocols are worked out as needed between the authors of the code and their users. The technology is extremely simple and pluggable so that interested parties can add extra services (such as unified bug tracking, reviews and automatic testing) without the approval of a central authority.

It is a model that fosters a mutual community of sharing at the level of the people doing the work, not by forcing them, but by exposing them and their work to other people, ideas and commentary. And since it's more a mindset then a technology it should be useful outside the domain of software development.

Speaker
Michael G Schwern works on the code that holds CPAN together. He trains programmers to do their own QA and is currently examining issues of communications amongst programmers.
Audience
This talk will be useful to programmers at any level and also people who make technical policy. It does not require any knowledge of a specific programming language. It can be tailored for very different audiences.
Availability
100%, even missing the flight won't stop me. :)
Location
Portland, OR, USA

Ricardo Signes <rjbs@cpan.org>

  • Module::Starter, CPAN::Mini developer
  • Occasional Test:: author
  • Can give free software-related talks about:
    • CPAN::Mini, especially regarding creating custom or private CPAN-like sites
    • ...and the reuse of CPAN tools for an internal (proprietary) code base
  • http://search.cpan.org/~rjbs/
 I think that is a laudable goal.  I think it would be excellent to get some
 people together to work on improving (a) TAP and (b) the existing tools for
 mucking about with it.
 
 Improving the install toolchain's expected behavior woulud be great all around
 (and would suggest asking Ken Williams if he's interested).  More heavily
 trusting META.yml for things like "bug tracker" would be great.  (Is Graham
 available for search.cpan.org improvements?  What about Brian Cassady for his
 search.cpan replacement that he's working on?)
 
 I'd love to talk more about re-use of the CPAN tools for internal
 organizations.  That is, I want a cpan-like repository of code, against which
 I can run cpants, CPANPLUS, a search.cpan-like interface for docs and bugs,
 and so on.  We (Pobox) are looking to explore this further, soon.
 
 I'd love to get more TAP use in other languages, even if only as "make
 existing tools capable of emitting this" or "write translator pipes."  My
 own PyTAP is  miserable, but has been useful.

Talk: XPAN - An External CPAN for Internal Distribution

Description
The Comprehensive Perl Archive Network (CPAN) is Perl's killer app, providing thousands of reusable libraries for nearly every problem space. It's an open system, with dozens of tools built around its very simple design. XPAN allows organizations to easily build their own archive of Perl code, using the structure of the CPAN and integrating their code with the actual CPAN and applying filters and constraints. The end result is an internal repository of code that can be used for deployment and testing, and which can benefit from all the tools built to enhance the core CPAN itself. Key modules may be smoked against customized XPAN sources, for example, to determine the way in which an upgrade would affect production.

It's the authors' hope that XPAN can also be used as the basis for CPAN-like archives entirely separate from the CPAN including for non-Perl projects like OpenJSAN.

Speaker
Ricardo Signes works for the venerable Pobox.com, making sure mail never, ever stops flowing. He and Hans Dieter Pearcey developed XPAN in 2008.
Audience
This project will appeal both to experienced Perl programmers who have customized development environments and to system administrators who must manage large Perl deployments.
Travel
Barring acts of God or major personal catastrophe, I will without a doubt be able to come to Oslo, if funded. I am in Bethlehem, PA, USA. I would probably fly out of Philadelphia, Newark, New York, or ABE.

Adrian Howard <adrianh@quietstars.com>

  • Author of Test::Class, Agile fanatic. User experience zealot. Testing bigot. CPAN Author. Guilt tripped into being a TPF Grant Manager. Adrian still hopes that Lisp and Smalltalk will take over the world, but in the mean time gets paid for playing with Perl and Ruby amongst other things. Location: Dorset, UK.
  • Can give free software-related talks about:
    • Test::Class and xUnit style testing (surprise!)
    • Test Driven Development
    • Why 100% test coverage isn't the point
    • Agile development
    • Integrating usability into agile teams
    • more five minute lighting rants than you can shake a stick at...
  • http://search.cpan.org/~adie/
  • Available to be in Oslo April 4th-10th
  • Emailed. Positive responese. Has committed, provided available funding.
 If you can get me there, more than happy to come.

