10 Oct

EuroStar eBook "A Lucky Shot at Agile", reviewed by Shmuel Gershon

10 October 2010 by Shmuel Gershon

Hello all!
For this weekend's read, I downloaded and printed the (free!) "A Lucky Shot at Agile" eBook written by Zeger Van Hese. It is a best-paper award piece from last year's EuroSTAR, and recently a webinar too. I had to miss the webinar due to scheduling conflicts, so having it on book now is a good opportunity to catch up. I will write a little bit about the eBook here, but I recommend you still read the full story on this link (this is a review only, not a summary).

The review conclusion is that I enjoyed the book a lot. Here are the details:

HITS: CONTENT

The content is very interesting. The testing project was a revamped version destined to replace an old big product. Those of you who already worked in a "completely changed version of an old big product" know how daunting a task it is: all the difficulties of "new", but with the same obstacles related to "old" too (and don't get me started on copy-pastes...). It seems from the book that they did a great job and customers+management were satisfied.

The book has a short introduction to Agile, just enough to understand the story. So if you want to learn about Agile, you have plenty of resources on the web or the "Selected References" chapter of the eBook.

You can't really disagree with the content, as it is Zeger's personal account of the facts and their interpretation.
Well, maybe we can argue a bit on interpretation… And one of the parts where I dissent is where Zeger considers "Manual tests still used" as one of the problems. I don't agree due to personal preferences for manual (the sapient/ brainual type) tests, but also because the paragraph expanding this part talks about coverage traceability and not really about manual tests.

There are some points in the story that sent me on side wanderings...

  • One of them is the depiction of their very organized company, that had all the process management tools necessary as part of the ISO13485:2003 certification, but where the requirements management system was not used by the programming team and the requirement traceability to the testing management system was ignored after a glitch. Makes you think about the real value of certification and their criteria.
  • Another was this statement: "For every agile practice we had embraced before, there seemed to be another practice that was carelessly neglected or even abandoned" (the phrase made me stand up from the chair). Very well said, and I feel it happens a lot (although it's hard to keep track). I'm pretty sure I've read a very similar account before, just can't point it. Why does this happen? Maybe it is because after feeling the benefits of the new processes people feel like they can drop the benefits of the other ones (in a risk homeostasis way)? Maybe it is because there's a limit to the quantity of processes a people would like to carry with?


HITS: FORMAT

It's got great readability and flow, maybe the most important attribute after content.
I had read Zeger's texts at his blog and knew he writes well, but writing a short blog post is not the same as writing a 17 page booklet. Zeger stood up to the challenge very well.

The story telling is compelling and engaging, you follow the development team (managers, testers and programmers) on its evolution, doubts and progress. As said earlier, I literally stood up from my chair and read standing up when reaching the beginning of chapter 8 (where problems started to happen).

Visuals are nicely done, very good design and pagination that makes reading easier. The 'tweetable' sections are a creative way of denoting block quotations :).

MISSES

There are some bugs in the eBook... Some of them minor (typos, formatting... which I'll write Zeger directly about), and some more important ones. I'll expand on these ones:

  • The story includes a transition from many prescripted tests to exploratory based test activities. It would be nice to have more details on how this was presented, executed and received.
  • Another thing I miss is stories on the regression suite. At one moment Zeger comments that "the automated regression tests proved to be invaluable once more", but it's unclear if the value was due to the sense of coverage, or due to actual bugs uncovered.
  • Perhaps the biggest point missing is a section about the author. Zeger's name appears in the cover but there's nothing more about him anywhere... They should've added an "About the author" at the end, even a short one like this one adapted from his blog's about page: "Zeger Van Hese lives in Belgium. He started out in IT in 1999 and rolled into software testing in 2000. His passions include exploratory testing, testing in agile projects and, above all, continuous learning from different perspectives."


So if you want a simple first person account of Agile transition, go, read the book!
And then leave your comments and questions here, I'll send the link to Zeger and am sure he will be reading and answering them.