Talk: Perl: A Culture of Transparent Quality

Description

Perl has developed a culture of transparent quality where practices (a focus on testing, excellent documentation, etc.) and tools (CPAN Testers, Anno CPAN, CPAN Forum, CPANTS, etc.) combine to allow everybody to view of the quality of Perl code. Making the advantages of open-source code that much

This talk gives a quick tour of some of ways that Perl infrastructure make the advantages of open-source development more transparent to everybody.

Audience

Folk with a technical background. No Perl experience needed.

Talk: Automating User Experience Testing

Description

Testing for "usability" is often seen as being something that can only be done by experts. I'll talk through various ways that you can use automated tests to do some of the slog work. Helping developers get the user experience right, and leaving your user experience team free to concentrate on the hard problems.

Audience
  • User experience folk with developer leanings.
  • Developer folk with user experience leanings.
  • People who have to manage developers and user experience folk.


Michael Peters <mpeters@plusthree.com>

Talk: Automated Build and Testing (BuildBot and Smolder)

Title
Automated Build and Testing (BuildBot and Smolder)
Description

Any large software organization should already have some sort of automated build and testing system. Usually it's a mix of shell scripts and custom programs. It get's even hairier when multiple machines and architectures are involved.

BuildBot is a continuous integration solution that allows you to easily setup an automated build system for your project (or projects). You can track various build metrics (warning counts, lint checks, image size, compile time) over time and see how various commits and branches compare to one another. It is language agnostic and supports lots of SCPs (CVS, Arch, Bazaar, Darcs, and SVN).

Smolder is test aggregation tool used to centrally report on automated tests. It uses the language agnostic TAP (Test Anything Protocol) and supports and testing infrastructure that can output TAP.

In this talk we'll look at how we can create a BuildBot setup that will automatically send it's reports to a Smolder server. Now we can easily test our project against multiple machines and multiple architectures. It's also a nice experiment in being pragmatic with language choice (Perl and Python) and in playing to both project's strengths.

Speaker
Michael Peters is the primary developer of Smolder. He manages the testing infrastructure for Plus Three, LP and loves trying to create easy testing systems where non programmers can provide testing input.
Audience
This talk will be useful to programmers, testing engineers and project manager. Any actual programming examples used will be fairly simple and be geared towards integration between BuildBot and Smolder.
Availability
100% provided funding can be arranged
Location
Silver Spring, MD USA


brian d foy <brian.d.foy@gmail.com>

  • Automated release testing with Module::Release
  • MyCPAN: Custom versions of CPAN
    • As part of the workshop, I want to fix up Module::Release to run `make test` with several different perls, perhaps some of which are remote.
    • I saw Ricardo's XPAN stuff on this wiki, and am working on something similar. I'd like to get together with him to see how we can join forces. :)

Talk: Automated Release Testing with Perl

Description
I maintain a many Perl modules, and I want to ensure that each time I release a new distribution it's thoroughly checked. Doing this by hand meant I'd often forget to do something or some something would slip by me. I created Module::Release to automate the process and to allow me to extend it as I found new things to check. Module::Release is superb at doing things the same way every time and not releasing something that is broken, which saves me a lot of time fixing things later. I couldn't maintain all of my modules without it.
Speaker
brian d foy is a partner at Stonehenge Consulting Services. He's the author of Mastering Perl, co-author of Learning Perl and Intermediate Perl, and publisher of The Perl Review.
Audience
Module::Release appeals to people who like to automate their work. Although I use it to release modules to CPAN, you can use it for whatever you like
Travel
I can come to Oslo if funded. I'd be traveling from Chicago.


Gabor Szabo <szabgab@gmail.com>

I moved the things I would like to work on to the Topics


Jos I. Boumans <jos@dwim.org>

  • CPANPLUS and CPANPLUS::Dist::* developer.
  • Can give free software-related talks about:
    • Perl Bad Practices (talk in doing for LPW this year)
    • XXX Perl (Talk I've done a few times, including for the NPW)
    • Something shorter on specifically the CPAN packaging initiative.
  • http://search.cpan.org/~kane/

Talk: Barely Legal XXX Perl

Description

This is the succesor of the talk '20 things you might not know about perl', which has been given at several other conferences and proved quite successful. So by popular demand here is its follow up.

The talk shows several features of perl you might not have known that existed, which are being (ab)used to run a program that was designed never to be able to run in the first place... It's a high paced, humourous and entertaining look at perls slightly less obvious features.

Note to organizers: this talk will go very well after Adam Kennedy's "Parsing, Analysing and Manipulating Perl (without perl)"

Speaker

Jos Boumans is an active Perl developer, counting over two dozen contributions to CPAN, including Archive::Tar and CPANPLUS and is responsible for 2% of all the code in perl5.10. He's also the driving force behind the CPAN packaging effort. He is a a regular speaker at Perl conferences on topics of perl deviance and toolchain.

Audience

This talk mostly targets people from a perl background of any level or familiar with other similar languages like Python and Ruby.

Preview

N/A. Talk has been given at among others the NPW 2007, LPW 2007, YAPC::EU 2007

Talk: The Good, the Bad and the Ugly

Description

Perl has a reputation of being unmaintainable and line noise. This talk delves into some good, bad and ugly practices, illustrated by Damian Conway's fantastic 'Perl Best Practices' book and the amazing 'Perl::Critic' module. This is another high paced, humourous and entertaining look at perls slightly less obvious features, this time from a QA angle.

Speaker

Jos Boumans is an active Perl developer, counting over two dozen contributions to CPAN, including Archive::Tar and CPANPLUS and is responsible for 2% of all the code in perl5.10. He's also the driving force behind the CPAN packaging effort. He is a a regular speaker at Perl conferences on topics of perl deviance and toolchai

Audience

This talk will be entertaining (and possibly useful) for anyone maintaining legacy code or who had the thought "It seemed like a good idea at the time".

Preview

N/A. Talk has been given at LPW 2007

Talk: Packaging the CPAN

Description

As an experiment, we have started http://debian.pkgs.cpan.org to automatically make certain CPAN packages available for debian.

As it turns out, this was a widely successfull experiment and we are now looking at adoptation in the packaging community. Debian is on the ball, who will follow suit?

Speaker

Jos Boumans is an active Perl developer, counting over two dozen contributions to CPAN, including Archive::Tar and CPANPLUS and is responsible for 2% of all the code in perl5.10. He's also the driving force behind the CPAN packaging effort. He is a a regular speaker at Perl conferences on topics of perl deviance and toolchain issues.

Audience

This talk will be of interest to people packaging perl modules for their favourite distrubtions and vendors alike

Preview

N/A


David Golden <dagolden@cpan.org>

  • Writer of Class::InsideOut, CPAN::Reporter, CPAN tester & Perl toolchain hacker
  • Former release engineer for Strawberry Perl
  • Can give free software-related talks about:
    • CPAN Testers and CPAN::Reporter
    • Inside Out Objects
    • Strawberry Perl and Perl on Win32 in general
  • Would be interested in hacking on
  • Emailed. Responded with interest.
  • Availability: 95% if at least partially funded, somewhat lower if not; depends on attendees -- if a lot of the CPAN Testers group is there to hack on CPAN Testers 2.0, then I'm more likely not to pass up the chance to work face to face

Talk: New Developments in Perl for Windows

Description
The 2007 Perl Survey found that 51% of respondents used Perl on Microsoft Windows at least some of the time. Historically, choices for Perl on Windows were limited; moreover, modules on CPAN varied widely in quality and support for Windows. However, the 2006 preview and 2007 release of the community-led Strawberry Perl for Windows drove a surge of new interest in fixing Perl on Windows and led to portability improvements in dozens of CPAN modules.
This talk introduces Strawberry Perl and will compare and contrast it with the free, but commercially-led alternative, ActivePerl. Other topics (depending on length) could include:
  • Tips and traps for Perl on Windows
  • Interesting/useful modules for Windows
  • Installing external library dependencies
  • Interacting with Windows applications
  • GUI toolkits
  • Resources online for further study
Length
The talk can be adapted anywhere from 20 to 45 minutes
Audience
The talk should appeal primarily to two groups:
  • those who already use Perl on Windows
  • those who fear they might someday need to
Speaker
David Golden was the release engineer for the preview version of Strawberry Perl. He is a prolific CPAN author, with over two-dozen modules released, including the groundbreaking CPAN::Reporter and Class::InsideOut. He has been a speaker at YAPC::NA, The New York Perl Seminar, and Boston.pm and has written articles for The Perl Review. He lives in New York City.


Sébastien Aperghis-Tramoni <saper@cpan.org>

  • Works with legacy support for CPAN modules.
  • Can give free software-related talks about:
    • How to maintain support for legacy Perls in current code (1)
    • How to find bugs thanks to tests (2)
    • Win32::Mock (3)
  • http://search.cpan.org/~saper/
  • Chatted on IRC. Has given positive feedback. No commitment.

(1) This is the talk I did this year at the NPW 2007 and at the YAPC::EU 2007. But it's not really about QA.

(2) This is more on the subject: I could talk about how I fully rewrote the test suite of Net::Pcap when I took its maintenance, and how it helped me find bugs. Same for Sys::Syslog when it was dual-lived and kind-of "handled" to me.

(3) It's a module I recently began working on because it's not easy for a Unix user to have access to a Win32 environment with all the toolchain. The aim is to provide a way to emulate most of Win32 API throught mocked versions of the Perl modules, and allow for easy instrumentation, which might even interest actual Win32 users.

But I'm far from being an expert wrt QA. I really want to attend this workshop, but more to learn than to teach ;)

No dates conflict right now so you can count me in, at least as attendant.


Anton Berezin <tobez@tobez.org>

 I'd be glad to come.  It is not a problem at all if the funding is available
 and we have a date at least 2-3 weeks in advance;  failing the funding, it
 is still possible that I'll come - let me talk to my employer first.
 
 I think the idea is nice, but it might be difficult to put people on the
 same footing. In the context of QA I am most interested in FreeBSD ports -
 not only of Perl itself, but also of various modules in the ports collection
 - we currently have almost 3000 of those. [...]
 I'd like to, but I did not discuss it with my employer yet;  I'll do that
 next week...  Or shall I say so in the wiki?  ;-)


Barbie <barbie@missbarbell.co.uk> (Withdrawn)

Talk: Distributed Testing of CPAN

Description
The Comprehensive Perl Archive Network (CPAN) is a repository of Perl code packages. In order to provide some credibility to these packages it is encouraged that authors upload tests within the packages. Theses test can be run during installation by the user to ensure that the functionality of the code is applicable and appropriate for that installation environment. In order to provide feedback to authors and pre-warn users of any potential issues, the CPAN Testers, numbering around 100 people, run automated testing on the CPAN packages over several hundred different installation environments, then publicly report on the outcomes.
With 100+ testers producing 80,000+ test reports per month, this talk will explain in more detail how CPAN Testers works, the methods and applications used, how we have implemented automated testing and the benefits that been gained from such a globally distributed network of testers.
Length
The presentation typically lasts for about 40 minutes. This can be reduced or extended as needed.
Speaker
Barbie works for MessageLabs, the largest independent messaging security company in the world today. Barbie has been involved in CPAN Testing, both as a tester and toolchain developer, since 2003 and now acts as a caretaker to the CPAN testing community to ensure that authors and testers use CPAN Testers appropriately.
Audience
This presentation will appeal to anyone with an interest in distributed testing and/or automated testing within many different installation environments.
Location
Birmingham, UK.
Availability
No longer available.


Philippe 'BooK' Bruhat <philippe.bruhat@booking.com>

  • Works on QA at Booking.com
    • Will come if send by his company
  • Can talk about:
    • How to test Apache handlers without Apache
    • Running smokes with TAP::Parser


Marcus Ramberg <marcus@nordaaker.com>

Catalyst Release Manager

Guess I'll be showing up as I live nearby. I'll be interested in doing things related to QA and automated testing and Catalyst. Specially GUI based testing, and making it effortless to test Web Apps with Catalyst and Selenium.

   22:42 <@marcus> jrockway: is Test::WWW::Selenium::Catalyst  good?
   22:42 <@jrockway> marcus: not really, too hard to get working
   22:42 <@jrockway> for example, i can't use it because firefox isn't in my PATH, 
                 and that chokes it
   22:43 <@jrockway> i have seen it work, though, so it is theoretically possible
   22:43 <@jrockway> just not easy

Andreas Marienborg <andreas@startsiden.no>

  • Does alot of work on automated QA as a part of his job.
  • Wants to work on automating the build process from sourcecode to deployment.

Mark Overmeer <mark@overmeer.net>

I worry about the quality of major components and important modules in our Perl environment. When the projects grow larger (and because of the speed increase of computers, you can write much larger Perl programs without penalty than a few years ago), the needs change as well.

  • CPAN6, rethinking CPAN; an alternative data distribution infrastructure has been designed, and the implementation is underway. It might once be used to install Perl6 (or even Perl5) modules, but first has to prove itself. It has many other applications.
  • XML::Compile and XML::Compile::SOAP, a new SOAP::Lite alternative; using XML correctly is too complicated (for Perl programmers) with strict data-types and name-spaces. XML::Compile removes that burdon and is close to a 100% implementation. Do XML automatically correct.
  • Log::Report, combining Log::Dispatch, Locale::TextDomain and exceptions; this tricky module adds three things at once. Translations are delayed to the IO-layer, like charset encodings, which has many advantages.
  • OODoc and Perl6 documentation system discussion; Perl6 needs more than a cleaned-up POD syntax.

http://search.cpan.org/~markov/ http://perl.overmeer.net

CPAN6

Last two years, a new software distribution system is being developed, which may become a follow-up for CPAN, but may be not: it first has to prove itself. By abstracting the distribution process, many existing data sharing schemes came together: sharing Perl modules, sharing photos within a family, maintaining ftp-servers, document-flows in a company.

CPAN is beautiful in its freedom of sharing, but should adapt better to changing world. Were will we store our Perl6, Parrot, and Parrot related code? How do we publish results of additional services, like cpan-testers? How do we guarantee that the author of the new version of a module is the same as for the old version? Perl6 has a different versioning structure, which CPAN cannot handle. Quite some special case situations currently require manual intervention from Andreas.

CPAN6 has a few very new sharing concepts, afaik not available in any existing application. For the Perl community, it is important to understand that it is only about the distribution process of packages: from upload by the author (pause) until download by the user. It is not re-inventing packaging or installation tools, but on the level of transport protocol.

When developing a large new system, with wishes for pluggable modules contributed by various people, you discover that Perl5 is actually quite poor. CPAN6 development therefore first focused on general purpose building bricks:

  • XML::Compile and XML::Compile::SOAP, because the meta-data will be XML based
  • Log::Report, because translation must be simple
  • OODoc, to nicely integrate (especially complex object oriented) manual pages

There are also a few hundred pages of detailed design documents available from http://cpan6.org There are also a few presentations on the subject, 30, 45 and 90 minutes long.

Jonathan Worthington <jnthn@jnthn.net>

  • Perl 6 and Parrot Hacker
  • More than happy to help people digging in to contributing to the Perl 6 spec test suite
  • Will spend time working with Mark on CPAN6
  • Happy to join any discussions/hacking where I'm of use; otherwise, I'll happily use the time and space and lack of $DAYJOB distractions to hack on Rakudo

Lars Thegler <lars@thegler.dk>

  • FreeBSD Perl packaging guy
  • I'll be working on improving the handling of dual-life modules (core and CPAN) in FreeBSD ports, together with tobez and des.

jonasbn <jonasbn@gmail.com>

  • Perl Hacker
  • Has experience with Module::Build and Test::Builder, interested in automated smoke testing
  • Can give talk on: Module::Build and Test::Builder

Cosimo Streppone <cosimo@cpan.org>

  • I'll be happy to work on Perl6 test suite refactoring
  • Maybe team up with JonathanW ?
  • Other random things: Test::Class, Selenium, PITA, bringing last Win32::API release into Strawberry Perl :-)

Nadim Khemir <nadim@khemir.net>

  • Help and learn at the same time is the goal. Teach me and I'll slave till late night
  • I like good documentation
  • ASCII diagrams are cool
              \o/
               | ----.
              / \    |     (\_/)
     \o/             |-----(O.o)
      |  ------------.     (> <)
     / \             |
.--------------------|------------.
| Norge .------------|----------. |
|       | Oslo       |          | |      \o/
|       |            .------------------  |
|       |            |          | |      / \
|       |            v          | |
|       |  .-------------------.| |
|       |  | Perl-QA hackathon || |
|       |  '-------------------'| |
|       '-----------------------' |
'---------------------------------'

Dag-Erling Smørgrav <des@des.no>

  • FreeBSD committer, author and maintainer of the FreeBSD tinderbox
  • Currently working on Light My Fire, the next generation of the FreeBSD tinderbox

Trond Michelsen <trondmm-qahack08@crusaders.no>

  • Perl Hacker at Norwegian Meteorological Institute

Øystein Torget <oysteto@ifi.uio.no>

  • Perl developer at DNV
Personal tools