. Shmuel

 

6 comment(s) for “EuroStar eBook "A Lucky Shot at Agile", reviewed by Shmuel Gershon”

  1. Gravatar of space jams for sale
    space jams for sale Says:
    But, boy oh boy we are also very sad that you both have to go through this so tough time. I know you will pull through, but it is also okay to cry. Louis. Coach Terry Murray said Clifford had a black eye, but his injury is "nothing out of the ordinary." .Wow, and if you're wondering about my Fakes, I have since contacted the buyer and the auction site. This rather classified as being "odd" and "different" form of Nike Jordan is truly one of a kind for its own distinctive reasons. It's because they know more than most of their friends, and might even be able to win a fight or two. Part of your training is to learn how to deal with defeat and not let it defeat you. space jams for sale http://www.google.la/webmasters/tools/richsnippets?url=run2013.org/ceshi12.html
  2. Gravatar of testwithus
    testwithus Says:
    SWIFT Interview questions on

    http://testwithus.blogspot.in/p/swift.htm

    For selenium solution visit
    http://testwithus.blogspot.in/p/blog-page.html


    For QTP interview questions

    http://testwithus.blogspot.in/p/qtp-questions.html


    www.searchyourpolicy.com


  3. Gravatar of Adam Brown
    Adam Brown Says:
    I definitely recommend watching the Webinar if you've not already done so. Found it both funny and educational!

    http://www.eurostarconferences.com/community/member/webinar-archive/webinar-23-a-lucky-shot-at-agile.aspx
  4. Gravatar of Shmuel Gershon
    Shmuel Gershon Says:
    Thansk for the link, Adam!
    It's a nice companion to the eBook.
    Now people have the choice -- do they see the movie or they read the book? :)
  5. Gravatar of Zeger Van Hese
    Zeger Van Hese Says:
    Hi Shmuel,
    Thank you for taking the time to read and review the e-book. I included my answers to your remarks below:

    >> The story includes a transition from many prescripted tests to exploratory based test activities.
    >> It would be nice to have more details on how this was presented,
    >> executed and received.

    We replaced all these existing test scripts we used to do manually (our old regression set) by a carefully chosen set of automated tests.
    We performed exploratory testing on all the new functionalities once they became (even partly) available. It wasn't full-blown session-based testing,
    but most of the time we *did* have a mission beforehand. This helped us to focus. There were no formal test reports in place anymore, we communicated issues and findings directly.


    >> One of the parts where I dissent is where Zeger considers "Manual tests still used" as one of the problems.
    >> I don’t agree due to personal preferences for manual (the sapient/brainual type) tests,
    >> but also because the paragraph expanding this part talks about coverage traceability and not really about manual tests

    Our exploratory tests were the "real" manual, sapient tests. My personal preference is also for the sapient/brainual tests, hence our switch to exploratory as main approach.
    The manual, scripted tests I talked about in the "misses" were NO sapient tests,
    but mainly high-level checks written in Quality Center to be able to flag off the tests as passed
    so our coverage module in Quality center would be updated.
    We also used these as acceptance tests afterwards, because they neatly listed all new features without too much distracting details.

    >> At one moment Zeger comments that "the automated regression tests proved to be invaluable once more",
    >> but it's unclear if the value was due to the sense of coverage, or due to actual bugs uncovered.

    Actually, the value came of both.
    Most importantly, the regression tests *did* find a lot of bugs during refactoring, and also when functionality was incrementally being added.
    I remember that that surprised me a little bit at the time since I had also heard stories about the pesticide paradox,
    but I guess that is something that primarily comes into play when always running the same tests on code that isn't changed.
    Concerning the sense of coverage, this was probably more of a mental thing. We had a set of automated tests that covered the most important areas, and
    we gave the areas that had proven more tricky in the past more attention.

    About the author thing, good point.
    The main reason is probably because my paper was transferred directly into the e-book. I think there is nothing written about me in the paper, too.
    So, thank your for including some credentials :-)

    -- Zeger
  6. Gravatar of Shmuel Gershon
    Shmuel Gershon Says:
    Thanks Zeger for taking the time to write these answers.
    They complement the review very well (and I think the book as well).

    Congrats again for the book!
    Shmuel

Leave comment:

Name:  
Email:  
Website:
Comment: